PB85-102523
Swirl and Helical Bend
Regulator/Concentrator for Storm and
Combined Sewer Overflow Control
Environmental Design and Planning, Inc.
Boston, M4
Prepared for
Municipal Environmental Research Lab.
Cincinnati, OH
Sep 84
I
J
U.S. topa trait of Commerce
(tetioml T4ctafci! Information
-------
EPA-600/2-84-151
September 1984
PB85-102523
Swirl and Helical Bend Regulator/Concentrator for
Storm and Combined Sewer Overflow Control
by
Will1am C. Plsano
Daniel J. Connlck
Gerald L. Aronson
Environmental Design & Planning, Inc.
Hanover, Massachusetts 02339
Grant Nos. S-805975 and S-802219
Project Officer
Richard Field
Storm and Combined Sewer Program
Wastewater Research Division
Municipal Environmental Research Laboratory (Cincinnati)
Edison, New Jersey 08837
MUNICIPAL ENVIRONMENTAL RESEARCH LABORATORY
OFFICE OF RESEARCH AND DEVELOPMENT
U.S. ENVIRONMENTAL PROTECTION AGENCY
CINCINNATI, OHIO 45258
-------
TECHNICAL REPORT DATA
f liuuut Irani on the rerene before completing)
(Please read /au/ueti'jnt on
I REPORT NO.
EPA-600/2-84-151
3. RECIPIENT'S ACCESSIOWNO.
10259?
«. TITIE AND SUBTITLE
"Swirl and Helical Send Regulator/Concantrator for
Storm and Combined Sewer Overflow Control"
B. REPORT DATE
September 1984
B. PERFORMING ORGANIZATION CODE
7. AUTHORIS)
William C. Plsano, Daniel J. Connlck, and Gerald L.
Aronson
B. PERFORMING ORGANIZATION REPORT NO.
». PERFORMING ORG 1NIZATION NAME AND ADDRESS
Environmental Design & Planning, Inc.
353 Circuit Street
Hanover, Massachusetts 02339
1O. PROGRAM ELEMENT NO.
11. CONTRACT/OR ANT NO.
S-805975, S-802219
12. SPONSORING AGENCY NAME AND AOORESS
Municipal Environmental Research Laboratory-C1n., OH
Office of Research and Development
U.S. Environmental Protection Agency
Cincinnati, Ohio 45268
IX TYPE Of REPORT AND PERIOD COVERED
14. SPONSORING AGENCY CODE
EPA/600/14
IS. SUPPLEMENTARY NOTES
Project Officer: Richard Field, Chief, Storm and Combined Sewer
Program, Edison. NJ 08837; Coml. (201) 321-6674; FTS 340-6674
IB. ABSTRACT
*
Swirl and helical bend devices were studied for tiiree years at Lancaster. PA. and West
Roxbury In Boston. HA. At Lancaster the study Included:
o a full-scale swirl regulator/solids concentrator (SRC) for combined sever overflow.
(CSO) control (24-ft (7.3-m) diameter) and
o a swirl degrltter for SRC foul underflow (8-ft (2.4-m) diameter).
At West Roxbury the study Included:
o a pilot-scale SRC for separate urban stormwater treatment (10.5-ft (3.2-m) dia-
meter) and
o a pilot-scale helical bend regulator/solIds concentrator (H8RC) for separate
urban stormwater treatment (60-ft (18.3-m) long).
Data from the Lancaster facility Indicated that the SRC Is an efficient treatment
device to remove heavier or "first flush'-related suspended solids and grit.' Treatment
efficiencies usually exceeded 601 for flows exceeding 20 cfs (566 l/s). The swirl de-
grltter did net function properly due to clogging of the bottom hopper apex.
' At West Roxbury. efficiencies were low for both units and appear to be related to
very low settling velocities of the suspended solids.
The full report was submitted In fulfillment of Grant Nos. S-80S975 and S-802219.
by Environmental Design S Planning. Inc.. under the sponsorship of the U.S. Environment-
al Protection Agency.
17.
KEY WORDS A*O DOCUMENT ANALYSIS
DESCRIPTORS
b. IDENTIFIERS/OPEN ENDED TERMS C. COS AT I Field/Group
IS. DISTRIBUTION STATEMENT
RELEASE TO PUBLIC
19. SECURITY CLASS (TallReport)
UNCLASSIFIED
21. NO. OF PAGES
36 f
SECURITY
UNCLASSIFIED
22. PRICE
EPA Form Z23O-1 (9-71)
-------
DISCLAIMER
"Although the Information described
-------
FOREWORD
The Environmental Protection Agency was created because of
Increasing public and government concern about the dangers of
pollution to the health and welfare of the American people.
Noxious air, foul water, and spoiled land are tragic testimony to
the deterioration of our natural environment. The complexity of
that environment and the Interplay between Its components require
a concentrated and Integrated attack on the problem.
Research and development Is that necessary first step In
problem solution and it Involves defining the problem, measuring
Its Impact, and searching for solutions. The Municipal
Environmental Research Laboratory develops new and Improved
technology and systems for the prevention, treatment, and
management of wastewater and solid and hazardous waste pollutant
discharges from municipal and community sources, for the
preservation and treatment for public drinking water supplies and
to minimize the adverse economic, scclal, health, and aesthetic
effects of pollution. This publication Is one of the products of
that research, a most vital communications link between the
researcher and the user community.
The deleterious effects of storm sewer discharges and
combined sewer overflows upon the nation's waterways have become
of Increasing concern In recent times. Efforts to alleviate the
problem depend In part upon the development of Integrated
technologies Involving non-structural best management practices
with structural storage and treatment concepts.
This report presents the summary results of a three year
field-oriented data collection and analysis effort aimed at
evaluating the feasibility and efficiency of flow
regulator/concentrators for control of urban stormwater
pollution. The practice of designing regulators solely for flow
rate control or diversion of combined wastewaters to treatment
plants and overflow to receiving waters has proved costly and
Inefficient. Sewer system management which emphasizes the dual
function of qualitative and quantitative control of storm Induced
water pollution has proved highly cost-effective under several
separate evaluation studies. The use of the Swirl
Regulator/Concentrator and Helical Bend Regulator for control of
both separately sewered urbar stormwarer and combined sewer
overflows (CSO's) have resulted In removal efficiencies
comparable to those achieved through the use of conventional
settlIng tanks.
Francis T. Mayo
Director
Municipal Environmental Research
Laboratory
iii
-------
ABSTRACT
This report summarizes the results of a three year study on
the applicability of full scale Swirl Regulator/Concentrators and
Helical Bend Regulators for qualitative and quantitative control
of pollutant loads released through storm discharges and combined
sewer overflows (CSO). Two separate studies were Incorporated
Into this report; I) an ex am (nation of the feasibility of using
the Swirl Regulator/Solids Separator and Helical Bend Regulator
for wastewater treatment of a separate storm sewer serving an
urbanized area In West Roxbury, Ma., and 2) a posf'constructIon
evaluation of an existing U.S. Environmental Protection
Agency/City of Lancaster, Pa. Swirl Regulator/Concentrator and
Swirl D.grltter complex located In Lancaster, Pa. for the
abatement of CSO pol lutlon. This report does summarize earl ler
design/construction details of the demonstration facility at
Lancaster.
The West Roxbury, Ma. Swirl, Helical Bend complex evaluation
was divided Into three separate phases. The first phase Involved
site Inspection and runoff flow evaluation. Phase two Included
design, fabrication and Installation of the two fullscale
treatment devices while the monitoring, evaluation, data
reduction and comparative assessments were Included In phase
three. Detailed site Inspections, low level flow monitoring
effort and detailed computer modeling to ascertain typical runoff
flows resulted In the design of a 10.5 ft (3.2 m) diameter Swirl
Regulator/Concentrator and 60 ft (18.3 m) long Helical Bend
Regulator with design flows each of 6 cfs (173 l/s) and maximum
capacity of 12 cfs or (346 l/s). Foul sewer underflow from each
unit was regulated by a Hydro-Brake at a cfs of 3) of the design
flow or 0.18 cfs or (.85 l/s). The second phase Involved the
fabrication and Installation of the units near an outlet of a
separate storm sewered 160 acre urbanized area. The stormwater
flow at the outlet was divided and Input Into Swirl and Helical
Bend Regulator/Concentrators constructed side-by-slde.
In the third phase of work, the two treatment units were
monitored during a 22 month period for approximately 15 separate
storm events covering a wide range of antecedent dry conditions,
rainfall Intensities and durations. This phase Included I) an
assessment of the hydraulic Influent/foul sewer waste stream
characteristics 2) testing the feasibility of foul sewer
discharge to an adjacent sanitary sewer for removal of waste
stream effluents, and 3) assessment of the I nfIuent,effIuent and
foul stream characteristics from both devices with respect to COD
and sol I ds.
1v
-------
Data reduction and desk-top analytical efforts resulted In a
comparative assessment of pollutant removal effectiveness of the
two devices. Suspended solids removal efficiencies observed for
the units ranged between 5% - 35< depending on the
characteristics of the event sampled.
The Lancaster, Pa., study Involved three phases of work for
demonstrating the applicability of the full scale swirl concept
for ?SO control. Planning, design and construction of the
facility was completed IP the first phase of fork. A 24 ft (7.3
m) diameter Swirl Regulator/Concentrator was constructed along
with an 8 ft (2.4 m) diameter Swirl Degrltter. Swirl
Regulator/Concentrator was designed for a design of 40 cfs (1132
l/s) with capacity for handling
-------
U.S. Environmental Protection Agency and the City of Lancaster,
and the final phase including report preparation was by Environ-
mental Desiqn and Planning, Inc. _ .
This reoort covers the period March 1979 to December 1981 and
work was completed January 1982.
vi
-------
TABLE OF CONTENTS
Foreword iii
Abstract iv
List of Figures xi
List of Tables xviii
Abbreviations and Symbols xix
Acknov I edgements xxi
Chapter I Introduction I
1.1 Foreword I
1.2 Background-The Urban Runoff Problem..... I
1.3 The Need for Improved Regulators 4
1.4 Historical Background and Purpose of
Lancaster Facility 10
1.5 Historical Background and Purpose of
West Roxbury Facility 15
Chapter 2 Conclusions. 20
A Technical Performance Conclusions
- Lancaster ..........20
B Instrumentation/Sampling Techniques
- Lancaster 22
Chapter 3 Recommendations 23
A Design - Lancaster 23
B Instrumentation/Sampling Methodology 25
Chapter 4 Descriptions of Pertinent Treatment Units 27
4.1 Foreword 27
4.2 Integral Parts of Swirl Regulator
/Concentrator Design 27
4.3 Swirl Degrltter 35
4.4 Helical Bend 40
4.5 Dlscostralner 40
4.6 Teacup Regulator/Concentrator 43
4.7 Dynamic Solids Separator 45
Chapter 5 Study of Area Description
- Lancaster Swirl Facility 60
5. 1 Foreword , 60
5.2 Area and So*age System Characteristics 60
5.3 Monitoring Results 61
vii
-------
Chapter 6 *tudy Area Description
West Roxbury Swirl Facility 66
6.1 Foreword 66
6.2 Study Area Selection - Background 66
6.3 Study Area Description 68
6.4 Analysis of Long-Term Rainfall Records ..68
6.5 Hydrologlc Mode'Ing of Catchment Area 73
Chapter 7 Design Phase - Lancaster Svlrl Project 85
7.1 For en or d . .......................... ............85
7.2 Design Considerations & Constraints.... 85
7.3 Site Characteristics 85
7.4 Hydraulic Design 86
7.5 Final Swlri Concentrator
Design Considerations 91
7.6 Swirl Degrltter Design Details 93
7.7 Other Facility Design Factors 93
7.8 Hydro-Brakes 96
Chapter 8 Design - West Roxbury Facility 98
8.1 Foreword 98
8.2 Site Description 98
8.3 Rainfall/Runoff Design
Flow Considerations 98
8.4 Swlrl/Kelleal Bend
Design Dimensions 99
8.5 Miscellaneous Design Details 99
Chapter 9 Construction - Lancaster Project 103
9.1 Fo: ewor d.. . s .103
9 .2 Genera I 1 03
9.3 Special Considerations 103
9.4 Construction Cost 112
9.4.1 Operation and Maintenance Costs 115
Chapter 10 Construction - West Roxbury Project 116
10.1 Foreword 116
10.2 Historical Overview 116
10.3 Site Preparation 117
10.4 Swirl Concentrator Fabr(cat)on/InstalIatlon..119
10.5 Helical Bend Fabrication/ I nstal I atlon 127
lO.o Influent Structure Appurtenances 134
10.7 Effluent Appurtenances 137
10.8 Security Measures 140
vi 11
-------
10.9 Measurement Devices 143
10.10 Simulated Combined Sewer Lift Station 143
Chapter 11. Evaluation Methods
11.1 Foreword 1 44
11.2 Evaluation Program Overview
- Lancaster Swirl Project
initial EOP Review and Project Startup....144
11.2.1 Details of Sampling Locations
at Lancaster 148
11.2.2 Initial Monitoring Problems 150
11.2.3 Specialized EDP Automatic
Influent Sampling Devices 151
11.2.4 Lancaster Swirl Evaluation
Project Flow Meter Calibration 158
11.4 Procedure for Detirnlnation of
Settleabi I Ity Characteristics 172
11.4.1 Settling Column Design 172
11.4.2 Lancaster Settling Column 179
11.5 Analytical Methodology 179
11.6 Efficiency Calculation Procedures 180
11.7 Settleabi I Ity Curve Manipulations 182
Chapter 12 Lancaster Swirl Evaluation 187
12.1 Foreword 1 87
12.2 Summary of Lancaster Evaluation ...137
12.3 Description of Each Sampling Location 189
12.4 Comparison of Swirl Influent Concentrations
Using the Mannnlng 6000 Unit and the EDP
Technologies Cross-Sectional Sampler 196
12.5 Detailed Event Analysis
- Swirl Concentrator 197
12.6 Reassessment of Early Program Data 224
12.7 Ancillary Sampling Considerations
Swirl Concentrator Tank Samples 226
12.8 SettleabiIIty Experiments -
Sw I r I/Concentrator Regul a tor 227
12.8.1 Comparative Summary of Swirl Influent
SettleabllIty Characteristics with
Discrete Sampler Results 227
12.8.2 Comparison of Theoretical with
Actual Performance Removal 231
12.9 Overview Summary
- Swirl Concentrator Performance 237
12.10 Degrltter Operation Summary 238
12.11 Lancaster Dl scostral ner Evaluation 245
12.12 Hydro-Brake 247
ix
-------
Chapter 13 Hast Roxbury Evaluation ?50
13.1 Foreword 250
13.2 Evaluation Overview 250
13.3 Detailed Event Analysis 258
13.4 Settleabl I Ity Experiments 315
13.4.1 Influent Settleab11Ity Characteristics 316
13.4.2 Efficiency Determination Using
Sett I eab M Ity Results 320
13.5 Performance Summary.......... 335
13.6 Dye Studies 335
References • .339
-------
LIST OF FIGURES
Number Page
1 Helical Bend and Swirl Concentrator. 6
2 Lancaster Swirl Project facilities layout. 12
3 Schematic flow diagram of Lancaster Swirl Project. 13
4 Swlrl/Helleal Bend site plan, W. Roxbury Project. 16
5 Isometric view of Swirl Regulator/Concentrator. 28
6 Comparison of sewage settling velocity
characteristics used In San Francisco
Swirl evaluation with APWA Results. 32
7 Isometric vtew of Swirl Concentrator as a
grit separator. 37
8 Construction and operating features of Dlscostralner. 41
9 Features of the Teacup Solids Separator. 44
10 Photographs of Vortex Separator constructed
In Bristol, England. 46
11 Hydro Dynamic Separator. 48
12 Photograph showing top view of Vortex Separator
and sludging Holding Tank, Pilot Plant,
Blackwoll (UK). 50
13 Photograph of Vortex Separator In operation
Blackwell (UK). 51
14 Drainage districts within the CMy of Lancaster,
Pennsylvania. 62
15 Stevens Avenue drainage district And
Swirl Project Site. 63
16 Lancaster Swirl Project site plan. 64
I/ Locatlonal map of West Roxbury Swirl site. 69
18 Land use map - West Roxbury demonstration project. 70
19 SWMM discretization. West Roxbury catchment area. 75
20 Stage/rating curve, 30 In. RCP drain,
New Haven Street. 77
21 Storm 1:7/4/78 Hydrographs. 82
22 Storm 2:5/15/78 Hydrographs. 83
23 Calculated combined sewer flows at swirl site vs
occurances - % less than or equal to. 88
24 Chamber diameters for 90% recovery. 89
25 Settleable solids recovery for Dl
'•1.5 cm. (3 ft.). 89
26 Genural Swirl Regulator/Concentrator design details. 90
27 General Swirl Degrltter design dimensions. 94
28 Chamber diameters for 90% recovery. 95
29 Swirl/Helical Bend Regulator ralnfalI/runoff
characteristics W. Roxbury demonstration facility.100
30 Helical Bend Dimension?, W. Roxbury. 102
31 Initial construction. 104
32 Lancaster Swirl Project. 105
33 Swirl floor and gutter construction. 106
xi
-------
Number
Page
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
Photographs
Photographs
concrete
Photographs
Vlev of gutters In Swirl floor.
Swirl Inlet configurations.
Emergency overflow weir.
Weir and floatables trap assembly.
The Swirl Oegrltter.
Photographs of completed facilities.
of site excavation.
of Influent Diversion Chamber
demo I Itlon.
of Influent Diversion Chamber
- sldewall construction details.
Photographs of Influent Gates and Flumes.
Photographs of shop fabricated
-Swirl Solids Separator components,
W. Roxbury.
Photographs of field assembly
-Swirl Sol Ids Separator, W. Roxbury.
Photographs of completed Swirl Solids
Separator W. Roxbury.
Photographs of Helical Bend wood fabrication.
Photographs of Helical Bend Regulation,
wooden fabrication details.
Photographs of Helical Bend wooden shell
(In progress).
Photographs of Helical
(compl eted) .
Photographs of Helical Bend
I nter lor .
Photographs of Helical Bend
Interior.
Photographs
Photographs
Bend woodan shell
aluminum sheeting
aluminum sheeting
of
of
Helical Bend on-slte Installation.
final surface preparation
for Helleal Bend.
Photograph of 8 In. x 10 In. 90o flanged
Hydro-Brake. (Helical Bend Foul Sewer).
Photographs of Foul Sewer Pump.
A. Top View of Foul Sewer Pump
B. Trash Pump
Photographs of control building, W. Roxbury.
Photographs of slteworK.
Lancaster Swirl Project sampling locations.
Diagram of 36 In. Influent Sampling Device (typical
Photographs of EDP Technologies Cross-Sectional
Sampler Installations.
A. Inserting Inflatlble Dam In Swirl Influent
B. Cutting Segment of 36 In. Swirl Influen* for
placing New Sampler.
108
109
110
111
113
118
120
121
122
123
124
125
126
128
129
130
131
132
133
135
136
138
139
141
142
149
).152
154
xil
-------
Number Paga
62 Photographs of EDP Technologies, Inc.
Cross Sectional Samplers.
A. EDP Cross-Sectional Swirl Influent
Sewer Sampler - Lowered Into Basement Floor.
B. EDP Cross-Sectional Swirl
Foul Sewer Sampler Installed. 155
63 Photographs of EDP Technologies
Cross Sectional Samplerst
A. Partial View of 36 In. EDP Technologies
Cross Sectional Sampler.
B. Partial View of 36 In. EDP Technologies
Cross Sectional Sampler. 156
64 Photographs of EDP Technologies, Inc.
Cross Sectional Samplers.
A. EDP Cross-Sectional Swirl
Foul Sewer Sampler - Completed.
B. Inner Assembly EDP Cross-Sectional
Swirl Foul Sewer Sampler. 157
65 Influent flow record, Lancaster, 4/1/81. 160
66 Influent flow record, (continuation) Lancaster,
4/1/81 161
67 Plot of Swirl level vs Mapco meter flow (10/2/80). 163
68 Plot of Hapco Meter Flow vs estimated flow (5/15/81).165
69 Plot of adjusted Mapco meter flow vs estimated flow*
(10/2/80). 167
70 West Roxbury Swirl Site sampling locations. 170
71 Front and side views of motorized settling columnn. 174
72 Photographs of motorized settling column. 175
73 Photographs of settling column mechanical details. 176
74 Photographs of settling column sampling operations. 178
75 Settling Column results, Lancaster (7/2/81). 183
76 Suspended solids remaining vs settling velocity, 184
Lancaster Cwlrl (7/2/81).
77 Normalized suspended solids remaining vs settling
velocity, Lancaster Swirl (7/2/81). 186
78 Suspended solids concentrations vs time,
Lancaster Swirl (6/22/81). 198
79 Comparison of Swirl Influent suspended solids
concentrations using EDP Technologies Cross
Sectional Sampler with Manning Sampler. 199
80 Suspended solids cone, vs time,
Lancaster Swirl Concentrator/Regulator (9/10/80). 201
81 Volatile suspended solids cone, vs time, Lancaster
Swirl Concentrator/Regulator (9/10/80). 202
x 1 i 1
-------
82 Settleable solids cone, vs time Lancaster
Swirl Concentrator/Regulator (9/10/80). 203
83 Volatile settleable solids cone, vs time,
Lancaster Swirl Concentrator/Regulator (9/10/80). 204
84 Suspended sol Ids mass vs time, Lancaster
swirl Concentrator/Regulator (9/10/80). 205
85 Volatile suspended solids mass vs time
Lancaster Swirl Concentrator/Regulator (9/10/80). 206
86 Settleable solids mass vs time, Lancaster
Swirl Concentrator/Regulator (9/10/80). 207
87 Volatile settleable solids mass vs time,
Lancaster Swirl Concentrator/Regulator (9/10/80) 208
88 Suspended sol Ids cone, vs time, Lancaster
Swirl Concentrator/Regulator (10/2/80). 210
89 Settleable solids concentration vs time,
Lancaster Swirl Concentrator/Regulator (10/2/80) 211
90 COO concentration vs time, Lancaster Swirl
Concentrator/Regulator (10/2/80). 213
91 Suspended solids concentration vs time,
Lancaster Swirl Concentrator/Regulator (7/2/81). 21*
92 Volatile suspended solids concentration vs time,
Lancaster Swirl Concentrator/Regulator (7/2/81). 215
93 Settleable solids concentration vs time,
Lancaster Swirl Concentrator/Regulate:- (7/2/81). 216
94 Volatile settleable solids concentration vs time
Lancaster Swirl Concentrator/Regulator (7/2/81). 217
95 Suspended solids concentration vs tlir.e,
Lancaster Swirl Concentrator/Regulator (5/10/81). 218
96 Settleable solids concentration vs time,
Lancaster Swirl Concentrator/Regulator (5/10/81). 219
97 Suspended sol ids concentrations vs time,
Lancaster Swirl Concentrator/Regulator (7/20/81). 221
98 Volatile suspended solids concentrations vs time,
Lancaster Swirl Concentrator/Regulator (7/20/81). 222
99 Settleable Solids cone, vs time, Lancaster
Swirl Concentrator/Regulator (7/20/81). 223
100 Volatile settleable solids cone, vs time,
Lancaster Swirl Concentrator/Regulator (7/20/81). 224
101 Lancaster Swirl Concentrator/Regulator, Influent.
Comparison of settling column analysis with
discrete sampler results. 229
102 Comparison of Influent/clear suspended solids
sel-tllng curves with discrete sampler results,
Lancaster Swirl Concentrator/Regulator (7/2/81). 231
103 influent and clear suspended solids settling column
results, Lancaster Swirl Concentrator/Regulator
(7/2/81). 234
xi v
-------
Number Page
104 Influent and clear suspended solids settling column
results* Lancaster Swirl Concentrator/Regulator
(7/20/81 am). 235
105 Influent and clear suspended solids settling column
results, Lancaster Swirl Concentrator/Regulator
(7/20/81 an). 236
106 Influent and clear suspended solids sottlIng
column results, Lancaster Swirl Concenirator
/Regulator (7/28/81). 237
107 Influent and clear suspended solids cone, vs time,
Lancaster Swirl Oegrltter (6/22/81). 243
108 Influent and clear settling column results,
Lancaster Swirl Oegrltter (7/20/81). 245
109 Photographs of grit removed by Swirl Oegrltter,
Lancaster. 247
110 Strlpcharts, Hydro-Brake Evaluation. 250
111 Suspended solids concentrations vs time,
W. Roxbury Swirl Concentrator/Regulator (6/29/80). 260
112 Suspended solids concentrations vs time,
W. Roxbury Swirl Concentrator/Regulator (6/29/80). 261
113 Sett I cable solids concentrations vs time,
W. Roxbury Swirl Concentrator/Regulator (6/29/80). 262
114 Volatile suspended solids concentrations vs time,
W. Roxbury Swirl Concentrator/Regulator (6/29/80). 263
115 Suspended solids concentrations vs time,
W. Roxbury Helical Bend Regulator (6/29/80). 264
116 Settleable solids concentrations vs time,
W. Roxbury Helical Bend Regulator (6/29/80). 265
117 Volatile suspended solids concentrations vs time,
W. Roxbury Helical Bend Regulator (6/29/80). 266
118 Suspended solids concentrations vs time,
W. Roxbury Swirl Concentrator/Regulator (7/29/80). 268
119 Volatile suspended solids concentrations vs time,
W. Roxbury Swirl Concentrator/Regulator (7/29/80). 269
120 Settleable solids concentrations vs time,
W. Roxbury Swirl Concentrator/Regulator (7/29/80). 270
121 Volatile suspended solids concentrations vs time,
W. Roxbury Concentrator/Regulator (7/29/80). 271
122 Suspended solids concentrations vs time,
W. Roxbury Helical Bend Regulator (7/29/80). 272
123 Volatile suspended solids concentrations vs time,
W. Roxburv Helical Bend Regulator (7/29/60). 273
124 Settleable solids concentrations vs time,
W. Roxbury Helical Bend Regulator (7/29/80). 274
125 Volatile settleable solids concentrations vs time,
W. Roxbury Helical Bend Regulator (7/29/80). 275
126 Suspended solids concentrations vs time,
W. Roxbury Swirl Concentrator/ReguI a tor (10/3/80). 277
xv
-------
Number Page
127 Volatile suspended solids concentrations vs tine,
H. Roxbury Swirl Concentrator/Regulator (10/3/80). 278
128 Settleable solids concentration vs tine*
N. Roxbury Swirl Concentrator/Regulator (10/3/80). 279
129 Volatile settleable solids concentrations vs time,
W. Roxbury Swirl Concentrator/Regulator (10/3/80). 280
130 Suspended solid? concentrations vs time,
H. Roxbury Helical Bend Regulator (10/3/80). 281
131 Volatile suspended solids concentrations vs time,
W. Roxbury Helical Bend Regulator (10/3/80). 282
132 Settleable solids concentrations vs time,
W. Roxbury Helical Bend Regulator (10/3/80). 283
133 Volatile settleabie solids concentrations vs time,
W. Roxbury Helical Bend Regulator (10/3/80). 284
134 Suspended solids concentrations vs time,
Swirl Concentrator/Regulator (10/25/80). 285
135 Volatile suspended solids concentrations vs time,
W. Roxbury Swirl Concentrator/Regulator (10/25/80).286
136 Settleable solids concentrations vs time,
W. Roxbury Swirl Concentrator/Regulator (10/25/80).287
137 Volatile settleable solids concentrations vs time,
W. Roxbury Swirl Concentrator/Regulator (10/25/80).288
138 Suspended solids concentrations vs time,
W. Roxbury Helical Bend Regulator (10/25/80). 289
139 Volatile suspended solids concentrations vs time,
W. Roxbury Helical Bend Regulator (10/25/80). 290
MO Setrleable solids concentrations vs time,
W. Roxbury Helical Bend Regulator, (10/25/80). 291
141 Volatile settleable solids concentrations vs time,
N. Roxbury Helical Bend Regulator (10/25/80). 292
142 Suspended solids concentrations vs time,
West Roxbury Swirl (6/9/81). . 294
143 Settleable solids concentrations vs time.
West Roxbury Swirl (6/9/81). 295
144 Suspended solids concentration vs time.
West Roxbury Helical (6/9/81). 296
145 Settleable solids concentrations vs time.
West Roxbury Helical (6/9/81). 297
146 Photographs of the Helical Bend
Regulator operation, (6/9/81). 299
147 Suspended solids concentrations vs time.
West Roxbury Swirl (6/22/81). 300
148 Volatile suspended solids concentrations
vs time West Roxbury swirl (6/22/81). 301
149 Settleable solids concentrations vs time,
West Roxbury Swirl (6/22/81). 302
150 Volatile settleable solids concentrations
vs time, West Roxbury Swirl (6/22/81). 303
xv1
-------
Number Page
151 Suspended solids concentrations vs time.
Nest Roxbury Helical (6/22/81). 304
152 Volatile suspended solids concentrations
vs time. West Roxbury Helical (6/22/81). 305
153 Settleable solids concentrations vs tine.
West Roxbury Helical (6/22/81). 306
154 Volatile settleable solids concentrations vs time,
West Roxbury Helical (6/22/81). 307
155 Suspended solids concentrations vs time, West
Roxbury Svlrl (8/4/81). 306
156 Volatile suspended solids concentrations vs time.
West Roxbury Svlrl (6/4/81). 309
157 Settleable solids concentrations vs time.
West Roxbury S»IrI (8/4/81). 310
158 Volatile settleable so*Ids concenlrations
vs time, West Roxbury Swirl (8/4/81). 311
159 Suspended solids concentrations vs tlm«,
West Roxbur/ Helleal (8/4/81). 312
160 Volatile suspended solids concentrations
vs time. West Roxbury Helical (8/4/81). 313
16' Settleable solids concentrations vs time.
West Roxbury Helical (8/4/81). 314
152 Volatile settleable solids concentrations
vs time. West Roxbury Helical (8/4/81). 315
163 Influent suspended solids sett IeablI!ty
characteristics, W. Roxbury. 318
164 Influent and clear suspended solids
settleabllIty characteristics,
W. Roxbury Swirl Concentrator/Regulator (6/9/81). 322
165 Influent and clear suspended solids settleabl!Ity
character'sties, W. Roxbury Swirl Concentrator/
Regulator (6/22/81). 326
166 Influent and clear volatile suspended solids
settlaabl11ty characteristics, W. Roxbury Swirl
Concentrator/Regulator (6/22/81). 327
167 Influent and cl*ar suspended solids sett'eablI Ity
characteristics, W. Roxbury Helical Bend Regulator
(6/22/81). 328
168 Influent and clear volatile suspended solids
settleab11 Ity characteristics, W. Roxbury
Helical Bend Regulator (6/22/81). * 329
169 Influent and clear suspended solids settleab11 Ity
characteristics, W. Roxbury Swirl Concentrator/
Regulator (8/4/81). 334
170 Influent and clenr suspended solids se t-f i eab f II ty
characteristics, W. Roxbury Helical Bend Regulator
18/4/81). 335
-------
LIST OF TABLES
Number Paga
I Outline of Report 19
2 Comparison of Solids Removal Effectiveness of
Vortex Separator with £uIescentSett I Ing Tank..,54
3 Dynamic Solids Separator Performance Summary
(Tubercero) Monterry, Mexico... 56
4 Dynamic Solids Separator Performance Results
(Aqua I noustr I al )Monterrey, Nexcto 58
5 Cooperative Analysis of Rainfall statistics
Logan Airport vs Blue HIM (B.H.) Observatory
(1964-1973) .^..72
6 Mea sured Storm Events 78
7 SWMM Model Calibration Model Data 79
8 SWMM Calibration Results 81
9 Equalization Optimization 84
10 Actual Lancaster Swirl Dimensions 91
11 Actual Lancaster Degrltter Dimensions 93
12 Design Dimensions of Swirl Concentrator
N . Roxbury. 1 01
13 Lancaster Swirl Facility
Construction Cost Breakdown 114
14 Summary of Swirl Facilities
Construction Costs, Lancaster 115
15 9/14/81 Analytical Data
6/21/78 Analytical Data 226
16 Summary of Swirl Concentrator Performance 239
17 Swirl Degrltter Operation Summary..... 240
18 Sieve Analysis of Grit From Swirl Degrltter Portion
of Storm 241
19 Influent Suspended Solids Settleablity
Analysis Summary 319
20 Swirl efficiency Based on Settleabl11ty
Measurements, 6/9/81 324
21 Theoretical Swirl Efficiency, 6/9/81 325
22 Swirl Efficiency Based on SettleablI Ity
Measurements « 330
23 Theoretical Swirl Efficiency, 6/22/81 331
24 Helical Bend Efficiency Based on Settleab11 Ity
Measurement, 6/22/81 333
25 Summary: West Roxbury Evaluation Suspended Solids
Removal and Efficiency Rates 337
xvi11
-------
ABBREVIATIONS AND SYMBOLS
Abbreviations
cc
cm
cm/ sec
cfs
cu yds
ft
fl3
g
gal
g/cc
g/i
g/s o
gal/ftVday
gpm
gpm/ft*
ha
hr
1n./hr
1
1/s
1pm «
1/s/m*
lb/ftj
Ib/m1n
kg 2
kg/i/
kg/mln
km
km/ ha
I3
m3
m /sec
mg/1
mgal
mgd
ml/1
rpm
sg
cubic centimeter
centimeter
centimeter per second
cubic feet per second
cubic yards
feet
cubic feet
gran
gallon
grams per cubic centimeter
grams per liter
grams per second
gallons per square foot per day
gallons per minute
gallons per minute per square foot
hectare
hour
Inch per hour
liter
IHer per second
liter per minute
liter per second per square meter
pounds per square foot
pounds per cubic foot
pounds per minute
kilogram
kilogram per square meter
kilogram per minute
kilometer
kilometer per hectare
meter
cubic meter
cubic meter per minute
cubic meter per second
milligram per liter
million gallons
million gallons per day
nrmmter per liter
revolutions per minute
specific gravity
xi x
-------
Symbol s
b? Swirl Concentrator Offset, Primary Gutter
BOO Biochemical Oxygen Demand
BMP Best Management Practices
~C Clear Overflow Concentration
CF Clear Overflow concentration Factor
CM Clear Overflow Mass Rate
COD Chemical Oxygen Demand
CSO Combined Sewer Overflows
D1 Swirl Concentrator Inlet Dimension
°2 Swirl Concentrator Chamber Diameter
03 Swirl Concentrator Scum Ring Diameter
04 Swirl Concentrator Clearwater Ue1r Diameter
D-j* Swirl Degrltter Inlet Dimension
02* Swirl Concentrator Chamber Diameter
D3* Swirl Degrltter Clearwater Weir Diameter
D.L. Diversion Level
E Efficiency
ffc Foul Flow Concentration Factor
FC Foul Flow Concentration
FM Foul Flow Mass Rate
Hi Swirl Concentrator Height to Clearwater Ue1r
hi Swirl Concentrator Depth of Clearwater Weir
h2 Swirl Concentrator Depth of Floatable Trap
Hi * Swirl Degrltter Depth to Weir
H~* Swirl DegrHter Depth of Weir
H\* Swirl Degrltter Depth of Cone
I Influent Concentrator
IM Influent Mass Rate
Q Discharge
RI Swirl Concentrator Primary Gutter Radius
Rg Swirl Concen* ator Primary Gutter Radius
R3 Swirl Concentrator Secondary Gutter Radius
R4 Swirl Concentrator Secondary Gutter Radius
Re Swirl Concentrator Secondary Gutter Radius
RR Removal Rate
Si Swirl with Diameter 1
5z Swirl with Diameter 2
S.L. Swirl Level
% Percent
vs Versus
"XX
-------
ACKNOWLEDGEMENTS
The authors would like to comnend so many of the staff of
EDP who worked so hard to make this program a success. This
study, through the dill (gent efforts of many, collected the
largest data base of its kind available, through conditions of
severe weather, workload and stress.
The Boston Water & Sewer Commission was the recipient of
grants from the Municipal Environmental Research Laboratory, U.S.
Environmental Protection Agency, covering construction evaluation
and reporting for the Nest Roxbury Project. The Division of Water
Pollution Control, State of Massachusetts also provided direct
grant research monlos for evaluation. The Department of
Environmental Quality Engineering, State of Massachusetts was the
recipient of grant money from the Non Point Source Program, Water
Quality plannlng, U.S. Environmental Protection Agency covering
portions of the evaluation program. The Metropolitan District
Commission provided on-loan a control building to house
Instrumentation for the Lancaster Project, the City of Lancaster
shared the cost of the project with the U.S. Environmental
Protection Agency (25f/75f).
Dr. William C. Plsano, President, EDP was the Principal
Investigator and overall project manager. Mr. Daniel J. Connlck,
Chief Analytical Services, EDP, supervised EDP's evaluation
programs for both facilities and worked with W.C. Plsano In final
report preparation. Mr. Gerald L. Aronson, Executive Vice-
President EDP, supervised construction of the West Roxbury
facility and the design fabrication and Installation of the new
Cross-SecMonaI samplers In Lancaster. EDP Technologies, Inc. a
subsidiary equipment manufacturing corporation of EDP fabricated
the new samplers for Lancaster, Randel Dymond, Huth Engineering,
Inc. provided draft report materials and graphics for design and
construction sections of this report.
xxi
-------
Project Officers/Sponsors
Richard Field, Chief
Robert Turkeltaub, Staff Engineer
Richard P. Traver, Staff Engineer
Storm and Combined Sever Section
Mastewater Research Laboratory
Municipal Environmental
Research Laboratory
U.S.. Environmental
Protection Agency
Dennis Athyde, Chief
Phil Somers, Staff Engineer
Non Point Source Branch
Mater Quality Planning
U.S. Environmental
Protection Agency
Charles Button, Chief Engineer
Dennis Seblan, Staff Engineer
Boston Water & Sever Commission
Boston, Massachusetts
Thomas McHahon, Director
Water Resources Commission
Warren Klmball, Staff Engineer
Massachusetts Division of
Water Pollution Control
Westboro, Massachusetts
Anthony D. Cortese, Director
Made!Ine Snov
Nancy Apple
Massachusetts Department of
Environmental Quality Engineering
Boston, Massachusetts
John Angstadt, Director
Michael Freedman
Robin Thomas
City of Lancaster
Department of PublIc Works,
Lancaster, Pennsylvania
xxii
-------
CHAPTER 1
INTRODUCTION
1.1 Foreword
Background Information Is presented In Section 1.2
describing the urban runoff problem and the control of urban
runoff. Section 1.3 discusses the need for Improved runoff
control technology such as the Swirl and Helical Bend
regulator/concentrators. A general revlev of the Lancaster
and West Roxbury Facilities Is provided In sections 1.4 and
1.5 respectively, Including each project's goals and
objectives. The purpose of this report Is presented In section
1.6 and a reader's guide to the report Is provided In
section 1.7. Detailed descriptions of the operational units and
addltloral background material not given In this Chapter are
provided In Chapter 4.
2 BACKGROUND
THE URBAN RUNOFF PROBLEM
Three major types of discharges constitute the urban
runoff pollution problem Including: combined snwer overflows
from predominantly older cities, separate storm drainage from
newly developed areas, and diffuse nonpolnt overland flow
encompassing pollution loads from all types of land use
activities.
A combined sewer Is a sewer which accepts both domestic
sewage and stormwater flow. These sewers feed both stormwater and
sewage to the treatment plant up to Its design limit. When wet
weather flows exceed this limit, combined sewer overflow (CSO)
occurs, resulting In degradation to receiving water quality due
to washout of sewage and contaminated runoff.
Untreated storm overflows from combined (storm and
sanitary) sewers are a substantial water pollution source
during wetweather periods. There are roughly 15,000 to
18,000 combined sewer overflow points In the United States
that can equal 40 to 80 percent of the total organic load
from municipalities durlno wet-weather flow perl ads. It has
been estimated that, on a national level the expenditure for
combined sewer overflow pollution abatement would be $25
bl I I lon.d)
-------
Erosion of hillsides and construction sites caused by
rainfall can produce extremely high concentrations of
Inorganic solids, frequently several tines higher than
those In municipal sewage. Unauthorized or Intentional
cross-connections with sanitary or Industrial sewers are
common. Diffuse non-point source pollution from overland
runoff and drainage Is extremely difficult to effectively
manage and control. Best Management Practices (BMP) of a non-
structural nature can often be effective to help eliminate
the problems by controlling, the original sources of
pollutants.
Within the past 20 years* It has been recognized
that waters discharged from separate storm sewers contain
pollutants. Even without the addition of sanitary and
Industrial wastewaters, storm sewer discharges are usually
high In suspended solids (TSS) and on occasion, may have
Blochemclal Oxygen Demand (BOD) concentrations approaching
those In municipal sewage. Rain falling on an urban area
picks up pollutants from the air, dusty roofs* littered and
dirty streets and sidewalks. traffic byproducts (tire
residuals/vehicular exhaust), galvanic corrosion partlcul ates,
and chemicals applied for fertilization, control of Ice,
rodents, Insects and weeds.
The stormwater runoff problem Is a direct result of urban
growth and development. Roughly eighty percent of the
population of the United States Is now urbanized with the
process of urbanization proceeding at an estimated rate of
about 1,500 square miles (3900 sq km) per year. For years
discharges from stormwater drainage systems were considered
to be non-polluting and therefore were ducted to the nearest
available water course on the basts of convenience and cost.
Impacts from storm drainage systems are often more severe
than municipal discharges In a given area.
In storm sewers, there Is a tendency for solids to
settle out during the latter storm stages as the flow
tapers off and velocities are reduced. Also, large
separate sewer systems may have relief points that allow
some surcharged sanitary sewers, even on rare dry weath'--
occasfons, to overflow Into the storm sewers. The soil
In the municipal sewage overflows may accumulate In sto
sewers under these conditions and contribute to a "f I r L
-flush" effect In storm sewers. A similar situation exists when
storm sewers have Illicit direct Industrial and sanitary
connections.
-------
The accumulated solids deposits may contain
and other organic matter undergoing decay. When the sewer
overflow Increases sharply during a storm, this solid
natter which will exert a high organic loading will
be discharged. Depending on the sewer system, the rainfall
Intensity and the number of antecedent rfry days, a "first-
flush" effect may result. If It does occur. It may last
for a few minutes or even hours. Solids scoured from the
upper reaches of the large system may take a long time
to reach a dlstanr overflow point. BOO concentrations
during these periods may often exceed those of the normal
untreated dry weather wastewater.
CONTROL Oj; URBAN RUNOFF
The control of combined and storm sewer pollution
loadings Is one of the more difficult and expensive
problems the Nation faces In the abatement of
pollution for enhancing the quality of Its rivers, lakes
and streams. During the next decade or so. It Is
expected that billions of dollars may be spent In the
United States to combat the degradation of water bodies by
pollutants released through storm discharges and combined
sewer overflows. The difficulties and the need for
solutions to urban stormwater pollution have spawned
over the past decade a major research and development
effort both In the United States and In other nations around
the world. The U.S. EPA, Storm & Combined Sewer Research
group has engaged In multiple research and development programs
and Investigating to Identify, control and correct the
causes of known problems relating to these storm
occurrences.
U.S. EPA has developed over the last decade a number of
different treatment strategies for abating urban runoff
pollution. Physical treatment alternatives are the most
promising class of techniques that have resulted from this
research effort. Physical treatment alternatives are
primarily applied for floatable, suspended and settleable
solids removal and their associated pollutants from
wastestreams, and are of particular Importance to storm and
combined sewer overflow treatment. Removal of these
contaminants IP I n I m a I I y enhances visual recreational
potential of surface waters. Further treatment for colloidal
and dissolved pollutants contained In urban runoff may
never be feasible due to the Inordinate costs of
handling large volumes.
-------
Conventional nhysica! treataent alternatives for
stormwater and combined sever control typically Involve
expensive highly structural measures such as
detention/sedimentation basins (high capital costs/moderate
operational costs) or filtration/screening facilities
(moderate first costs/ high operational costs). U.S. EPA
sought through research new Inexpensive combination of methods
with low operational costs providing comparable solids
removal efficiencies.
There Is no "best" technological alternative for all CSO
problems. The most efficient control procedure Is highly
dependent upon the Individual problem. Some of the more feasible
alternatives are street sweeping, periodic flushing of sewers,
cleaning of catchbaslns, micro-screens for solids removal* Ir-
I I ne and/or off-line storage of storm flow and replacement of
the older combined sewers with newer separated sewers. However,
many of these alternatives are costly and require substantial
maintenance. Also, many of these alternatives control either
the quality or the quantity of CSO, not bo+h. The most recent
control philosophy has been referred to as the "two Q"
concept where the emphasis Is on both the quality and quantity
of the combined sewage overflow.
Since the regulator In the combined sewer system Is the
overflow point and Is a logical point of control, the practice
IP the United States of designing regulators exclusively for
flow-rate control or diversion of combined wastewaters to
the treatment plant and overflow to receiving waters Is
reconsidered as part of the "two Q" concept. Sewer system
management that emphasizes the dual function of combined sewer
overflow regulator facilities for Improving overflow
quality will pay significant dividends. The dual function
Is concentration of wastewater solids to the sanitary
Interceptor, and diversion of excess storm flow to the
outfalI.
1.3 The Need for Improved ReguIators
In considering wet weather water pollution
abatement, attention must first be directed to control of the
existing combined sewerage system and replacement (or
stricter maintenance) of faulty regulators. Consulting and
municipal engineers will agree that regulator mechanical
failures and blockages persist at the overflow or diversion
points resulting In unnecessary bypassing, which Is also a
problem during dry weather. Malfunctioning overflow rtructures,
both of the static and dynamic varieties, are major
contributors to the overall water pollution problem.
-------
Sever system management that emphasizes the dual
function of combined sewer overflow regulators for
Improving overflow quality by concentrating wastewarer
solids to the sanitary Interceptor and diverting excess
storm flow to receiving water body has been recognized
as a practical, efficient method to reduce receiving
water pollution. A new class of solids separation
devices have emerged from research seeking to Improve
combined sewer regulators. These devices are termed
secondary flow solids separators. These devices can either
operate as a regulator with capacity to provide treatment
during wet weather or alternatively, can be viewed as
strictly off-line treatment units. The Swirl and Helical
Bend regulators are two such devices. These new devices
can be used to remove floatable, settleable, suspended
solids and their associated pollutants from wastestraams
and are particularly Important to stormwater end CSO
treatment. The Swirl also has application for Industrial
wasre treatment requiring solids separation.
In the mid 70's U.S. EPA also perceived a need to
develop and have reserve of control hardware for urban
stormwater pollution from separate storm drainage, and
to effectively reduce the associated high cost Implications
for conventional storage tanks. It was felt that the
SwIrI/He!leal type regulators previously thought of for
application only to combined sewer overflow could be
Installed on separate storm drains before discharge. The
resultant concentrate can be stored In relatively small
tanks since concentrate flow Is only a few percent of
total flow. Stored concentrate can later be directed to the
sanitary sever for subsequent treatment during low flow or
dry weather periods or If capacity Is available In the
sanitary system, the concentrate may be diverted to It
rtthout storage. This method of stormwater control would be
cheaper In many Instances than building huge holding
reservoirs, and It offers a feasible approach to the
treatment of separately sewered urban stormwater.
SWIRL REGULATOR CONCENTRATOR
The Swirl Regulator/Concentrator Is of simple annular-
shaped construction and requires no moving parts. It
provides the dual 'unction, regulating flour by a central
circular weir spillway, while simultaneously treating
Influent wastewater by swirl action which Imparts
sol l**s/1 Iquld separation. (See Figure 1.)
-------
INLET
CHANNEL FOR
OVERFLOW
OUTLET TO
STREAM
\
TRANSITION
SECTION
'STRAIGHT (INFLUENT
\ SECTION
rr i
3L\
HELICAL
BENO 60°
OUTLET TO
SEWER
ISOMETRIC VIEW OF
HELICAL BEND REGULATOR
SCUM BAFFLE
WEIR
SECTION
/^CONCENTRATE
RETURN TO
SEWER
EMERGENCY
WEIR
;EFFLUENT TO STREAM OR
ADDITIONAL TREATMENT
-EMERGENCY
*mwsr- WEIR
CONCENTRATE
RETURN TO
SEWER
EFFLUENT
TO STREAM
PLAN
SWIRL CONCENTRATOR
Figure 1. Helical Send and Swirl Concentrator^.
-------
The combined sewage flow Is fed to the unit tangent (ally. A
slow circular notion Is Induced by the flow's kinetic energy and
separation of the larger, denser solids occurs at the bottom of
the swirl. This concentrated flow Is then transported either to
a pump station for conveyance, or directly to a sewage treatment
plant. Cleaner flow at the top of the circular tank overflows to
a discharge point, reducing the amount of flow requiring further
treatment.
Investigation of this mechanism fir CSO control has been
axtenslve. There have been a number of studies on the
applicability of secondary flow mechanisms for abatement of
combined sever overflow problems. These Include studies performed
by the American Public (forks Association, (APWA) La Salle
Hydraulic Laboratory, Ltd. and the University of Florida.
However, the Initial vortex concept was first studied In Bristol,
Englan^ In the early !960's.
Studies were carried out In Bristol for the purpose of
determining the hydraulic and performance characteristics of the
vortex mechanism. (2) It was found that this mechanism would
concentrate combined sewage solids from the liquid portion of the
flow and tnat the majority of the solids In the flow could be
collected In an "underflow foul sewer" for conveyance to a sewage
treatment plant. Further discussion of this research Is provided
In Chapter 4, section 4.7. At this point U.S. EPA decided to
Investigate this "vortex" mechanism for use In combined sewer
overflows In the United States. The resultant Investigation,
which was conducted by the APMA Initially consisted of modeling
the required hydraulic and mathematics conditions for the optimal
concentration of solids from the applied flow.
The La Salle Hydraulic Laboratory, Ltd. (Laboratorle
D'HydraulIque), LaSalle, Quebec, Canada, conducted hydraulic
modeling studies and the General Electric Company, Re-Entry and
Environmental Sytems Division, Philadelphia, conducted
mathematical modeling for the APWA study. Both modeling efforts
were combined In a report published by U.S. EPA In September 1972
(3).
The Swirl Concentrator model In the hydraulic study was 36
In. (91.4 cm) In diameter, 40 In. (101.6 cm) high and made of 1/2
In. (1.3 cm) plexiglass. The Inflow to the unit was seeded with a
known concentration of gllsonlte to simulate combined sewerage.
During the modeling studies. It was determined that agitation
caused by the vortex mechanism provided for mixing of the solids
which resulted In reduced efficiencies for solids concentration.
Therefore a flow deflector was constructed In the model to
prevent full vortex motion, resulting In a gentle "swirling"
notion which Improved the efficiency of solids concentration.
-------
Various alterations to the swirl model were tested In order to
optimize performance. The tested variables Included inlet
aperature size, foul outlet location. Inlet slope, gutter depth
and configuration, weir diameter and the floatables trap
configuration. After consideration of these variables In the
unit's performance* the study arrived at a recommended
configuration and design procedure.
At this point U.S. EPA desired to test the Swirl
Concentrator In a full-scale mode. The City of Lancaster,
Pennsylvania worked with U.S* EPA to Institute the first major
full-scale Swirl Regulator/Concentrator for CSO control. Other
much smaller Installations were also being built and tested
during this period end are discussed later.
A supplement to this report was published In July 1974
entitled "Relationship Between Diameter and Height for the Design
of a Swirl Concentrator as a Combined Sewer Overflow Regulator."
(4) This report augmented the original hydraulic modeling studies
at LaSalle Hydraulic Laboratory. Ltd. The purpose of this report
was to further develop the necessary design parameters of the
Swirl Regulator for maximum solids concentration efficiency.
Various geometric modification curves were developed to
facilitate design when head or existing system constraints
exist. This study has proven very helpful for the design of
Swirl Regulator/Concentrators and was utilized In the design of
the Swirl unit for the City of Lancaster.
During the Initial modeling of the swirl concept. It was
noted that solids of particular sizes or specffic gravities were
more effectively separated from the flow. Therefore, It was
logically deducted that the swirl concept could also be used as
a mechanism for grit removal as well as for the concentration of
CSO's. In June 1974 U.S. EPA published a report entitled "The
Swirl Concentrator as a Grit Separator Device."(5) This report
summarizes hydraulic studies undertaken to evaluate grit removal
with a swirl unit, which were also parformed at the LaSalle
Hydraulic Laboratory, Ltd. Consequently, these hydraulic studies
focused on the many variables In the swirl grit unit's design.
The University of Florida Department of Environmental
Engineering Sciences completed a report entitled "Storm Mater
Management Model: Refinements, Testing and Decision-Making" In
June 1973.(6) The goal of this study was to Improve and update
the EPA Storm Water Management Model (SWMM) by Including urban
erosion prediction, modeling of new treatment devices, monlrorlng
of pollution sources and modeling pollution loading from large
areas with flexibility. The computer model was programmed to
predict runoff hydrographs and po!Iutographs during a storm
event. The purpose was to predict quality and quantity of
8
-------
storrawater runoff In any brea given the local characteristics.
The model was tested in the Stevens Avenue Drainage District In
the City of Lancaster, Pennsylvania where field measurements
compared favorably with the predicted hydrographs and
pol I utographs. This modeling tool proved valuable In estimating
the characteristics of the combined setrege to be treated by the
Lancaster Swirl Regulator/Concentrator.
In addition to the full scale Swirl Degrltter and Swirl
Regulator/Concentrator for CSO control at the Lancaster
facility there have been several pilot-type Installations
throughout the United States. The City of Denver, Colorado
utilized a Swirl prototype for grit removal In the early 1970's
(7). Syracuse. New York Installed a 12 ft (3.6 n) diameter Swirl
Regulator/Concentrator and subsequent operation Indicated
successful separation of CSO solids. (8) Rochester, New York
Installed and successfully tested a pilot Swirl Concentrator In
series for CSO treatment. (9) Toronto, Ontario, Canada Installed
a pilot swirl primary separator which was tested for primary
sedimentation of sanitary sewage.(IO) The above pilot facilities
generally were successful and supported the feasibility of the
swirl concept In several various applications.
riaJIcal BjBjLd fion Rflgiilatgf - The Helical Bend flow
regulator Is based on the concept of using the secondary
helical motion Imparted to fluids at bends where a total
angle of approximately 60 degrees Is employed. (See Figure
1) Hydraulic model studies of this device, carried out at
the University of Surrey, England, Indicated that this Is
a feasible means of separating solids from the overflow.
(11)
The basic structural features of significance In the
Helical Bend -ares a) the Inlet from the entrance sewer
section to the device) b) the transition section from the
Inlet to the expanded cross section of the straight-run
section ahead of the bend) c) the overflow side weir and
scum board, or so-called drip plate) and d) the foul outlet
for the concentrated solids removal In the secondary flow
pattern together with the means for controlling the amount
of this underflow going to the treatment works.
The heavily polluted sewage Is drawn to the Inner wall.
It then passes to a semicircular channel situated at a
lower level leading the treatment plant. The proportion of
the concentrated discharge will depend on the particular
design. The overflow passes over a side weir for
discharge to the receiving waters. Surface debris collects at
the end of the chamber and passes over a short flume to
join the sewer conveying the flow to the treatment plant.
-------
The model study Indicated that even the simplest
application of the spiral flow separator will produce an
Inexpensive regulator that will be superior to many
existing types. (11) It was also stated that further research
Is necessary to define the variables, the limits of
applications and the actual limitations of the spiral
flow regulator. A prototype Helical Bend regulator has been
designed and constructed at Nantwhlch, England.
A Helical Bend hydraulic model was constructed at the
La Salle Hydraulic Laboratory. Configuration changes and
Internal auxiliary flow control and effluent Improvement
modifications, as well as detailed studies of the helical
flow patterns developed and suspended solids removal
efficiencies achieved, were studied. (12) As part of this
effort. Mathematical model and computer simulation of the
device was conducted by the General Electric Company.(12)
The hydraulic model studies and the computer mathematical
simulation of the Helical Bend combined sewer overflow
regulator Indicate that this "flash-method* of solids
removal, without use of mechanical appurtenances, can produce
excellent efflclences with reasonable size units In
combined sewer systems.
The ultimate purpose of these studies was to correlate and
confirm findings by both hydraulic and mathematical means
and to develop design criteria which will enable
engineers to utilize the helical flow principle for solids
removal from combined sewer flows and to properly
regulate The overflow of clarified wastewater to
receiving waters or points of retention and/or
treatment.
A design manual for secondary flow pollution control devices
Including Swirl Concentrator and the Helical Bend Regulator was
recently prepared by APWA In cooperation with U.S. EPA (13). This
report Is meant to coalesce piece-meal reporting from the various
research and demonstration projects Into one unified document.
1.4 HISTORICAL BACKGROUND AND PURPOSE OF LANCASTER FACILITY
The City of Lancaster, Pa. experimented with a full scale
swirl Installation due to severe CSO problems and U.S. EPA's
commitment to partially fund a demonstration project for the
Swirl/Regulator Concentrator.
Following U.S. EPA's decision to use a relatively Isolated
drainage district within the City, a cooperative effort between
both the City and the U.S. EPA resulted In the Initiation of
10
-------
project design. Preliminary design of the Lancaster Swirl
Project was completed by Meridian Engineering, Inc.,
Philadelphia, PA In 1972. (14) At that time the project was known
PS the Silo Project. The prototype swirl foul underflow
containing the concentrated solids would be pumped to an
Interceptor for conveyance to the City's South Mastewater
Treatment Plant (MNTP), with the clean overflow frotfc the swirl
stored In a large, aerated silo-shaped storage tank for
release to the Conestoga River following chI or I nation. As an
alternate, the overflow could be pumpod back to the
Interceptor sewer during a dry period. This concaot was not
used due to the prohibitive cost of the silo construction.
In February '974, Huth Engineers, Inc. of Lancaster, PA
was retained by the City of Lancaster to Initiate the design
of the Swirl Regulator/Concentrator. A facilities layout and a
schematic flow diagram of the project are shown In Figures 2
and 3, respectively. The new concept called for a prototype
Swirl to receive the combined sewage with the foul outlet
flow being controlled and directed to a second, smaller Swirl
device for grit removal while the overflow from the larger
Swirl Regulator/Concentrator unit wo'ild be discharged to the
Conestoga River. The Inclusion of the Swirl Oegrltter unit was
for the purpose of preventing excessive wear due to grit
particles In the raw sewage pumps which are downstream of the
degrltter clear effluent. The separated grit would be collected
and weighed after each storm event to aid In determining the
degrltter's removal efficiency. A Dlscostralner was also Included
to evaluate It for removing solids from the flow stream. The
overflow frou the Swirl Degrltter would then be pumped to the
Interceptor sewer via the existing Stevens Avenue Pumping
Station.
The 24 ft (7.3m) diameter duaI -functI on Ing Swirl
Regulator/concentrator treats CSO from the Stevens Ave. District
(one of the city's six overflow points to the Conestoga River).
The 215 acre (88.5 ha) drainage area served Is approximately 50$
residential and 50$ parklands and undeveloped fields.
The Swirl was designed to operate at 40 cfs (1.1
ffl3/sec)wlth capacity to be surcharged to 90 cfs (2.6 mVsec).
During dry weather sewage enters the Swirl and Is directed
along the floor gutters to the foul underflow outlet and then
directly to the pumping station. A Swirl Degrltter and grit
handling system was constructed to remove grit from the
foul underflow of the Swirl Concentrator before the
underflow Is pumped to the municipal sewage treatment plant.
11
-------
-
-
•
v. .
-••••• • " >:' • '^!feV^'Jli%
•^>".
^.---^
"•>**' . i*^;
_
«
'
•
CONTROL BUILDING
COfUCENTRATOR
I!
.
?
'S§f.-- ".""'''^S^',^^^^^
'^^^Sl-
K2l'**1^
HB»^ MM. CONCfNTNMW
• •
_^ _ iv
MMMHMHI
•K--.*-- - , • **&*!&;»*.
-
MPUUCNT
.
|
.
TO CONESTOC*
Figure 2. Loncaster Swirl Project fcdlHiet
-------
Ifr'
. '*iK
-'
,
STEVENS
PUMP 8T4TION
•«.•
-m?
:-«:
g --
. .
.
flow diogrom of Loncoittr Swirl
.
'
&
'.
-------
The Huth project design concept called for a bar screen to
protect downstream facility components from large debris. The
Swirl facilities would be automatically ac-rivared when storm
flow entered the swirl. A sonic level sensor would activate
a sequence of valves to change the mode from dry to wet weather
operation. All clear overflow would be chlorinated before
discharge to the river. A chlorinated pipe extending over the
clear overflow weir assembly can dose the effluent Just
before It Is discharged. Flushing lines located on the side
walls above the water line and under the' floatables trap would
wash deposits from the Swirl after each storm. The deposits
would be washed to floor gutters and sent to treatment plant.
An emergency overflow velr occupies one quadrant of the Swirl
wall. Th» weir la 6 In. (15.2 cm) higher than the clear
overflow weir and would provide an additional 19 ft (6 m) of
weir length for flows that exceed the hydraulic capacity of
the Swirl.
The final design of the project was completed by Huth
Engineers, Inc. and approved by the U.S. EPA In November, 1975.
Construction began In April 1977 and was completed In June
1978 with operation of the facility commencing shortly
thereafter. Initial evaluation efforts In 1978 and 1979 were
plauged by Instrumentation problems.
Environmental Design & Planning, Inc., assjmed
responsibility for revamping critical elements of the
facilities Instrumentation, assisting the City In a further
evaluation effort beginning In September, 1980 and lasting till
July, 1981 and preparing the final report.
PROJECT GOAL
The outlined project goals for the Lancaster Swirl Project
are shown below. They aret
1. To successfully demonstrate the applicability of the
full-scale swirl conceot for use In the qualitative
and quantitative control of CSO.
2. To determine the actual solids removal efficiency of
an operating large-scale Swirl Regulator/Concentrator.
3. To demonstrate tho efficiency and applicability of a
Swirl Degrltter In a CSO control Installation.
4. To develop a background of cost and design experience
for use In future Installations. This Includes noting
any particular scale-up or design problems
encountered.
14
-------
1.5 HISTORICAL BACKGROUND AND PURPOSE fl£ WEST ROX3URY FACILITY
During the mid 1970's the Sw'r! Regulator/Concentrator was
being examined by U.S. EPA and others as a suitable compact
device for removing solids and partlculate material from
combined sever overflows and primary treatment for grit
separation at treatment works. The performance of the device was
believed to be good and extremely economical. It was considered
(and still Is) as a viable control option In a number of combined
sewer management studies throughout the country. The swirl had
yet to be tried for removing pollutants from a separate storm
sewered area.
The generation of operational experience using the Helical
Bend regulator vas still In Its Infancy. The full-scale field
demonstration of the Helical Bend regulatr-- had yet to be
conducted. It was felt In the mid 70*s that non-point control
would soon move Into the Implementation phases for a number of
areas throughout the country. Thus, It was deemed Important to
the federal government at that time to demonstrate the field
feasibility of using Swirl and Helical Bend regulators to
handle storm sewer emissions. The aim of the West Roxbury
project was to determine the effectiveness of a fuH-scale
Installation using both devices for removing storm sewer
pllutant emission loadings.
A plot plan of the West Roxbury, Boston research and
demonstration facility Is shown In Figure 4. Discharge from
160 acres (65 ha) of separated storm sewer system enters the
site through an 87 In. (2.2m) reinforced concrete pipe (RCP)
which connects Into a 120 In. (3.05 m) conduit discharging to
the Charles River 1000 ft (305 m) downstream. Storm drainage Is
diverted by gravity Into the site and to the two units. Flow
Is split evenly to the two units by motor-driven bottom
opening sluice gates. In the second mode of operation, sewage
from a 27 In. (.68 m) sanitary trunk sewer (with severe Inflow
problems) Is pumped over the 120 In. (3.05 m) drain and mixed
with Incoming storm drainage to simulate combined sewer overflow
conditions.
Clear water discharge from the 10.5 ft (3.2 m) diameter
Swirl drains by gravity Into the 120 In. (3.m) reinforced
concrete pipe. The foul sewer drains from the unit and Is
discharged by gravity Into the foul sump tank. Discharge In
the foul sewer underflow Is limited to 3£ of the design
Influent. The design flow Is 6 c*s (170 l/s) with a maximum
flow of 12 cfs (340 l/s) and a 0.18 cfs (5 l/s) underflow. The
Xest Roxbury Facility Is described In more detail In Chapters
6,8,10 and 13 of this report.
15
-------
87 In. STORM MAIN
120 In. STORM DRAIN
27
ME.VFiM amitf.; I'Aia
Figure 4 Swirl /Helical Bend site plan,W. Roxbury Project.
at-
«•
-------
EBflJf£1
divided Into three
with detalled site
efforts and detailed
runoff flows for
second phase of
££.!££ !!.¥£: The envisioned project was
separate phases. The first phase dealt
Inspections, low level flow monitoring
computer modeling to ascertain typical
the design and construction phase. In the
work. Swirl and Helical Bend
Regulator/Concentrators were designed* fabricated and Installed
near the outlet of a separate storm sewered urbanized area.
The flow at the outlet was divided and Inputted Into the two
units constructed side by side. In the third phase of work,
the two treatment units were monitored during a 12 month
period for approximately 13 separate storm events covering
a vide range of antecedent dry conditions, rainfall
Intensity and duration. A comparative assessment of the two
units were made from these results.
The objectives of the project are as follows:
1)
2)
3)
4)
To design, construct, and Install a Swirl and Helical
Bend regulator operating on storm sewer discharged from
a residential area for a 12 month period (15 storm
events).
To test the operational pollutant reduction performance of
a Swirl and Helical Bend regulator on a storm drainage
discharge from a residential area.
To assess the hydraulic Influent and effluent/foul sewer
waste stream char actor IstIcs for various meteorological
e* ents of the two treatment devices.
To test the feasibility of foul sewer d
an adjacent sanitary sewer for the remove:
s+ream effluents over an extended period.
^charge to
of waste
1.6 Purpose of Report
The Intended purpose of this report Is to provide the
user-community with a synopsis of the Information available
on the control of urban runoff pollution using secondary
flow separation technologies. This report references various
publications available on the abatement of stormwater runoff
pollution In urbanized areas. A documentation of the results
of the evaluation of stormwater regulator/concentrators Is
placed on the methods of analysis and
of the two primary full scale
provided. Emphasis Is
performance results
facilities at Lancaster, Pa. and West Roxhury, Ma.
17
-------
1.7 Report Coord 1 nates
This report contains fifteen chapters. Summary conclusions
and recommendations are presented In chapters 2 and 3. The
remaining chapters can be classified under the heading of
Process Description, Facility Description and Facility
Evaluation. Table 1 lists the major groupings und their
relevant chapters. Chapter 4 Includes control process
descriptions andhlstorles for the swirl concentrator, swirl
degrltter, swirl primary separator, dynamic separator and
helical bend. Chapters 5 through 10 provide study area
descriptions, design and construction of the Lancaster and
Nest Roxbury facilities respectively. Evaluation methods and
results for each facility are discussed In Chapters 11 through
13, respectively.
18
-------
TABLE 1
OUTLINE OF REPORT
A. Process Description
Chap-tar
4
B. Facility Description
Chanter
5
6
7
8
9
10
C. FaelI Itv Evaluat Ion
Chapter
11
12
13
Tltla
Process Descriptions/History
Title.
Study Area Description -
Study Area Description -
Design - Lancaster
Design - W. Roxbury
Construction - Lancaster
Construction - If. Roxbury
Lancaster
W. Roxbury
Evaluation
Evaluation
Evaluation
Tltla
- Methods
- Results
- Results
Lancaster
West Roxbury
19
-------
CHAPTER 2
CONCLUSIONS
Conclusions derived from this Investigation are divided Into
two road categories Including: technical solids remove!
performance conclusions. Instrumentation and sampling technique
concluslons.
A. Taehn I e-al Performance Cone I us Ions
Lancaster
1. The Swirl Concentrator/Regulator proved to be an
efficient device for removing gross "first-flush" related solids
from combined sewage. The design of the Swirl Concentrator at
Lancaster was determined through laboratory and prototype
experimentation to remove (treatment and not by flow-splitting)
90? settleable solids (artificial media meant to replicate
combined sewage solids) at a design flow of 40 cfs (1126 l/s).
Actual suspended solids treatment efficiencies monitored In fall,
1980 through summer, 1981 exceeded 60$ for Influent discharges
exceeding 20 cfs (563 l/s). Typically "first-flush* occured at
the onset of storm events when the flows rapidly peaked.
Treatment rates then drastically reduced for later portions of
events. Most of the suspended solids were Inorganic settleable
grit In nature. Persistent Influent flow meter malfunctioning
precluded accurate estimates of flow-weighted efficiency rates.
2. Settling column tests performed on the Influent during
storm events Indicated suspended solids particle distributions
having significantly less larger particles (settling velocities
greater than I cm/sec or 0.0328 ft/sec) than artificial media
used to develop the design. These apparent differences may have
been attributable to settling column sample collection
techniques. (See conclusion 13).
3. Swirl Concentrator efficiency conclusions reached above
were determined using data obtained by automatic discrete
sampling techniques. Swirl Concentrator efficiencies determined
using differences In measured Influent and clear overflow solids
particles vs settling velocity distributions were much lower.
These apparent differences may have also been attributable to
settling column sample collection techniques, (see also
concluslon 13).
4. Swirl Concentrator COD treatment performance was noted
for several storms to be comparable to suspended solids removals.
Swirl Concentrator efficiency rates for volatile suspended
20
-------
solids, settleable solids and volatile settleable solids were
comparable to efficiency rates noted for suspended solids.
5. Swirl Concentrator efficiency rates seemed to
deteriorate with flows below design flow. Much lower efficiencies
(5f-20$) were typically noled near the end of the sampling events
when both the Influent flow and solids concentrations were the
lowest. Influent flow melers typically malfunction at the start
of the event when flows were low and "first-flush" concentrations
high.
6. Evaluation results for the Swirl Oegrltter and Disco-
Strainer are Inconclusive. Both units rarely functioned correctly
due to facility design problems.
7. The Hydro-Brake used to regulate the flow from the
Swirl Concentrator foul sewer to the Swirl Degrltter during wet
weather operation appeared to function correctly and did not clog
(soda tin cans which has passed through the Hydro-Brake were
frequently noted floating In the Swirl Degrltter).
Wast Roxbury
8. Th* Swirl Concentrator and Helical Bend Regulator
provided treatment of stormwater related suspended solids.
Treatment efficiencies ranged from 5% to 40$ and were not flow
related. The Swirl Corcentrator had been sized to provide 80$
removal of settleable solids (artificial media) at a design flow
of 6 cfs (170 l/s). The Helical Bend Regulator had been sized to
provide 90% removal.
9. Settling column tests performed on the Influent to the
facility Indicated stormwafer suspended solids particle
distributions having significantly less larger particles
(settling velocities less than 1 cm/sec or 0.0328 ft/sec) than
artificial media used to develop the design for treatment
combined sewage solids. As with Lancaster (but to a lessor
extent), these apparent differences may have been attributable to
settling column sample cc'lectlon techniques (same manual
sampling procedure used In West Roxbury as In Lancaster see
conclusions 13).
10. Swirl Concentrator efficiencies noted In conclusion 8
were determined using data obtained by discrete sampling
techniques. Comparable suspended solids efficiency levels were
determined using differences In measured Swirl Influent and clear
sol Ms particles vs settling velocity distributions. The actual
21
-------
efficiency noted from settleablIIty analyses for several events
compared favorably vlth estimates computed using measured data
with theoretical performance formalisms developed In the APVA
vork (4).
B» I nstrumentatton/Samp I Ing TaehnIques
Lancaster
11. Special Cross-Sectional samplers capable of taking
complete vertical "slices* of fluid over an entire pipe cross
section In an automatic discrete mode vere developed* tested and
Installed In Lancaster In Spring, 1981 by EDP Technologies. These
devices vere placed on the 36 In. (.91 m) Influent conduit to the
Swirl Concentrator and on the 12 In. (30.5 cm) line to the Swirl
Degrltter. These devices were Installed during the evaluation
program to obtain representative solids samples of discharges to
the Swirl Concentrator and the Swirl Degrltter. During four storm
events suspended solids concentrations of the Swirl Concentrator
Influent were determined on samples obtained by the new Cross-
Sectional sampler and on samples simultaneously collected by a
Manning model 6000 sequential sampler. An analysis of the data
Indicated that suspended solids of samples collected by the new
sampler were 6 to 7 times more concentrated than samples
collected (at the same Instant) by Manning sampler during the
first 10 minutes of the storm events peak of "first-flush"
passage. Concentration factors reduced to 2 to 4 times for
samples collected mid-event and then down to 1.5 to 2.0 for end -
of-event samples.
12. The set of Swirl Concentrator Influent suspended solids
multiplicative factors cited In conclusion 11 were ised to
reassess seemingly poor efficiency performance of an event
monitored In fall, 1980 (before special Cross-Sectional samplers
were Installed) and an event monitored In 1978 during the Initial
evaluation period. Both events showed good solids removal
efficiency using the Influent multiplicative factors. It had been
concluded (prior to EDP's Involvement) that the Swirl
Concentrator had been providing negligible to negative solids
treatment efficiency.
13. Large volume Swirl Concentrator Influent samples taken
manually for settleablllty analyses typically contained suspended
solids concentrations much lower than concentrations of samples
taken using the Cross-Sectional and Manning) samplers during
comparable time periods. It appeared that much of the heavier
grit-like material was missed by the hand sampling manual
operation which precluded adequate representation of the bed load
material.
22
-------
CHAPTER 3
RECOMMENDAT IONS
Recommendations derived from this Investigation are
presented In two major categories. Including unit operation,
design modifications and Instrumentation samp I Ing methodology
recommendations.
A. Das Ijn
Lancaster
1. The diversion chamber at Lancaster should be revamped to
minimize dry weather debris accumulations. The bar screens should
be eliminated or placed after the Swirl Concentrator. The
diversion floor chamber should be cambered to prevent shoaling.
The relative directions of the Influent/overflow lines to the
chamber should be reversed. The Influent flow comes In one side
of the chamber and exists to the Swirl Concentrator after
traversing tho width of the chamber and completing two 90 degrees
flow changes. The bypass conduit controlled by the sluice gate
operation Is on a direct line with the Influent to the chamber.
2. A bypass to the Swirl Concentrator should be constructed for
diverting dry weather and normal weather and normal wet weather
flow around the unit.
3. The concrete walls In the Swirl Concentrator chamber i-hcvld
be trowel-finished or epoxy-coated as to minimize n*>te-lul
clIngi ng.
4. The bottom floor and primary and secondary gutters of the
Swirl Concentrator should have greater pitch than the present
design. There are portions or the secondary gutter where the
slope Is minimal (and even negative). Debris and large sediment
accumulations were visually noted at the upper end of the
secondary gutter.
5. The Swirl Concentrator Inner assembly should be constructed
out of stainless steel In floating lap-seam sections as to
minimize thermal warpage/overfIow welr-IevelIng problems.
Stainless steel would also e'lmlnate excessive rusting and
corrosion. Portions of the Inner assembly which was constructed
out of "weathering" steel have badly corroded.
6. Permanent washdown piping attached to Swirl Concentrator
walls and underside of the weir plate assembly should be
eliminated since materials catch on these protrusions. A good
23
-------
hose washdown system with sufficient length to permit washdown of
underside of weir plate would be adequate, (accumulations vould
be minimized If recommendations 4 and 5 were Instituted).
7. The Swirl Concentrator foul underflow should have'its flow-
limiting device, that Is, the Hydro-Brake placed within the
Concentrator floor slab and not downstream after a stretch of
pipe. Both In Lancaster and In West Roxbury, Hydro-Brakes are
located on foul sewer lines away from treatment units. Sediment
accumulations constantly occurred within these segments and were
a problem In West Roxbury.
8. The activation level within the Swirl Concentrator tanx that
Is presently used to switch the dry weather flow pattern to wet
weather operating conditions should be lowered below 18 In. (45.8
cm). Flows as high as 8 cfs (226.4 l/s), that Is, part of the
•first-flush" can be d'scharged without Swirl treatment under
present setup. Lowering of this level can only be accomplished
after the Swirl Oegrltter operating characteristics have been
enhanced.
9. The Swirl Oegrltter requires major modification to enhance
Its operation. Severe solids "bridging" constantly occurred
during the evaluation program which clogged the unit and
precluded an adequate evaluation of the Swirl Concentrator/Swirl
Degrltter configuration. The major deficiency Is the tightness
of the apex cone on the bottom of the Degrltter. Either the
bottom portion of the Oegrltter cone (solids drawoff point)
should be cutoff and/or the slope of cone should be lessened to
alleviate the severe "brIdglng"/clogglng problem. Epoxy coating
the Interior of the Swirl Degrltter or fabricating the unit out
of stainless steel would have lessened the "bridging" problem.
The bottom cone of the Swirl Degrltter should also be out-fitted
with a pumped counter-current flush/withdrawal system to permit a
reasonable and sanitary unclogglng procedures.The present design
I Imlts unclogglng to manual "crow-bar" operation with a single
tap for f I ushwater/draln. Unclogglng the Degrltter usually
results In spillage of Degrltter solids on the basement floor of
the control building.
10. The screw-conveyor/link belt system for lifting grit from
the Swirl Degrltter up from the basement floor to a dumpster
outside the control building needs splash-plates to avoid
spillage during operation. The first segment of this system from
the Swirl Degrltter Is a screw conveyor which should be replaced
with a link-belt conveyor. A screw conveyor will tend to compress
compactable material In a reverse direction from the forward
flow. This problem may have aggravated the solids "bridging"
problem In the Swirl Degrltter chamber.
24
-------
II. The telescoping washdown boos located on the Swirl Degrltter
screw conveyor was meant to wash organ Ics from the grit. Organ Ics
would flow back Into the Degrltter and overflow to the MWTP. This
system has never functioned since Its operation was dependent
upon solids not "bridging" within the Degrltter. A simpler method
for achieving the same grit/organic separation would Involve
permitting variable flow Into the Oegrltter and adjusting the
overflow throughput hydraulic rate such that only the organlcs
and fIoatables would pass over the clearwater weir (supernate to
WNTP) and grit would settle and be removed. The original design
permitted this possibility since a 12 In. (30.5 cm) pinch-valve
was. orlgl nal ly used to regulate flow. It vas replaced with the
fixed discharge Hydro-Brake because the pinch-valve frequently
malfunctioned and clogged. A much smaller sized pinch-valve would
have permitted both a larger throat opening (would not have
clogged) and variable flow adjustment.
Wast Roxbury
12. The He!leal Bend Regulator can be economically constructed
by a formed concrete guntte operation on a poured slab
foundation.
13. Small diameter Swirl concentrators, that Is, less than 12 ft
(3.63*m) In diameter should be fabricated of carbon steel with
Interior epoxy-coated or fabricated using stainless steel. The
Swirl Concentrator at West Roxbury was constructed of steel and
the Interior was epoxy-coated. No washdown system was Included
and there was no problem encountered due to clinging material.
B. InstrumantatI on/Samp I Ing Mathodoloyy
14. The Cross-Sectional sampler Is recommended for use In any
application where careful sampling Is desired of a randomly
varying heterogenous waste stream of unknown bed load, suspended
load and wash load fractions. The rectangul-ar Intake silt In the
present slider configuration allows a uniform sampling (and
Integration) over the entire flow depth. An Inverted trapezoidal
shape would probably allow a more accurate mass representation of
different fractions probably present.
15. Large volume samples (30 gal - 113.4 I) samples taken for
settling column sett IeabI I Ity analysis should be taken by a
modified version of the Cross-Sectional samplers used In
Lancaster. These devices delivered 1-1 samples and could be
operated as f>equently as one cycle (1-1) per minute. Collection
of the required 30 gal over a 10-20 minute period would be
Impossible using the prssent configuration. The slit opening
could be varied and a different type of carousel used to handling
larger volumes.
25
-------
16. Because of the flow monitoring problems In Lancaster, It was
Impossible to ascertain hot* the Swirl Concentrator solids removal'
efficiency varied with flow rate, particularly during "first-
flush" period when significant removals were noted. It would
therefore be deslreable to assess full-scale Swirl Concentrator
performance under sustained and even flow rates over the course
of a storm event. The envisioned pump-fed Swirl Concentrator
demonstration project upcoming In Saglnaw, Michigan will permit
this opportunity.
26
-------
CHAPTER 4
Descriptions of Pertinent Treatment Units
4.1 Foreword
Integral parts of the Swirl Regulator/Concentrator
and design considerations are presented In Section 4.2.
Description of several full-scale swirl installations are also
provided In this section. A variation of the Swirl
Regulator/Concentrator - the Swirl Degrltter Is discussed In
Section 4.3 along with examples of Its practical application.
The Integral parts of the Helical Bend Regulator are described
In Section 4.4. The Dlscostralner, an alternative solids
separating device vhlch operates by straining, sedimentation
and filtration Is discussed In Section 4.5. Section 4.6
describes yet another regulator/concentrator - the Teacup
Solids Separator. Finally, the most recent Innovative solids
separating device - the Dynamic Solids Separator Is presented
In Section 4.7 along with capsulated reviews of four different
evaluation efforts. The Dynamic Solids Separator has It's
origins with continuing British research to Improve the Vortex
"operators tested In Bristol, England during the early 60's (2).
.2 I ntagral Parts of Swirl Regii I ator/Coneantrator Daslgn
As discussed In Chapter 1 the Swirl Regulator/Concentrator
Is of simple annular-shaped construction and requires no
moving parts. An Isometric view of the final form of the
device was shown In Figure 5. The Swirl provides a dual
function, regulating flow by a central circular weir spillway
while simultaneously treating combined wastewater by swirl
action which Imparts so I Ids/I Iqu I d separation. Dry weather
flows are diverted through a cunette-llke channel In the
floor of the chamber Into the bottom orifice or foul
underflow located near the clear water down sheft to the
Intercepting sewer for subsequent treatment at the municipal
plant. During higher flow storm conditions, the low-volume
concentrate (3-1 Of total flow) Is diverted via the same bottom
orifice leading to the receiving sewer Interceptor and the
excess, relatively clear, high volume supernatant overflows
the central circular walr Into a downshaft for storage,
treatment or discharge to a waterway.
For an essentially static device to perform efficiently
under varying flowrates and suspended solids concentrations,
special attention mv^t be given to the various elements
within t!ie chamber. (Refer to Figure 5).
27
-------
'*i:,
•
•'
•
11
'
• "*" '''SB
• • • .:.' -
• f.
i
•
INLET STRUCTURE
W «J VV^P
3 OVERFLOW WEI* AMD HORIZONTAL WEIR PLATE
t SPOILERS
^^^^1^^^^
G FOUL SEWER OUTLET
H FLOOR GUTTERS
FigurtSi Isometric view of S*«rlRequ!ator/Conc«ntfotor
28
-------
Xh£ J.XLL&1 RaJDJi &&d Tf apsl^lofl 5fi£±J.aJl should Introduce
the Incoming flow at the bottom of the chaofter and sliow
solids to enter at the lowest position possible. Th.> Inflow
should not be turbulent to prevent solids from being
carried directly to the overflow weir along with the water
and the floor of the ramp should be V-shape to allow
for self cleaning during periods of low flow.
Tha Flow Da-f I actor Is a vertical free-standing wall that
Is an extension of the Internal wall of the Inlet ramp.
It directs the flow around the chamber, setting up the swirl
or spiral hydraulic patterns, thereby creating a longer
particle path and a greater chance for solids separation.
The purpose of the Sfium JLLJLfl Is to prevent floating
solids fro* overflowing. It should be extended a minimum of
6 In. l\5 cm) below the level of the overflow weir crest. The
Overf low Weir and Weir PI ata connects to a Central Downshaf t
carrying the overflow liquid to discharge. It's underside
acts as a storage cap for floating solids that are directed
beneath the weir plate through the EJoa.tflfaJflS IL3UL* The
vertical element of the weir Is extended below the weir plata
to retain and store floatables. When the liquid level In the
amber decreases after the rainfall, the floatables exit
ough the foul sewer underflow.
The Spol lars are radial flow-guides that extend from the
downshaft to the scum ring and are vertically mounted on the
weir plate which break up the turbulent vortex conditions
that If allowed to exist, would Impede proper function.
The Foul ifljtfi£ Qutl at or Underf low Is strategically
located on the floor of the chamber which allows dry weather
flow and concentrated storm flow to exit. It Is placed at the
point of maximum settleable solids concentration and Is
designed to reduce the clogging problems •hat often
Incapacitate conventional regulators.
JLfi£.oj).4.a£.y. £J.£fi£ &u.±±fi.r..s are designed for
peak dry weather flow and are semi-circular In shape to
prevent shoaling and solids deposition.
An Emergency S tda Ovarf low Watr Assamb ly allows *he Swirl
to be overdriven to bette*" than twice Its design fiowrate
and still maintain ef • * I v e setrleable solids removal
efficiencies. This is accoi shed by short- circuiting excess
flow and maintaining the Integrity of the chamber's
separation flow patterns.
29
-------
Flush Bin a As££jaJl.LL.as located around r'.e upper
portion of the inferior wall of the Swirl chamber and beneath
the floatables trap allow for easy clean-up operations
folloving a vet weather treatment event.
Evaluations of the Swirl Regulator/Concentrators have been
performed In various locations throughout the U.S. Full-scale
facilities have been evaluated In Lancaster, PA, K. Roxbury, MA.,
Syracuse, NY and San Francisco, CA. Swirl Concentrator
demonstration project In Saglnav, Michigan Is currently (1981) In
design phase. A detailed discussion of the Lancaster Facility
and Its evaluation Is presented In Chapters 5,7,9 and 12 of
this, report. The Vest Roxbury Facility Is described In detail
In Chapters 6,8,10 and 13 of this report.
Syraeuaa. Naw York
A 12 ft (3.66 m) diameter Swirl Regulator/Concentrator
was Installed In Syracuse, NY. (8) Design flow to the swirl
was based on maximum carrying capacity of a 24 In. (61 cm)
diameter combined sewer Inlet of 8.9 mgd (.39 o»3/s).
The facility was designed for Immediate response to an
overflow condition. A flowmeter on the 12 In. (0.3 m) foul sever
outlet line measures dry weather flow and the foul concentrate
to the Interceptor during wetweather flow. The average dry
weather flow range Is approximately 0.50 to 0.75 mgd (1.3 to
2.0 mVmln).
Since the downstream capacity of the Interceptor Is
1.3 mgd (3.4 m3/mln), maximum foul sewnr flow Is quickly reached.
Flow in excess of 1.3 mgd (3.4 m'/mln) discharges over the
Swirl's central overflow weir where It Is measured by another
flowmeter, disinfected and discharged to the receiving stream.
When the overflov subsides, a foul sever pump activates
and lowers the water level In the Swirl chamber to allow
free gravity flov In the floor gutter to prevent solids
from settling. Scour velocity Is maintained between storms.
Sampling was performed at the Inlet and outfali locations
during overflow events.
The coarse fIoatabIes/scum removal mechanism worked
satisfactorily. During overflows floatables ware contained by
the scum ring In the outer ring of the chamber and forced
Into the floatahles trap and under the weir plate for wt»t weather
containment. These pollutants were subsequently drawn down and
removed to the foul sewer during dry weather. Relatively good
suspended solids removal efficiencies were determined over the
entire storm flow range of the Syracuse prototype. Tests
30
-------
further verified that the device Is capable of functioning
efficiently over a wide range (10:1) of CSO rates, and can
offectlvely separate suspended matter at a small fraction of
the detention time required for conventional sedimentation or
flotation.
The capital cost of the Syracuse prototype was
$160,930 or $23,670/mgd ($1040 m3/Sec) and $2,'30/acre
($1210/ha) (ENR-2700). It should be noted that the Syracuse
design closely matched full-pipe flood conditions. Design
capacity of the Swirl was not realized In that the facility was
hydraulleally "under-driven" relative to Its design. Withstanding
this limitation, the cost of the swirl technology which performed
effectively at this site was on an order of magnitude* cheaper
than conventional technology.
San Franc Iseo. CA
A 6ft (1.8 m) diameter Swirl Regulator/Concentrator
with a design capacity of 1 cfs (28.3 l/s) was tested In San
Francisco* CA using pumped raw sewage from the Richmond-Sunset
district. (15) Results of the San Francisco analysis found no
significant or measurable benefit attributable to the
concentrating effect of the Swirl Concentrator when operating
on raw sewage. Concentration reduction ranged from 0 - 5%. Mass
removal ranged from 1-4? above that attributable to the flow
split. It was later determined after the experiments were
completed that the primary difference between these results and
the U.S. EPA design tests (3,4) for the unit was the use of
well-mascerated raw sewage with very little of the expected
grit and settloable material. The design parameters for the
Swirl Concentrator were developed using a waste that contained
far greater grit than found In San Francisco Richmond-Sunset
sewage.
The two top curves In Figure 6 "Proposed Grit" and
"Proposed Organlcs" were materials that the APWA working group
specified as being desired to remove In the Swirl
Regulator/Concentrator (4). The third curve from the top In
Figure 6 shows the settling velocities of particles In a
six ft (1.8 m) diameter prototype chamber that were
represented by tne gllsonlte used In the 3 ft. (0.9 m) diameter
Swirl model. It can be seen here that over most of the range
the gllsonlte represented finer materials than those proposed.
The exception was at the lower end of the organlcs where 30?
were not covered.
The bottom two curves on rlgure 6 show the dry
weather sanitary sewage that was analysed at Philadelphia for
the primary settler tests (16) and at San Francisco (15).
31
-------
10.
1.0.
.1.
.01.
'Propose*! G»wi!—*. s*^"^ _. —
(S.G.-2.65) ^" ^-~
'Proposed Organic* "
(S.G.-2.63)
I
/ *Gil»oniteusedon model of 3ft diameter
/ scaled up to 6ft. chamber.
-LR.A.-90 used on model for primary settler tests. (10)
Scaled up to 6ft. chamber. ~~
Philodelphio Dry Weather Sewage 06)
/
/ Asan Francisco Dry Weather Sewage (15)
Ian/sac = 0.03 ft./see.
Graph Preparation (17)
OlOIO SO 40 60 GO 70 80 90 WO
Percent of Particles, with Settling Velocity less than.
Figure 6. Comparison of sewage settling velocity characteristics used in
San Francisco Swirl evaluation with APWA results.
32
-------
Finally, the middle curve Is the artificial media used in the
swirl primary Bottler evaluation tests (10) scaled up to the 6 ft
(1.8 m) diameter chamber In San Francisco.
Several conclusions can be made upon Inspection of
Figure 6. Settling velocities of the San Francisco dry weather
sewage falls entirely outslda the runge of settleable solids
settling velocities for which the Swirl Regulator/Concentrator
was designed. In fact, settling velocities of the San Francisco
sewage were less than those In the reference materials used for
the evaluation cl the Swirl Primary separator. In sum, the
prototype Swirl Regulator/Concentrator should not have been used
to treat wel l-mascerated sewage from the Richmond-Sunset
district. No conclusions about the appropriateness of the swirl
technology can be drawn from the San Francisco experiments.
Sag I na* fjf< I eh I gan 5jiJ.j;J. Qpngentraj-or De.m.o.PaJrBlAQJl EJLfl.lfl.cJt
A. Prior CSO FaelIIty Plan
Abatement of pollution from combined sewer overflows of
mixtures of sewage. Industrial wastes and stormwater from the
City's combined sewer system Into the Saglnaw River began In
1960. An engineering plan focusing on the combined sewer overflow
problem considered utilizing to Its maximum the existing system
of combined sewers. Intercepting sewers, and stormwater pump'ng
stations and proposed the location of storage and treatment
facilities at certain locations determined by the hydraulics of
the system along the route of the Interceptor. (18) The concept
of storage and treatment was considered to be the state-of-the-
art CSO technology during that time.
The CSO control plan focused on utilizing In-line and off-
line storage. In-system storage was available for
retention/attenuation of combined sewage through modification of
regulators, construction of weirs, restraining of tide gates and
other means. In addition, off-line storage at seven locations was
recommended to be constructed to retain and treat wet weather
flows In excess of the collection and transmission system's
hydraulic capacity. Disinfection was recommended to be provided
by Injection of chlorine Into the Influent conduits of each
storage basin by means of feed pumps and diffusing arrangements.
Costs for construction of seven off-line storage/treetment
facilities were estimated to be $25 million in 1972, $37 million
by 1980 standards.
One of the seven basins of the CSO facility plan has been
Implemented at the site of the Hancock Street storm and flood
water pumping plant at a total cosh of $7.1 million. The Hancock
33
-------
Street facilities serving about 1600 acres (700 ha) of combined
sewer'area consist of an Integrated system of In-line storage
contra!!cd ty asdlfled regulator stations; a flood protection
pumping station; and a storage/treatment basin capable of
treating and disinfecting ell overflows.
B. Nat. CSO Fact I Ity Plan
In view of these new technological developments and the
rising Inflationary costs of both The envisioned construction
program and the ensuing high degree of operational requirements
already experienced In the City of Saglnaw applied for, and
received a Section 108 grant from the U.S. EPA Great Lakes
program. Region V. The grant Is entitled, "Demonstration Grant
for Control and Treatment of Stormwater from a combined Sewer
Overflow In Saginaw," and Is divided Into the following four
phases.
Phase I: Preliminary Engineering Feasibility Study
Phase 2: Design of CSO Abatement Technology
Selected In Phase 1
Phase 3: Construction and Installation of Equipment
Phase 4: Facility Start-up and Evaluation
Environmental Design & Planning, Inc. (EDP) completed
the Phase 1 study In June, 1980. (19) The study recommended using
secondary flow pollution control devices (Swirl Concentrators) In
lieu of storage/treatment basins together with extensive
utilization of In-line storage and minor structural system
iiiodl f Icat I on to Increase flow performance of the existing system.
A major element of the proposed CSO plan Is envisioned as the 108
demonstration facility.
Upon request by City of Saglnaw and U.S. EPA, EDP
expanded the Phase I Feasibility Study Into an alternative CSO
Facility Plan for the City. (20) Public hearings were held In
January 1981 and the new CSO facility plan was formally adopted
and submitted to the Michigan Dept. of Natural Resources. Capital
costs for the recommended CSO abatement program were estimated to
be 15.7 million dollars.
The costs do not Include the presently operating
Hancock Street CSO facility.
Webber Street CSO FaclIIty
One of the major overflow points In the system was
selected as the location to construct a major Swirl Concentrator
demonstration facility funded by the U.S. EPA 108 Great lakes
Program and the City of Saglnaw. The facility Is located at the
34
-------
Webber Street Pumping Station and treats overflows from roughly a
1200 acre (523 ha) service area. EDP completed a design
development document for the facility In June* 1982 (21).
construction plans and specifications will be ready for bidding
during early 1982. Construction Is envisioned by late spring.
The facility will consist of an In-line control
structure capable of generating 220,000 ft3 (6235 n»3) temporary
storage (wet well) for a 80 cfs (2260 l/s) lift station feeding a
40 ft (12.2 m) feeding a 40 ft- (12.2 m) diameter Swirl
Concentrator.
The In-line control structure will permit 20 cfs (566 l/s) of
combined sewage to constantly pass to a downstream Interceptor system. The
lift station Is Instrumentated to activate when the temporary storage Is near-
Ing exceedance. Foul sewer underflow from the swirl concentrator will drain by
gravity to an Interceptor system.
4.3 Swirl Dayrit+ar
The swirl separation principle was recognized to be
useful for other than combined sewer overflow regulation. A
natural application of this relatively "flash-type* solids-
liquid phase separation Is the removal of heavier grit
from wastewater flows because such solids are more readily
treatable due to their higher settling velocities. It Is a new
Innovation In the separation of heavy Inorganic solids from
lighter organic materials by selective use of longitudinal
flow velocities. It also offers the opportunity to effectively
remove grit from either the underflow concentrate (foul sewer
discharge) of a Swirl Regulator Concentrator regulator or from
normal dry weather and wet weather Influents Into WWTP.
Hydraulic model tests Indicate that the swirl
concentrator principle can be utilized to provide the same
high degree of performance In settling and removing grit
particles as conventional devices.(7) The dynamic action of
the Swirl Degrltter appears to wash the grit and may result
In a minimum of organic materials settled and entrapped with
the Inorganic grit particles.
The small size, high efficiency, and absence of
mechanical equipment In a Swirl grit chamber facility appear
to offer advantages over conventional devices. In conjuctlon
with a Swirl Regulator/Concentrator serving as a combined
sewer overflow treatment device. It can provide an efficient
means of removing the extremely large concentrations of grit
which can be anticipated In the foul sewer wastes. The device
can be used to provide removal of relatively large quantities
of grit and larger organic material which may cause deposition
problems for the receiving sewer or WWTP.
35
-------
The {idgrirtlng of wastewator is common practice. It Is
one of the conventional pro-treatment stages In sewage and/or
Industrial wastes treatment plants. The removal of Inorganic
grit Is provided to prevent excessive wear on subsequent
handling operations such as pumping, comminuting and screening of
sewage and pumping of sludge. Elimination of Inert solids
prevents deposition cf such material In settling tanks* sludge
hoppers, sludge digestion chambers, aeration chambers,
pipelines and other i cations.
Design of severs Is bated on the principle that
average sewage sol ids - organic and Inorganic In character -
can be held In suspension In a so-called "self-scouring" sever
at flov velocities over 2 ft (0.61 m) per second. Similarly,
grit chamber design Is based on the principle that heavier
grit will settle at velocities of flow of 1 ft (0.3 m) per
second, while lighter organic^ will be held In suspension
under these hydraulic conditions until they reach settling
chambers where flow velocities are reduced to rates In the
general range of I ft (0.3 m) per minute more or less. This,
then, Is the basic criterion for the separation of sollds-from-
sollds In grit units, and the separation of sol Ids-from-l(quid
In clarification or settling chambers.
The application of the Swirl Concentrator phenomenon
to the task of grit removal Is dependent on the ability to
provide flow velocity conditions and Internal hydraulic
patterns which will separate heavier, larger solids particles
from lighter, smaller materials and to allow the two
separated classifications to be collected and removed at
separate points.
The general features of the Swirl Oegrltter are
Identified In Figure 7, and are described as follows:
(a) Inlet - The Inlet dimension Is normally designed to
allow a Inlet velocity of 2 ft (0.61 m) per second. On this
basis the Inlet diameter becomes the controlling dimension
for sizing the unit. A set of curves has been developed
to express the relationship between flov. Inlet dimension,
chamber width and depth. (5) The flow Is directed
tangentlally so that a "long path" pattern, maximizing
solid separation In the chamber, may be developed.
(b) Covered Inlet - The Covered Inlet Is a square extension
of the Inlet which Is the straight I I no extension of the
Interior wall of the Inlet extending to Its point of
tangency. Its location Is important, as flow which Is
completing Its first revolution In the chamber strikes, and
36
-------
A. INLET STRUCTURE
& DEFLECTOR
C. WEIR AND WEIR PLATE
0. SPOILER
E. FLOOR
F. OOMCAL HOPPER
Figure 7. Isometric view of Swirl Concentrator as a grit separator.
37
-------
Is deflected Inwards, forming an Interior water mass which
nakes e second revolution !n the chamber, thus creating The
•long path" flow pattern. Without the deflector, the
rotational forces would quickly create a free vortex within
the chamber, destroying the solid separations efficiency. The
Height of the deflector Is the height of the Inlet port,
Insuring a head above the elevation of the inlet, a feature
which tends to rapidly direct solids down towards the floor.
(c) Qvarf low yeJr Aflii Jfil£ Plate - The diameter of the
weir Is a function of. the diameter of the chamber, and of
the Inlet dimension. Under normal condltons, the weir
diameter Is equal to the chamber diameter minus twice the
Inlet dimension. The depth or vertical distance from the weir
to the flat floor. Is normally twice the Inlet dimension.
The height, or rise, of the weir plate Is normally
0.25 the Inlet diameter. The weir plate connects the overflow
weir to a central column, carrying the clear overflow to the
Interceptor and primary treatment. The horizontal leg of the
downshaft should leave the chamber parallel to the Inlet.
(d) 5jifl.LJL.fl£.s.jL - Spoilers are radial flow-guides,
vertically mountec o the weir plate, extending from the
center shaft to :' -3 edge of the weir. They are required
to break up the rotational flow of the liquid above the
weir plate, thus Increasing the efficiency of the weir and
the downshaft. The height of the spoilers Is the same as the
Inlet diameter. This proportionately large size, as compared to
the Swirl Regulator/Concentrator Is required because of the
possible large variations In diurnal flow which may be
anticipated.
(e) EJ.fi££ - The floor of the unit Is level and Is In
effect a shelf, the width of the Inlet.
(f) Qenfrral M&lUl.fi£ - The central hopper Is used to
direct the settling grit particles to a single delivery point
where they may be removed to a conveyor for washing and
removal from the system. The hopper Is at an angle of 60
degrees to the floor. If the angle Is less thun 45
degrees, particles will build up at the lip. As the angle Is
Increased, the problem decreases to an optimum condition at
60 degrees.
The downshaft elbow must be sufficiently below the
floor to prevent formation of eddy currents. Th!s depth
appears to be the Inet diameter. Structural supports for tne
elbow and actual pipe connections must be designed to
prevent rags from being caught on a protruding bolt
head, flange or strut. The downshaft should exit parallel
38
-------
to the Inlet to assure a minimum hydraulic Interference for
settlIng particles.
Full-scale evaluations of the Swirl Degrltter have
been accomplished In Denver, Colorado and In Tamsworth,
Austral la.
Denver. Colorado
A prototype Swirl Degrltter was tested by the
Metropolitan Denver Sewage Disposal District No. 1.(7) The unit
«as designed to duplicate the grit removal device needed to
degrlt the underflow from the proposed swirl concentrator at
Lancaster. Degrlttlng Is considered In Lancaster to protect
pumps and prevent sIItatIon In the Interceptor.
The 6 ft (1.8) diameter device was designed for a flow
of 1.5 mgd (656 l/s). It was found that under the physical
arrangments In Denver, and testing with domestic sanitary
wastewater, that the Swirl Degrltter performed at slightly less
efficiency than the conventional aerated grit unit which was
operating at less than twice the normal flow-through rate. The
characteristics of the grit removal from the Swirl Degrltter
were excellent and particles of .008 In (0.2 mm) were removed.
Analyses of grit removal was accomplished with three
Chaslck sampling units. Blasting sand was added to provide
extremely high concentrations of .008 In. (0.2 mm) particles
(lower definition of grit) to duplicate the concentrate from the
swirl regulator. It was found that the unit could efficiently
remove the small particles at the high concentrations.
TamsMorth. NSW. Austral la
A full scale 22 ft (6.7m) diameter Swirl Degrltter
was Installed as part of the river Intake works of the
Tamsworth Supply Augmentation project In New South Wales,
Australia. (22) The Swirl Degrltter was designed for 1he pre-
treatment of river water prior to Its entrance Into the rising
main In order to reduce wear and tear on the raw water pumps
and also to reduce the solids loading of the rising main and
that of the bal ance tank of the water treatment works. The rate
of flow through the system Is governed by variable speed
pumps which Is automatically set by telemetry according to the
water level In the balance tank of the water treatment works and
the available water In the river. Influent flows vary from
2380 gpm (150 l/s) and 1340 gpm (850 l/s). Grit collected In the
bottom of the hopper Is removed with a water-Jet educter pump
through a grit discharge pipeline back Into the river. The
field evaluation program was Initiated with the overall
39
-------
objective of providing Information on the behaviour of a
full- scale Swirl Degr|t+«r, designed and constructed in
accordance with the shapes and proportions developed during
model studies. Results of the solids removal had been
evaluated In terms of three parameters: solids larger than 0.2 mm
- the classical size aimed at In grit chambers -, solids
larger than 0.088 mm and total settleable sol Ids. In general
the tests proved the validity of the laboratory results and at
design flow rates 98% removal efflclences were achieved.
4.4 Hal leal Bend
The Helical Bend fIov regulator Is based on the
concept of using the secondary helical motion Imparted to
fluids at bends when a total angle of approximately 60
degrees Is employed. Figure 1, Chapter i Illustrates the device.
Model studies have confirmed the pattern of solids
deposition In the deeper channel portion of the Helical
Bend* located along the Inner circumference of the bend
section, and the ability of dry-weather flows In this
restricted deep channel to self-scour the deposited solids
Into the foul sewer outlet.(11) The basic structural features
of significance In the Helical Bend aret the Iflifll Sgetlon
from the entrance sewer to the device} the Transition SaetIon
from the Inlet to the expanded cross section of the Straight
SaetIon ahead of the bend; the overflow side weir and scum
board (Treatment SacjtJonJf and the foul trough, together
with the means for controlling the amount of this underflow
going to the treatment works.
A more detailed description of the Helical Bend and
Its evaluation and testing Is given In Chapters 6,8,10 and 13
of this report.
4.5 DIgeostraInar
The Dlscostralner Is a disc-straining device which, because
of Its design. Is able to separate solids by a combination
of straining, sedimentation and filtration. The device
consists of a series of stainless steel mesh-covered discs
In a specifically designed chamber (see Figure 8). Each pair
of discs forms a separation unit. The Influent Is routed
Into the cavity belween pairs of discs and flows outward
through the wire mesh leaving the captured solids behind.
Inese trapped solids form a pre-coat material which may aid the
filtration effect. A spray system Is used to flush the
trapped solids from the mesh hack I ntr the cavity between the
discs. The spray system can utilize filtered water with a
built-in reelrcuI at Ing pump or an external water supply. As
40
-------
SIDE VIEW
DISC NOTATION
SCREEN MESH SUPPORT STRUCTURE
^_^_ I
SPRA»SYSTEM
FRONT VIEW
DISCS
DISC HIM
f LKXIBLK SEAL
INFLUENT PR6COAT
FORMATION
EFFLUENT
TOP VIEW OI8C» DISC SNAFT 8OUOS
SCHEEN MESH
SUPPOR' .TRUCTURE
Figure 8. Construction and operating features of discostramer.
4 i
-------
the solids concent.-atlon between the discs Increases, the
slurry In the downstream end of the cavity becomes
sufflcently thick so that the solids can be swept up by
the rotating discs and out through the solids discharge
opening, or adjustable weir window*
In an evaluation of the Dlscostralner a high speed
reclrculatlng Jet was used to mix the concentrated slurry and a
variable-speed peristaltic pump was used to meter the
concentrated slurry Into the mixing tube of the Influent line.
(23)
The DI s*?ostr a I ner was operated with Inflow rates cf
189, 379, 473 and 568 I pm (50. 100, 125, 150 gpm), fluxes of
126, 253, 315, 379 Ipm/m2 (3.1, 6.2, 7.8, 9.3 gpm/ft2), and
Inflow solids concentrations of 850 to 2270 mg/l. The headloss
through the screens was determined by measuring the difference
In elevation of the water level In the Inlet chamber and the
bottom edge of the discs. The Influent and effluent samples
were taken at the Inlet chamber and effluent pipe,
respect I- vely. The samples were analyzed for solids
concentration using the same procedure as for the tube
settler. Flow was established In the Dlscostralner by first
setting the primary flow of clear water, then starting the disc
drive and sprayer pump. The slurry feed system was then turned
on to Inject the concentrated sediment slurry Into the
I nfI went I Ine.
Representatives of the local distributor of the unit,
Northwestern Power Equipment Company, were notified and observed
the Initial tests. The Initial runs with the Dlscostralner
were entirely unsuccessful. Immediately after the slurry flow
was started the mesh plugged or blnded, and the water level In
the Inlet chamber Increased until the water discharged through
the overflow outlet. The binding continued as long as the flow
was maintained, even after the slurry feed line was turned off.
The sprayer system cleared the mesh as the discs rotated past
them but the mesh blnded again as soon as It »-e-entered the
flow. There was sufficient material trapped between the *1lscs
to cause the nesh to bind even when no additional slurry was
added. This rapid binding was partially caused by a deficiency
In the slurry feed system which caused a surge of high
sediment concentration to enter the Dlscostral ner when the
slurry flow ras first turned on. The slurry feed system was
subsequently modified so the Initial solids concentration
could be controlled an** so long as the Inflow concentration
was carefully monitored, the screens could be prevented from
blinding. However, the persistence of this problem Illustrates
the sensitivity of the square-mesh screen to plugging by
granular particles.
42
-------
Reou I
The Teacup Regulator/Concentrator resembles a hydro-
cyclone In external appearance (seo Figure 9). A properly
designed Teacup unit generates two Independent velocity
profiles. One Is a "forced- vortex" velocity profile which
propels dense solids to the periphery of the Teacup due
to high angular velocities, and Induces a "Teacup"
secondary velocity capable of sweeping sol las toward the
center of the bottom of the Teacup for concentration
In the underflow directed to further -treatment. "The forced
vortex" Is Induced In the unit by the top tangential
Introduction of the Inflow. The second Is a "free
vortex" velocity profile which Is characterized by a
constant angular momentum which can be adjusted to
separate less dense organic solids and floatables from the
dense Inorganic solids forced to the periphery of the
unit. Thus, the "free vortex" car* be "tuned" to achieve
classification of lighter from heavier matter.
A full-scale Teacup solids separation system was put Into
operation on a carrot washing system at Magglo-Tostado, Inc.,
hermai, CaI IfornI a.(24) The Teacup solids separator was 4 ft
1.22 n) In diameter and had a design flow of 500 gpm (32 l/s).
Throughout this test period, however, the unit operated at
333 gpm (21 l/s). Typically, the Teacup removed In excess
of 95% of the total suspended solids. The remaining suspended
solids were essentially colloidal In nature. It was also noted
that even greater removal levels were achieved when alum
was Inadvertently added to the waste stream.
During 1978, a Teacup solids separator was evaluated for
treating raw domestic wastewater at the Sacramento City Main
Wastewater Treatment Faci I I ty.(25) The specific objectives were
to evaluate removal efficiencies for settleable solids and
floatable materials, and to compare the Teacup's performance to
that of the City Facility's conventional grit removal process.
The pilot unit was 5 ft (1.52 m) In diameter; had Influent,
overflow, underflow and floatable flow line diameters of 6 In.
(15.2 cm), 10 In. (25.4 cm), 4 In. (10.2 cm), and 2 In. (5.1 cm),
respectively; and treated flows up to 900 gpm (57 l/s). Solids
removed In the underflow were very similar to grit with an
average solids concentration of 62$ and average volatile
fraction of 13$. Analyses of solids taken from tfre City's
grit chambers showed 92$ solids and a volatile fraction of 23$.
The City Facility staff reported that 0.5 ft3 (0.14 m3) of grit
was removed per Mgal of wastewater treated with minor
fluctuations of 0.1 f J-3 (0.003 m3) or less. The Teacup removed
43
-------
Tcp tangential introduction
inducts forcod vortox
and Mcondory "teacup" ""v.
velocity profile y
Bottom-center opening for
entrapment of eettleoblo
eolfde
Quieecent thickener
ConHr drowoff induces free
vortot velocity profile
I*P_
L Fore
port tele trajectory:
ced vortex and "teacup"
•eoondary velocities
carry particl* to bo**om
•enter opening
t.Free vortM «xclud*t
dl*n«* Mttltobto partieltt
from drawofff
TMckonod underflow drown
off at roquirtd
Figurt 9. Footurts of tht Teacup Solids Separator.
-------
between 3.05 and 3.44 ft3 (C.086 and 0.097 m3) of solids per Mgal
treated or from 6 to 7 times as much as removed by the grit
4.7 Dynamic Sol I ds Separator
The Swirl Regulator/Concentrator had Its origins from a type
of stormwater overflow built In Bristol, England In 1963-64
under the name of Vortex Separ ator s.(2) Considerable
research work had been carried out In Bristol In an
attempt to Improve the stability of the flow pattern In
sedimentation tanks for sewage treatment and to develop a
compact form of stormwater overflow treatment for a location In
a very congested urban area. It was not expected that any
significant separation of solids could occur at the high
velocities necessary In such compact structures, but both
the model tests and later tests of the full sized structures
showed that a considerable proportion of the solids In the
flow was retained In the foul water outlet.
Several Vortex Separators constructed In Bristol were
described In a paper given at a Symposium on Storm Sewage
Overflows held at the Institution of Civil Engineers In
London In 1967. (2) This paper led to the continuation of
the research by the APWA under the aegis of the U.S.
Environmental Protection Agency with a specific project at
Lancaster, Pa. under consideration. The name of the device
was changed to a Swirl Concentrator to avoid the use of
the word "Vortex" which Is almost universally associated with
"free" vortices, a form of vortex which tends to mix rather
thin separate. The pattern of flow In the Vortex Separator
approximates a "forced" vortex and not a "free" vortex.
An 18 ft (5.5 m) Vortex Separaror stormwater overflow
cor.structed In Bristol out of brick had a free outlet for
the ' orm sewage overflow. (Photographs of this unit are
shown . .- Figure 10). This design and the design of the Swirl
Regulator/Concentrator need a drop of several feet between
the Incoming sewer and both the foul outlet and the
storm outlet, in most existing drainage systems this Is not
available and the device can only be used where the
receiving water course Is below the level of the combined
sewer or where the pumping Is feasible. In completely new
drainage systems the necessary head to use Swirl
Concentrators as stormwater controls can be "built- In". The
really serious problem, however. Is In older systems
which are overloatiaa and often In areas where reconstruction
Is prohibitively expensive.
45
-------
A. Influent section 1n upper
left background
B. Overflow Weir
Figure 10. Photographs of Vortex Separator
constructed in Bristol, England.
46
-------
In an endeavour to solve this problem, research has
continued In England during the same period as the svlrl
technology evaluations In U.S. A patented device known as the
Dynamic Separator has been developed. Figure II shows an
Isometric view of the units with labelled components and a
photograph of a 3.3 ft ( 1m ) unit. This device carries out
the same function as the Swirl Regulator/Concentrator but
with the minimum possible head loss, only the energy needed
to rotate the flow being taken from the sewage flow.
The main difference between the Dynamic Separator and
the Swirl Regulator/Concentrator Is that the decanted flow
Is taken out of the main chamber upwards through the
top of the tank not over a weir and down through the
base of the tank. The chamber Is kept 'nearly full even
under dry weather flow conditions and Is continuously
rotating under the Influence of the tangential Inlet. The
velocity of rotation Increases as the Inlet flow
Increases.
Vortex Separators have been successfully used for treating
a wide variety of different types of waste streams. Capsulated
reviews of four pilot plant evaluations not appearing In
American literature are reported here. The first effort
Involved use of the Vortax Separator to remove humus from
a trickling filter plant InBlackwell, England In 1966.
The second two Investigation demonstrates that the Hydro
Dynamic Solids Separator Is particularly well suited to the
needs of Industry for the removal of particles from waste
process water, thus enabling reuse. The final • «t o
Investigations document the value of this unit for normal WMTP
applIcatlons.
BlaekwelI PI lot Plant
In 1966 Maynard, Froud and Stevens Consulting Engineers
(26) proposed a pilot plant treatment system for the grossly
overloaded sewage works at BlaekwelI, England. The works
constructed In 1928 consisted of a grit extraction with coarse
screening, primary settlement In horizontal flow settlement
tanks (originally providing 4 hours detention), percolating
filters that were breaking up under the load and ponding, humus
settlement In horizontal flow tanks and sludge drying lagoons.
The proposed pilot plant consisted of a Flocar pvc
filter with 64 cu yds (.8 m3) capacity measuring 8 ft x 12 ft
x 18 ft high (2.4m x 3.7m x 4m). It was raised to discharge 8
ft (2.4 m) above ground level to provide the necessary head to
feed to a Vortex Separator. The feed to the Flocor filter was
by pump up to the top where a ring pressure main del ivered
47
-------
-p.
CD
KEY
A. Inlet pipe
B. Treated Flow Outlet
C. Solids Outlet
D. Cross Section Flow Pattern
E. Inlet Flow Control
F. Treated Flow Outlet
G. Dip-plate to trap Floating Solids
to a Fleatables Trap
H. Rotiding Eye
J. Base Cone
K. Solids Constrained Under Cone
L. Outlet Flow Control Vanes
M. Solids Collection /one
Figure 11 Hydro Dynamic Separator.
-------
to sparges « I th splasher plates fixed to ensure fairly even
distribution over the F'.ocor. This even distribution was
disturbed when any of the sparge pipes blocked with solids In
the sewage and the pipes had to be checked and cleared dally.
The pump was a Mono pump delivering 125 gpm (4721/m) but a
bypass pipe enabled the flow to be varied by bleedlng-off some
of the flow.
The vortex-type settlement tank was In an 81 in.
(2.06 IB) diameter tank* 5 ft 6 In. (1.68 m) deep with a central
weir. The feed was tangential through an orifice 6 In. (15.2 cm)
long under a pressure of from zero to 3 ft (.9 m) head of
water. The sludge removed from the central sludge well In
the tank passes continuously Into an associated sludge
holding/thickening tank. The foul flow from the Vortex Separator
was maintained by a constant bleed-off from the sludge tank.
Figure 12 shows a photograph of the top view of the Vortex
Separator and the holding tank.
Up to this point the vortex-type settlement tank
had never been tested for use as a sedimentation tank. The
vortex tank does not work as a centrifuge* the velocities of
flow are not sufficient to give an appreciable centrifuge
effect. The principle on which the separation works Is where the
water Is spinning, the lighter particles move faster and
reach the center before the heavier particles. The lighter
particles leave the heavier particles to fall behind, outwards
and down to the bottom of the tank.
The flow pattern In the vortex Is not simple. The
splitting of the flow Into dense and less dense regions causes
a break-up of the forced vortex that exists In the outer
third of the tank. There Is a region of turbulence where a
secondary epI eye lie vortex motion forms. Two of these vortices
form In this region about the central weir with their centers
always opposite. These epicycllc vortices are a critical factor
for efficient tank performance and wheii they are eliminated, the
tank performance Is negligible. They consistently form In the
tank and their effect Is to trap the denser material and pull
It to the bottom of the tank where the flow Is towards the
center and where the sludge well In the center collects all
the solids. The less dense flow, moves round these vortices
forming a free spiral vortex at the centre of the vank and
passing over the centra! weir. A photograph of the unit In
operation Is shown In Figure 13.
The pilot plant was tested by sampling and analyzing the
effluents from each unit. The sampling points Included raw
sewage, flow to Flocor. f I or» to vortex-type tank and ftr.al
ef fIuent.
49
-------
Figure 12.
Reproduced from
b«st available copy.
Photograph showing top view of Vortex Separator
and Sludging Holding Tank, Pilot Plant, Blackwell
(UK).
50
-------
*tprodue«d from
b«s> •vii COBV
Figure 13. Photograph of Vortex Separator 1n operation.
Blackwell (UK).
51
-------
The flow pattern through the treatment plant was as
follows. Raw sewage passed through grit channels and then to
primary settlement to the dosing chamber. From the dosing
chamber the main flow went through percolating filters, humus
settlement and discharge to the River Kenn. The pilot plant feed
was drawn Initially through the Flocor and then on to the
vortex-type settling tank and returned either to the dosing
chamber or at the Inlet to the works or straight to humus
settlement. Later modifications enabled the pilot plant to draw
raw sewage.
For the eval uatlon period the pi lot pi ant was set at a f I ow
rate of 98 gpm (370 l/m). The flow from the pilot plant was
returned during this period either to the dosing chamber or to
the Inlet of the works. In either case the returning effluent
diluted the Incoming sewage and reduced the strength of the
flow to the pilot plant. Further operation of the pilot plant
Included a series of tests In which the vortex-type settling
tank was run parallel with a quiescent setrlement tank In
order to compare their efficiencies. The first test was
carried out using crude sewage passing straight to the vortex
tank and to one of the detritus tanks adapted for this purpose.
The quiescent tank or adapted detritus tank had a 540
gal (2049 I) capacity. Temporary dams were built to Isolate the
tank from the normal fiow. Its depth was 3.25 ft (.99 m) and
length 10.5 ft (3.2 m). The Inlet to the tank was over one of
the weirs and control of the flow was accomplished by
allowing some of the flow to splash outside the tank. Smooth
Inlet Into the tank was then ensured by a series of baffles and
temporary dip-plates.
The efficiency of both tanks using raw sewage as feed was
low. The average efficiency of the vortex -tank was 15.3$ and
the qu'escent tank 12.5$ removal of suspended solids. The
vortex was 1.22 times as efficient as the quiescent
settlement tank.
Further testing Invol/ed altering the Influent TO both
tanks to evaluate settling after Flocor treatment. Feed to both
+anks averaged 193 mg/l suspended solids. Tests were carried
out with detention times of from 50 to 90 minutes In the
quiescent tank and 25 minutes In the vortex tank. The results
were more consistent, the vortex tank oelng on average 82$ as
efficient as the quiescent tank. One explanation of the poor
removal results on these runs could be the unusual
characteristics of :he sewage. Testing of the waste Indicated
52
-------
a high level of unsettleable or col It/Ida! particles. The a*e of
a coagulating agent was recommended If used In primary
sedimentation prior to Inflow Into the turbulent vortex
chamber.
To arrive at the settleable solids removal efficiency of
the vortex and quiescent tanks, an analysis of the sewage
after two hours settlement In the laboratory was taken. The
results showed that the vortex was 19% as efficient on total
suspended solids and 77) as efficient on settleable suspended
solids as the quiescent settlement tank. One run of samples
was taken to confirm the above results In which the sludge
from the separator was sampled as an extra check. The
results showed that the vortex tank was 80) *>z efficient on
settleable suspended solids as the quiescent tank.
It was concluded that the normal performance of the vortex
tank would provide removal efficiency of 60-80) the effclency
of a large quiescent settlement tank, depending on the
strength of sewage. This loss of efficiency Is In the removal
of very small particles that can floculate and be removed In
the quiescent settlement tank but cannot form In the vortex
tank. A summary comparison Is presented In Table 2.
The cost of the vortex-type sedimentation tank Is
estimated to be approximately 1/10th the cost of corresponding
conventional sedimentation ranks. Therefore with the use of
the vortex separation the cost of treatment would be only
10) of thi- of conventional settlement tanks and provide 80)
of the sol •£ removal as realized In conventional tanks.
53
-------
Comparlsgn of
TABLE 2
Solids Removal Effectiveness of Vortex Separator
with Quiescent Settling Tank
PERFORMANCE fl£ VORTEX TANK
(a) Total Solids
Ran
01 ear
Efficiency
26% of the total sol Ids
• vent to siudge
(b) Settleable sol ids
Raw
Cl ear
Efficiency
53* of the settleable solids
went as sludge
10.92 Ib/hour
7.12 Ib/hour
33*
4.60 Ib/hour
1.72 Ib/hour
62.6%
PERFORMANCE ££ THE QUIESCENT TANK
(a) Total solids
Raw
Cl ear
Efficiency
(b) Settleable Sol Ids
Raw
Raw
Efficiency
Detention time
1 kg » 2.2 Ib
0.8 Ib/hour
o.47lb/hour
41$
0.3 Ib/hour
0.08lb/hour
73*
In Vortex «• 22 minutes
In Quiescent tank 130 minutes
(Tuaacero) Monterrey r Max I to
Hydro Research de Mexico was commissioned In 1981
to Investigate the technical feasibility o* treating and
reusing wastewater from a .nijor steel pipe manufacturing plant
(Tubacero) In Monterrey, Mexico.(23)
54
-------
Principle water consumers at the plant are the pipe
washing section, the high pressure pipe expander and pressure
testing machine. The main consumer being the pipe washing
section using an average of 40,000 - 53,000 gal (150 to 200 m3)
of water per day producing two effluents. The first waste stream
Is from the scrubbing and first-rinse cycle and Is heavily laden
with rust particles, mill scale. Insoluble oil, detergent and
potassium dlchromate. The other waste stream Is from the flnal-
rlnse cycle and Is relatively lightly contaminated with the
pollutants listed above.
In the analysis of the system the mode of operation
consisted of the effluent being taken from a settlement tank
ind pumped up to the header tank where a constant level was
i ilntalned by the overflow pipe. The header tank supplied
«. .'luent to a 3.3 ft (1.22 m) diameter. Dynamic Separator through
a 4 In. (10.2 cm) and 2 In. (5.1 cm) flexible hose connected to a
butterfly flow valve, enabling the adjustment of the flow rate
to the device. The treated effluent and underflow were taken
off respectively through 4 In. (10.2 cm) and 2 In. (5.1 cm)
flexible hose and returned to another settlement tank. After
filling the device the system ran for 15 minutes allowing
the process to stabilize before sampling. Four different sets
of experimental conditions were run with the pilot plant.
Once the color of the effluent from the header tank
overflow, the treated effluent and the underflow was visually
noted to be consistent for 5 to 10 minutes, samples were taken
using Imhoff cones for volumetric analysis of the effluent.
Values reported In Table 3-A are for volumes of suspended
solids (In ml/1) recorded after half an hour settlement.
Values reported In Table 3-B represent the gravimetric
analyses of results. Efficiency levels noted In Table 3
are defined as the ratio (f) of foul sewer solids mass
rate divided by sum of foul sewer and clear solids mass
rates. Applied overflow rates ranged from 1700 to 3500
gal/«t2/day (.80 - 1.60 !/s/m2>.
Efflcle;.cy adopted In this report for noting
performance of the West Roxbury and Lancaster demonstration
facll Itles Is defined as the percentage of fo'H mass divided by
Influent mass percentage less the percent foul underflow. (See
section 11.6) Comparable solids removal of T Iclencles In Table 3-B
would range from 39$ to 65%,
55
-------
TABLE 3.
Dynamic Solids Separator Performance Summary
(Tubercero) Monterrey, Mexico
A. Vol umatr Ic Ana-lys ts
Vol. of Sol Ids (ml/1) Applied Rate Flovof Eff. +
of Flow (l/s) Sol Ids (ral/s) $
« •• »•• * •« •«« « «* «•«
28
10
12
26
2.2
8.5
9.0
2.5
550
52
40
175
1
1
1
0
.87
.87
.92
.94
1.79
1.61
1.36
0.80
0.08
0.26
0.56
0.14
52.4
18.7
23.0
24.4
39.0
13.7
12.2
2.0
42.9
13.5
22.4
24.5
52
49
64
92
.0
.7
.7
.0
B. GrayImetrIe AnalysIs
Rate of Flow of Sol !ds
Suspended Solids (g/sec) Efficiency
(g/l) % Underflow ++
* *• •«• • «» «*•
0.255
0.200
0.120
0.758
Key:
0.267
0.197
0.112
0.273
4.646
1.210
0.457
6.350
0.48
0.37
0.23
0.71
0.48
0.32
0.15
0.22
0.36
0.31
0.26
0.89
43
49
53
80
4
14
29
15
* Influent
** Clear
»»• Foul Underflow
+ Efficiency » underflow solids flow/(InfIuent & clear
solids flow)
++ Ratio of foul flow to applied flow
I gal/s - 3.78 l/s
56
-------
(Aqua IndustrI a I ) Monterrey. MQX teo
The Dynamic Separator was also Investigated by Hydro
Research de Mexico for the treatment of raw sewage at Aqua
Industrial, Monterrey, Mexico on June 23-25, 1981.(24) Sewage Is
purchased, treated and then mixed with groundwater to serve as
process water for a large Industrial manufacturing plant. This
Is a common practice In watershort areas In Mexico. The
aforementioned unit tested at Tubercero *as set up alongside
the open supply channel at the Intake works to tne Industrial
pi ant.
Raw sewage was fed through the Dynamic Separator via
the header tank. The rate of flow through the Separator was
controlled by either adjusting the level of the header tank
before starting the test or by use of the butterfly valve on
the Dynamic Separator Inlet pipe. The treated effluent and
underflow discharge pipes were taken over the measuring tank
and kept at a constant* height during normal flow, and when
actually measuring the rate of flow of effluent or under flow.
The proportion of underflow to treated effluent was varied by
means of the ball valve on the underflow discharge pipe to
determine the efficiency of the Dynamic Separator at varying
ropot-tlons of underflow to treated effluent.
At each setting samples of raw sewage treated
effluent and underflow were taken In Imhoff cones. Volumetric
readings for settled solids wero taken at Intervals of 15
minutes and 30 minutes for each sample. The treated effluent
and underflow samples were taken 5 to 10 minutes after the raw
sewage sample In an attempt to take Into account the time the
raw sewage takes to pass through the system and to come out as
treated effluent and underflow. The raw sewage sample was taken
from the overflow at the heeder tank. Immediately after
sampling, the rates of flow of treated effluent and underflow
were measured using a measuring container and a stopwatch. For
small flows less than 8.7 gpa '0.55 l/s), a 4 gal (15 I) capacity
bucket was used and for flows greater than 8.7 gpm (0.55
I/sec) a 53 gal (200 I) oil drum van •jied.
Agua Industrial did not possess the required
filtering apparatus to permit the accurate measurement of
coarse sand and girt. The larger particles which separated
rapidly from the rest of the sample In the Imhoff cones
were therefore only measured vo I umetr I ca I I y and are not
Included In the gravimetric results for the total suspended
solids. Both volumetric and gravimetric solids performance
results are presented In Table 4.
57
-------
Sufficient design da+a MBS obtained from the testing
program to permit the design and predict the performance of
Dynamic Separators up to 4800 gpm (300 l/s) capacity. '• 13 ft
(4o) diamoter Dynamic Separator was recommended to provide the
additional primary treatment capacity required by the Aqua
Industrial plant. It will provide the equivalent treatment of two
74 ft (22.5m) diameter primary sedimentation tanks. The Dynamic
Separator Is considered to be far more economic to purchase,
Install, operate and In Its use of the available land than
equivalent primary settlement tanks.
TABLE 4
Dynamic Solids Separator Performance Results
(Agua Industrial) Monterrey, Mexico
Test
No.
Say
I' ay
Day
Day
Day
Day
1
1
1
2
2
2
.late of
Flow
l/s
2.19
1.87
1 .72
1 .37
0.92
1.36
linden 1 ow
Ratio
.16
.16
.32
.41
.36
.20
Hydraul Ic
Surface
Load! ng
l/s/m2
1.93
1.65
1.52
1 .21
0.81
1.21
Efficiency
*
20
32
40
55
52
25
($) Efflcl
**
27
41
51
55
88
28
ency
* Total so!Ids removal (gravimetric)
** Settleable Solids Removal (Volumetric)
Efficiency (£) * 100 (underflow mass rate)/(underfIow clear
mass rate)
Mote: 1.0 i/s/n»2 - 2113 gal/day/f+2
I gal/s » 3.78 l/s
58
-------
TlI bury Sewage Works
In November/December, 1981 Hydro Research & Development
(UK) Ltd. under subcontract to Peterson Candy International (UK)
Ltd. Investigated the potential application of the Dynamic
Separator at the Tilbury Sewage Works for the Anglian Water
Authority. The purpose of the Investigation was to ascertain
whether grit or. ly could be classified or removed from the mixed
liquor effluent of a pilot plant deep shaft so that the organic
fraction treatment process could continue to aerobic
settlement/digestion and ultimately ocean disposal. The mixed
liquor contains approximately 5000 mg/l suspended solids* badly
foams and contains considerable floatable material.
DegaslfIcatlon would normally be necessary prior to application
of conventional sedimentation.
A 3.3 ft. ( I m) diameter Dynamic Separator was used
for the pilot study. Influent overflow rates of deepshaft
effluent to the Dynamic Separator were varied from 450
gal/ft2/day (.22 l/m2/s) up to 4000 gal/ft2/day CI.91 l/m2/s). At
a.i applied hydraulic load'ng of 450 gal/ft2/day (.22 l/m2/s) an4
an underflow ratio of 17.6$, the resulting suspended solids
concentrations of the overflow and foul sewer were 610 mg/l and
11,730 mg/l respectively, providing an efficiency of 80.5$ (foul
mass/1 nfIuent miss) or . net efficiency of 62.9$. At an
Intermediate hydraulic loading of 1500 gal/ft2/day (.70 l/m2/S)
and an underflow ratio o< 10.9$, the resulting suspended sol Ids
concentrations of the overflow and foul sewer were 300 mg/l and
12,730 mg'l respectively, providing a gross efficiency of 83.9$
of a net efficiency of 73$. The desired classification (grit only
In foul) occurred at a hydraulic loading of about 1.91 l/m2/s. An
unexpected floculatlon effect (and benefit) occurred at the lowbr
vdraullc loadings. Further Investigations are proceeding as of
mid-December, 1981.
59
-------
CHAPTER 5
Study Area Description - Lancaster Swirl Facility
5.1 Foreword
General area and sewerage system characteristics are
presented in section 5.2. RaInfa I I/runoff and water quality
monitoring results obtained for the area are given In
section 5.3.
5.2 Area and Sawaga System Character I sties
The City of Lancaster Is located In southeastern
Pennsylvania, approximately 50 miles (80.5 kilometers) west of
Philadelphia. The City Is situated on the western bank of the
Conestoga River, which Is Included In the Susquehanna River
Basln;
The City of Lancaster Is the urban and socio-economic
center of the Lancaster, Pennsylvania Standard Metropolitan
Statistical Area (SMSA), which encompasses the area of
Lancaster County. In 1970 the City population numbered 57,589
residents, which Is 18* of tho SMSA's population. The City's
area Is 7.2 square m I I as (18.6 square kilometers) with the
remainder of Lancaster County being primarily farmland dotted
with several smaller rural towns.
The City of Lancaster was founded In the early 1700's
and Is the oldest Inland city In the United States. The
first City sewage facilities were InsJ-ailed In the 1860's.
During tills early period of construction, brick arch sewers
were Installed to combine and convey both wastewater and
stormwater runoff from the central sections of the City
to the Conestoga River at several overflow points. This
brick sewer construction continued throughout the City
until the mld-1930's when most of the major Interceptors
were completed. However, construction following the turn of
the century separated the flows Into parallel sanitary
and storm sewers.
In 1934, two separate 6 MGD (262 l/s) secondary sewage
treatment plants, the North and South Plants, were
constructed to treat the sewage flows. Existing CSO outfall
locations were connected to pumping stations that carried
the flows to the South Treatment Plant. The North Plant
received only combined gravl+y flow. Since must of the
City's system Is --omblnod sanitary and storm sewerage, the
system contained diversion chambers at the pumping stations.
60
-------
During peak stormwater flows, when the capacity of the
pumping stations was exceeded, the diversion chambers sou!*
channel excess sewer directly Into the Conestoga River. In
addition, if the peak hydraulic capacities of the treatment
plants were exceeded, excess flov would also be diverted to
the river drainage district areas are shown on Figure 14.
The Stevens Avenue drainage district Is located In the
eastern side of the City (see Figure 14). The district
covers 227 acres (5? ha) of which 152 acres (36 ha) are
serviced by combined sewers draining to a 60 In. (1.5 m)
sewer upstream of the Swirl site. The District boundary,
the location of the sewers, and the Swirl site are
Indicated on Figure 15.
The Stevens Avenue Pumping Station (see Figure 16) Is
also located on the swirl project site and provides a
portion of the total flow to the South Wastewater
Treatment Plant. The South Plant presently has a
capacity of 12 mgd (524 l/s), serves approximately 69,000
people and has significant Industrial Input. The older,
combined sewers service about half the residential area.
The City of Lancaster and the U.S. EPA elected to
utilize tne Stevens Avenue District as a demonstration
site because of Its combined sewers, history of wet
weather overflows, compact size and topographic
characteristics. In addition, the Stevens Avenue Pumping
Station was the only station owned by the City with
sufficient land available for the construction of the
proposed swirl project.
5.3 MonI tor Ing RBSUIts
The 250 acre (100 ha) demonstration area Is adjacent to
the Conestoga PIver In the southeastern portion of the
city, as shown In Figure 14. The developed portion of the
area, about 125 acres (50 ha) Is residential. The rema'nlng
area Is parklands or undeveloped fields.
The developed area has an estimated runoff coefficient
of 0.59 and would contribute 7,580 u»3 (2. Mgal) of runoff to
the swirl through the combined sewers, from a 1 IP. (2.54 cm)
rainfall. The undeveloped half of the drainage area Is
assumed to contribute runoff directly to the river.
In preparation for construction of the Swirl
Regulator/Concentrator facllty at 1he S+evens Avenue outfall
to Connestoga Creek, rainfall/runoff and water quality
monitoring from the 13' ac (54.2 ha) Stevens Avenue catchment
61
-------
ro
SOUTH MAMMf
wsrmcT
NORTH OftAINAM WSTWCT
ft • 0.3 1«
Figurel 4. Droinogt districts
MOfffM TM49WNT KANT
STEVENS AVENUE
OMAINACC MtTWCT
City of toncotttr, PfMNylvonio.
-------
tf
Figurtl 5 .Sltv*nt Ar^nut droino
-------
«
\
•: '.3
\
**
\ SOUTH BROAD STf
•i
• n, -
HIT
IVEftSION CHAMBER
ONTROL iUILOING
SWIRL CONCENTRATOR
•CALK IN PICT
1 ft • 0.3 In
Figure iS.Loncostcr Swirl Project site pton.
-------
ttas performed In 1973-74 by the City of Lancaster and
Meridian Engineering of Philadelphia. (14) Data reduction vas
performed by the University of Florida.(6)
Rainfall data were obtained using a weighing bucket
ralngage with five minute sampling Intervals. Depth
measurements were made In 60 In. (152 cm) RCP sewer by
Controlotron Corp 290-1 sonic water level sensor. Continuous
strip chart records at depth were converted to flow using
Manning equation, (n • 0.013, slcpe « 0.035).
Supercritical flov at the measuring point eliminated
backwater effects. The automatic depth gage recorded values
at 18 second Intervals and later averaged to provide 1.5
minute weighted averages. Computed flows were highly
variable and changed by over 100 cfs (2.8 mVsec) within
minutes owing to the steep "flashy" character of the
catchment. Six storms were monitored between September, 1973
through January, 1974.
65
-------
CHAPTER 6
Study Area Desertptlon-West Roxbury Swirl Facility
6.1 Foreword
The rationale used In selecting a site for the
demonstration SwIri/Helleal Bend regulator treatment complex Is
presented Fn section 6.2. Descriptions of the study area are
given In section 6.3. An analysis of long-term rainfall
characteristics within the Boston metropolitan area Is
presented In section 6.4. Utilization of the U.S. EPA
Storawater Management model (SMNM) to estimate a composite
runoff factor for the selected area Is presented In section 6.5.
The runoff factor was used in a further analysis described In
chapter 8 In which the design flow for the West Roxbury
facility was selected.
6.2 Study Area SalectIon - Background
Ir 1977 the Swirl Regulator/Concentrator was being
examined by U.S. EPA and others as a suitable compact device
for ramoving solids and partlculate material from combined
sewer overflows. At that time the swirl had yet to be tried
for removing pollutants from a separate storm sewered area.
Generation of operational experience using the Helical Bend
regulator was still In Its infancy. Full-scale field
demonstration of the Helical Bend regj'ator had yet to be
conducted. The project aim was 1 o determine the
effectiveness of a full-scale Installation -jslng both
devices, side by side, for removing storm sewer related
pollution emission loadings.
A design flow of approximately 15 cfs (.43 m/s) with
an even split Into the two regulators was arbitrally chosen
as the envisioned project level on a full-scale basis.
The underlying rationale for this choice was the economics of
limited research funding and the expected hydrologlc yield
from approximately 100 acres (43.4 ha) of urbanized land.
Runoff from a 100 acre parcel seemed to offer real world
potential. The project scale was Therefore tempered by both
the "reasonableness" of a real-world control situation and
dvoilable research money.
Another factor Influencing site selection was the
env'sloned mode of treating Intermittent stormwater
dIscharges.The hydraulic flow considerations for the regulators
at the envisioned site differed from past CSO experience In
Lancaster and Syracuse. First, dry weather sewbge base flow
66
-------
Is not present and the units would not be operating In the
CSO mode where treatment begins when downstream pipes cannot
hydrauI lea I Iy handle discharge and overflow occurs. In this
situation, these devices simply treat all Intermittent vet
weather flows. The foul sewer connection of both devices to
the receiving sanitary sewer could discharge considerable
stormwatar runoff for Iow-Intenslty/short duration storms.
This Implied that the selection of the design flow range where
reasonable treatment efficiencies could be expected for most
storm events was crucial since the foul sewer discharge Is
taken to be 3-5} of design flow.
Another factor Influencing site selection was the notion of
choosing design flow on the basis of "pollution control"
concepts. It seemed at the time that the selection of the
particular design discharge should emphasize probabilistic
pollution-reduction concepts versus the more conventional
notion of size selection on the basis of some reasonable
extreme hydro!oglc/hydrauI Ic event. The cost of mosl wet
weal her control facilities designed on the basis of high
flow extreme event hydrology Is most often much higher
than the ensuing costs of control programs aimed at
probabilistic pollutant reduction levels. This argument had
significant historical prescedent at the time since the
general federal viewport on the control of stormwater
related pollutant load!- s was to encourage municipalities
to reduce storm sewer emissions by and large* through non-
structural "public works" oriented practices. The funding
reality at that time (and still is) was that there was no
federal funding mechanism to deal with stormwater related
pollution. Limited solids reductions in storm sewered
loadings, on the order of 30-50$ were common section 208
planning recommendations.
The results of a rudimentary simulation analysis
Indicated that a residential area (It was decided to
concentrate on a residential area) of approximately 100
acres (43.4 ha) would be required to meet the criteria
previously discussed. Urbtn sectors In the greater Boston
area tend to be densely populated making the location of a
physical site difficult. A site of roughly 100 ft x 100 ft
(32.8 m x 38.8 m) would be required In an open area
(Inexpensive construction) somewhere at the end of a separate
storm-sewered residential catchment area. To further complicate
matters (truly the most difficult aspect) was the requirement
for gravity operation especially In the case of the swirl.
Since the receptor of the swirl foul sewer affluent would
be a sanitary sewer , an elevation difference of roughly 3 to
5 ft (.91 to 1.5 .n) below the storm drain (Invert to Invert)
would be required. The sewer II re would also have to have
67
-------
sufficient capacity end fluid shear stress to handle and
transport solids from the foul sewer effluent. Adequate head
would also be required for gravity operation of the clear water
overflow.
After reviewing roughly 15 areas In five communities
for over a year, EOF finally came upon a site In West
Roxbury, Boston that was reasonable and met the aforementioned
criteria for the demonstration project.
6.3 Study Area DescrIptIon
Figure 17 shows a nap of Greater Boston and the
approximate location of the facility. The proposed catchment
area lies In Most Roxbury, Just north of the Dedham line
approximately eight miles southwest of downtown Boston. West
Roxbury Is a section of Boston, and the sewers In
question are therefore totally under the Jurisdiction of
the Boston Water & Sewer Commission (BW&SC).
The land use of the West Roxbury (Boston) study area 1s
depicted In Figure 18 and may be typified as moderate Income
residential, with mixed commercial occurring primarily along
several main road tracts. General housing density In the area
Is moderate, with lawns and streets In generally good
condition. The sewer system In this area Is entirely separated
(as constructed), with the exception of a few streets. The
approximate median age of the housing In the area Is 40 years.
Drainage from the area discharges primarily Into the Upper
Charles River through an extensive drainage system. Topography
In the area varies from moderately hilly to flat.
6.4 Analysls of Long-Tarm RaInfalI Records
Magnetic tapes containing unedited precipitation records
were obtained from the National Climatic Center, North Carolina.
One tape contained 10 years of hourly rainfall records
(1964-1973) for the following gaging locations: Logan Airport
and Blue Hills Observatory (see Figure 17). The tape required
careful editing to detect missing gaps (less than .8} of the
total record). Most often, cumulative estimated end-of-the
storm totals were given for missing sequences of hourly
records. A temporar/ data file on disc storage containing
edited data, was generated from the magnetic tapes for each
station. The data files for the two primary stations contained
approximately 1000 wet days each, spanning over the 10 yecr
period.
68
-------
Figure 17. Locotionol mop of West Roxbury Swirl sitt.
69
-------
CATCHMtKT 500NOAR.Y
JMUe* *e»l6R
«.4ANHOL£ *
3W;*L/MCL»CAL MMP 8»ft LOCATIOM
QUARRVlNOrr MOT OF OWCMMCKTO
MtDKJJA DftMtltY
Figure18, Land use map- West Roxbury demonstration project
-------
The next phase of the computer analysis Involved defining
discrete events from the continuous odltod time series of
hourly precipitation records. A sequence of antecedent dry
periods and storm events were constructed In the following
manner. A storm was defined when the cumulative rainfall
exceeded 0.05 In. (1.3 mm) of rain. During the first three
hours of rainfall, If the hourly precipitation value was less
than 0.03 In (0.8 mm) the precipitation was Included In the
total rainfall o* the event but the periods were not Included
In the duration of the storm event. A separation In rainfall
of a minimum of six hours defined a new storm. It was assumed
that storms characterized by a total rainfall of less than
0.05 In (1.3 am) Just wet the system and would cause little. If
any, runoff. These periods were considered as part of the
antecedent conditions. The same assumption was made when the
first three hours o< rainfall were less than 0.03 In. (.8 mm).
For every storm event, the following Information was
computed: a) the year, month and day the storm began; b) the
hour of day the storm began; c) the length of the dry
period preceding the storm event; d) the total amotnt of
rainfall In the storir even?; e) the total duration of the
storm event; f) the maximum hourly precipitation In the storm
vent; and g) the hour of m ax I..1 urn rain within the storm event.
Another set of data files were prepared containing this
Informational set for the two stations. Statistical
computations were then performed on the following six storm
event parameters: a) dry days; b) total rainfall of the storm
event (In.); c) duration of storm event (hours); d) average
Intensity of the storm event (In./hour); e) mtxlmum hourly
rain In the storm (In./hr); and f) hour of maximum rain ,n the
storm.
Overt!I statistics are given In Table 5 for
antecedent dry days between storm events, total rainfall (In.),
duration (hours), average Intensity (In./hr), max. hourly
intensity (In./hr) and hour when maximum rainfall occurred.
Seasonal statistics were also computed but are not presented.
71
-------
TABLE 5
Comparative Analysis of Rainfall Statistics
Logan Airport vs Blue HIM (B.H.) Observatory
(1964 - 1973)
* ** *«« +
Notes;
A - Antecedent dry period - days
B - Total Rainfall per event - In.
C - Duration of storm (hr)
0 - Average storm Intensity - In./hr.
E - Maximum hourly Intensity In./hr.
F •- Hour of maximum rainfall Intensity
No. events? Blue Hills - 786 storms
Logan Airport - 747 storms
tt
A
B
C
0
E
F
*
*«
«*«
t
tt
Logan
B.H.
Logan
B.H.
Logan
B.H.
Logan
B.H.
Logan
B.H.
Logan
B.H.
Maan
Standard
Coeff Icl
Ml n Imum
Maximum
4.5
4.3
0.53
0.59
8.9
9.1
0.07
0.08
0.15
0.15
4.0
4.1
devlat
ent of
4.1
3.8
0.59
0.71
8.1
8.9
0.07
0.08
0.14
0.14
4.5
4.7
1 on
Variation
0.9
0.9
1.1
1.2
0.9
0.9
1.0
1.0
0.9
0.9
1.1
1.1
1 In.
0.25
0.25
0.05
0.05
1.0
1.0
0.008
0.008
0.01
O.C1
1 .0
1.0
0 2.54 en
23.1
26.8
5.1
7.7
48.0
76.0
0.87
1 .30
1 .56
1 52
39.0
39.0
72
-------
6.5 Hydro logle Mode I Ing of Catchment Area
The U.S. EPA Storm water Management Model (SWMM) was
used to calibrate several measured "snap-shot" rainfall/runoff
events In the study area. The purpose of this effort was to
derive a reasonable overall estimate of the watershed's
composite runoff coefficient which was later used In the
long term (1C year) simulation analysis described In Chapter
7.
BASIN SCHEMATIZATION;
The deaonstratlon site catchment area shown In Figure 18 was
Initially Identified by overlaying a topographic map onto a
map of the existing storm sewer system (from Boston Wate** &
Sewer Commission) within Most Roxbury. The basin outer
boundary was first approximated as to gird the sewer system
and then by observation of topographic divides all land area
draining to the storm sewer system was Included. As a check
on the accuracy of the contour-derIvea approximation, an
Inspection during a storm event was conducted of the storm
sewer lines near the periphery of the basin to ascertain the
direction of flow at predicted outer boundary locations. Total
catchment area was p I art Imetered to be 207.2 acres (84 ha) with
46.5 acres (18.8 ha) draining Into the quarry catchbasln. Since
the quarry Intermittently pumps water (usually only during
business hours) the remaining 160.7 acres (65.2 ha) alone were
considered to comprise the study's catchment area that Imparts
a runoff loading directly during a storm event. Roughly 66$ of
the wet weather contrlbutary catchment area contains
residential housing. The balance Is Idle or vacant open land.
Other characteristics of the urbanized portion of the
catchment area are Indicated below:
e street surfaced - 10.7 miles (17.1 km)(93? asphalt/7*
concrete)
• street density a 0.10 lane-miles/acre (.06 km/ha)
• population density ° 15 persons/acre (37.5
persons/ha)
e overall slope of area ° 290 ft/ml 0.053
• portion of streets with swale ditches » 2%
73
-------
major sediment sources
In catchment areas
Quarry (pumping after
rainstorm) (8:00 am -
5i00 pm)
No. catchbaslns:
Catchbasln density:
Average catchbasin
dimensions: a)
b)
c)
d)
85
1.25 catchbasln/acre
(.51 basln/ha)
SUMP volume •
total volume'
diameter •
depth »
65ft3 (1.8 m3)
97ft3 (2.7 m3)
4.25 ft (1.3 m)
7 ft (2.1 m)
RUNOFF BLOCK DATA PREPARATION
In vie* of the relatively small size of the basin, and
the homogeneity of basin land uses (essentially only single-
family and open park land)* a rough discretization grid
composed of six subcatchments was selected as adequate for
modeling purposes, (see Figure 19) With number of subcatchments
chosen. Individual subcatchment boundaries were delineated by
the following criteria: sewer trunk line location; Intra-
basln topographical divides; homogeneity of land uses and ground
slope characteristics; and subcatchment compactness.
Subcatchments I, II, III and V have a fairly uniform
single-family residential land use pattern. Subcatchment IV
Is roughly half single-family residential and half open land.
Subcatchment VI Is entirely undeveloped land characterized
by a steep gradient with a ground surface of shelf-rock.
Subcatchment VII, draining Into the West Roxbury Quarry, was
not modeled, for quarry water Is pumped only Intermittently
to the storm sewer system and was not pumped during the
storms monitored for modeling.
Readily determined SNMM Runoff Block data for each
subcatchment were as follows: area slope, characteristic
width, land use % ImpervIousness. Ground slopes and
characteristic widths were approximated by procedures
described elsewhere. (30) Other data required for entry Into the
Runoff Block such as overland flow resistance factors,
relentlon depths, Infiltration rates, and Infiltration rate
decay coefficients were not available. These factors were
assumed to vary from the default values pre-programmed Into
the model and were used as secondary calibration parameters.
74
-------
-------
STORM SEWER MODELING £ TRANSPORT BLOCK DATA PREPARATION
Of roughly 130 In-stream saver elements (manholes and
conduits), 41 main-stem elements vere selected for Transport
Block simulation. After numbering of modeled elements In
network conflguraton sewer maps were used to determine diameter,
slope and length. Sewer slopes were calculated as
differential manhole Invert levels over length of connecting
conduit. Manning roughness coefficients of roughness were
assumed to be In the range of 0.014 to 0.015 except for an
open brook section where a value of 0.035 was used.
CALIBRATION & VERIFI CAT I ON
For design objectives, SWMM can be an accurate
productive tool for storm-event simulation whent extensive data
defining basln-speclfIc parameters are available) extensive
testing Is done to determine site specific substitutions for
general default values offered In SWMM; rain gauging Is
conducted at or near the controld of the drainage basin (and
preferably at other locations In the basin as well); and, the
full capacity of the Runoff (200 subcatchments) and Transport
(159 sewer elements) Blocks are used.
Flow calibration was accomplished by adjusting
calibration parameters to match both the sum of predicted and
measured total volumes and the sum of predicted and measured
flow peaks. The former criteria Is achieved by adjustment of
subcatchment percent ImpervIousness values the latter criterion
by either adjusting number of sewer elements modeled In
Transport, changing Manning roughness coefficient values for
Transport Block elements, or by adding/subtracting aii
artificial conduit Inducing further routing delay and thus
attenuating /Increasing peaks.
STORM EVENT MONITOR IMG
A ralngage had been placed on top of the Spring
St. West Roxbury Post Office, outside the outer boundary of
the basin, and at a distance of roughly I ml (.6 km) from The
centrold of the drainage basin. Rainfall Intensities, for
entry In 15 minute hyetograph time-steps In the Runoff Block,
were read directly. A Manning liquid-level automatic recorder
was placed at the end of the catchment area on New Haven Street
to measure storm event depth of flow for a three month period.
Time of travel tests were performed at varying depths of
flow to enable calibration of a stage-discharge rating curve
(see Figure 20). This curve was determined via a least-squares
fitting procedure (31).
76
-------
30 -
25 -
20 -
5
0
15
10 -
5 -
O CALIBRATION POINTS ( 1978)
O CALIBRATION POINTS (1979) ^
~ * BEST FIT SLOPE EQUATION
*f
*
^f
J—
<
s
A
T
/
1
J—
I
/
^
/
/
i
.
I
/
/
/
/
/
/
y
¥
¥\
T
f
i
!T
V
r i T i
5 10 15 20 25
1 in = 2. 54 cm DEPTH (in.) ^ cfs a 28>3 1/s
Figure 20. Stage/ rating curvt,30in. RCP droin.Ntw H
-------
It Is desirable to embody a number of storm events
for calibration supplemented with a smaller Independent storm
event set for verification, both storm sets being
representative sf typical variations In intensity ard
deviation. Unfortunately the three month mon'torlng period was
dry and only two storms of reasonable volume were monitored.
Table 6 presents a characterization of those storm events.
TABLE 6. MEASURED STORM EVENTS
Date
7/4/78
Storm/
1
Duration
Length of
(hrs.)
17
Maximum
Intensity
(In./hr)
0.14
Total
Ralnfal
(In.)
1.08
Comments
1
Most
rel (able
data set
3/16/78
0.33
(I In. - 2.54 cm)
0.5 Less reliable
data set
difficulties
encountered
Interpretlng
ralngage data
CALI BRAT I ON/VER!F »CAT I ON
METHODOLOGY £ RESULTS
The primary objective of applying the SNMM Model
was to derive a reasonable overall estimate of the basin's
composite runoff coefficient, from which swirl regulator design
flow frequencies can be estimated. Unfortunately, the paucity
of storm event data for modeling necessitated a modification
of the more desirable caI I oratton/verIfIcatIon appronch
cited above. Storm event 01 was selected as a primary
calibration storm and storm 92 as a secondary calibration
storm.
Three calibration runs were made. For
parameters were estimated based on
schematlzatlon of the basin. Adjustments for
were derived fro* volume and peak comparisons
the first run,
the Initial
the second run
of Run 1 whore
only the primary calibration
calibration parameters for Run
(.dak comparisons of Run 2
storm results.
storm was used. Alteration of
5 were based on volume nnd
reflecting both calIbratron
A tabulation of parameter changes for the three
calibration runs for the two storm events ere given In Table 7.
78
-------
TABLE 7. SWMM MODEL CALIBRATION MODEL DATA
»•«•»••»••••••••••• WMVM^W^tB •••••••»•«••«••••••••«•••
CALIBRATION RUN
• •••••••«••••»•»•••••••••
1 2
STORMS & PARAMETER VALUES
Event 1 (7/4/78)
Primary
CalIbratlon
Event 2 (5/16/78)
Secondary
CalIbratlon
COMPOSITE
RUNOFF
COEFFICIENT
SUBCATCHMENT
IMPERVIOUSNESS
DETENTION ( in.)
DEPTH
MINIMUM
INFILTRATION RATE
(In./hr)
DECAY RATE
COEFFICIENT OF
INFILTRATION
0.376 (datuum)
1 45
2 45
3 43
4 27
f 45
6 20
0.03
0.30
0.005
0.396 (+5$)
47
47
45
29
47
25
0.03»
0.10
0.005
0.468 (+1
55
55
52
40
55
34
0
0
0.00
8*)
.03
.52
115
* Detention depths were varied In Run §2 ror more
pervious subcatchments, 14 and 16 from 0.03 In. to 0.02 In.
(I In. » 2.54 cm)
79
-------
Figure 21 a A An Initial run of storm 1 Incorporated a
composite runoff coefficient of 0.376) the resultant
modeled flows ^ (as Illustrated In Figure 21) closely matched
measured flows, particularly during the first two-
thirds of the storm event. Near the end of the storm,
however, modeled flows tended to be lower than recorded
flow. While most of the. peaks. were reasonably well matched,
the under estimation of total volume warranted an Increase In
the composite runoff coefficient for Run £2.
Both calibration storms were run. The runoff
coeffclent was Increased by 5% over Run II to 0.396 and
the minimum Infiltration rate decreased by 61%. Both changes
were made to Increase total predicted runoff volume. For both
storms the peak flows were accurately modeled. Total
modeled runoff still underestimated measured total runoff.
film £Ji In Run 13, the composite runoff coefficient was
Increased to 0.468, and Infiltration calibration parameters
returned to the default value recommended In the SHMM User's
Manual. In both Runs 2 and 3, the percent I mperv I ousness
attributed to subcatchment VI was Increased at a greater rate
than those of the other subcatchments In order to more
adequately simulate pervious runoff from the sloping shelf-rock
type terrain. For storms 1 and 2, total volume were
slightly overestimated while peak flows were over-estimated
by 19*.
On the basis of storms 1 and 2 an extrapolation
procedure was developed to optimize total value and total peak
equalization of measured and predicted values. Calibration
volume and peak flow summations for Run 7. and Run 3 were then
compared and Juxtaposed with corresponding composite runoff
coefficients. (See Table 9). By linear extrapolation the
composite runoff coefficients yielding unity ratios for total
volume/total peak flows are 0.4466/0.3921 respectively.
Consequently a mean value of 0.419 for the composite runoff
coefficient was selected for Incorporation In design flow
f ormul atlons.
80
-------
TABLES.- SUNK CALIBRATION RESULTS
CALIBRATION
STORMS
STORM
DATE
(1) 7/4/78
1(2) 5/17/78
SUMMATION OF
CALIBRATION VOLUMES
AND PEAKS
COMPARISON TO MEASURED
MEASURED .
TOTAL FLOW (ftj)
95.616
48.303
11 • 95,616
11*2-143,919
MEASURED
PEAK
(CFS)
2 90
2.40
2.76
3.76
0.70
2.25
9.0
11 • 12.52
11*2-23.77
RUN 1
Predicted
Total Flow
(«3)
78,513
— —
"
M/P-1 .21 78
Predicted
Peak Sum
(cfs)
3.26
2.4i
2.78
2.48
0.46
m m
"
1 - 11.4
M/P-1 .093
RUN 2
Predicted
Total flow
fft5)
81,775.8
41.414.89
1-123,190.7^
M/P-1 .1683
Predicted
Peak Sun
(cfs)
3.40
2.5:
2.90
2.64
0.46
3.'0
9.0
1 1*2-23. 93
M/P-0.993
RUN 3
Predlctud
(Total flow
(ft5)
106.481
48.487
C • 154.968
H/P-0.929
Predicted
Peak Sum
(Cfl)
3.97
2.94
3.40
3.09
0.53
3.35
10.20
C • 27.49
N/P-.865
1 ft* - 0.028 cubic meter
1 cfs - 28.3 1/s
-------
<0
"o
CO
Figure 21. Storm 1=7/4/78
Hydrographs
Hydrograph Key
Measured
Runt Predicted
Run 2 Predicted
Run 3 Predicted
7/4/78
7/5/78'
1 cfs - 28.3 1/s
TIME I hour I
-------
10r
CD
CO
O
Figure 22 .Storm 2*5/15/78
Hydrographs
Hydrograph Kay
Measured
Run 2 Predicted
Run 3 Predict* d
4:
1 cfs - 28.3 1/s
(pm)
5:OO
TIME (hour)
5/17/78
-------
TABLE 9. EQUALIZATION OPTIMIZATION
RUN 2 RUN 3
Composite Runoff 0.396 0.468
Coefficient
Sun of 0.993 0.865
Peak Flows
SUB of Total Voluaes
(Using Parenthetic 1,168 0.929
Values reflecting as
veil residual flows)
84
-------
CHAPTER 7
DESIGN PHASE - LANCASTER SWIRL PROJECT
7.1 FOREWORD
General design conditions and site characteristics are
provided In sections 7.2 and 7.3, respectively.
Rainfall/runoff considerations and selection of design flow
for sizing the facility are given In section 7.4. Final
Swirl Concantrator design details are given In section 7.5.
Swirl Degr'tter design Information Is provided In section
7.6. Other facility features Inc'udlng chlorlnatlon and
Dlscostraln«r are cited In section 7.7. Final ly, description
of the Hydro-Brake used to control the foul sewer flow from
the Swirl Concentrator to the Swirl Degrltter during wet
weather Is given In section 7.8.
7.2 DESIGN CONSIDERATIONS AND CONSTRAINTS
There were many design aspects of the Lancaster Swirl
project which required careful consideration and Included
the hydraulic sizing of the units as well as the
sampling capabilities of the system In order to
correctly evaluate the solids removal efficiency of
the Installation. Also, the general layout and design of
the control building housing the Swirl Degrltter and
various electrical controls was an Important
consideration.
The constraints of the project design were minimal. In
the Stevens Avenue drainage district, there was sufficient
hydraulic head to drive the Swirl Regulator/Concentrator. If a
Swirl Is Installed at the same elevation as an existing
sewer, then either the Influent line must be surcharged
or the foul underflow be pumped. At the Stevens Avenue
Swirl Project, the 60 In. (1.3 m) Influent sewer Is
constructed on a steep slope mandated by the topography near
the Concstoga River. This surcharging, or flooding, of
the Influent sewer, d'
-------
The Swirl facility site Is approximately one acre
(.4 ha) and consists of a gently sloping floodplaln area. The
site topography ranges from 268 to 294 ft (81.7 to 74.4m)
above sea level at an average slope of 4$. So!is are
characterized as being a sandy clay alluvium, underlain with
a shale-limestone at depths ranging from 14 to 19 ft (4.3
to 5.8 m). These soils are characteristic of the floodplaln
area In which the site Is located. It Is estimated that
the 100-year frequency flood would cover the site to elevation
260 ft. (79.2 m). Therefore, the control building was
constructed with a first floor elevation equal to the 100-
year flood elevation to assure proper operation during major
flooding stages of the Conestoga River.
7.4. HYDRAULIC DESIGN
The Initial design phase for the Lancaster Swirl
Project concentrated on the determination of the design storm
flow, with consideration given to provide a unit which would
operate at or above design capacity at least several times
per year. One major purpose of this philosophy was to try to
evaluate the swirl performance during flows greater than
the design flow. For this reason and for economics, the
swirl was sized to handle a design flow which would
occur relatively frequently (5 to 10 times per year).
To determine this peak flow, the Rational Formula was
used. This formula depends on the assumption of a steady,
uniform rainfall Intensity which will cause a maximum flow
rate at the time of concentration, or the time of flow
from the most distant corner of the drainage basin to the
point of discharge. This maximum flow Is found using the
rational formula,
Q - C I A
where Q • peak runoff rate, cubic feet per second (cfs)
C • dimension I ess runoff coefficient, expressing
the ratio of rate of runoff to the rate of
ralnfalI
I » average Intensity of rainfall lasting for the
time of concentration '. In/hr)
A » drainage basin area (acres)
86
-------
The runoff coefficient1 1C) Is estimated according to
the soil surface available, the amount of development In
the basin and the topography. For the Stovens Avenue
drainage area, the runoff coefficient was estimated for
each minor drainage section and a composite value of
0.55 was calculated.
Hourly rainfall data frTSm the Pennsylvania State
University weather station at Landlsvllle for the three-year
period from January 1973 through December 1975 were used
as base data for the design flow determination. During the
three year base data period there were 147 rainfall
events which would have caused an overflow to the
Conestoga River at the Swirl project site. All^ralnfall
Intensities less than 0.1 In./hr. (2.5 mm/h) which produced
a runoff rate of less than 8 cfs (226 l/s) were assumed
to be controlled without bypassing the Stevens Avenue
Pumping Station. Flow rates which were assumed to have
caused an overflow Into the swirl site were correlated for
the 147 rainfall events and plotted on the probability
graph shown In Figure 23. Of these 147 overflow events, 20
events or 6.67 events per year, produced a runoff rate
In excess of 40 cfs (1,131 l/s). A runoff rate of 40 cfs
(1,131 l/s) was selected as the design flow for the swirl
concentrator. From the probability analysis It Is Indicated
that 127 of the 147 events or 90 percent of the rainfall
events, would produce an overflow rate less than or equal
to the design flow while 10 percent would be greater than
the design flow. This will, on an annual basts, allow the
Swirl Concentrator to process approximately 50 storm events per
year of which 10 percent would be In excess of the design
flow. The Swirl Concentrator unit Is designed to
hydraulleally accomodate flows to 125 cfs (3,534 l/s).
After determination of a design storm flow, the next step
In the design was sizing the Swirl unit. U.S. EPA's design
publication (4) was followed. Initially, the Inlet pipe diameter
(Dj) Is set which. In this case, was 3 ft (0.91 m). This size
was selected for maintaining a scour velocity for the unit.
Assuming a desired solids recovery efficiency of 90% and a
flow of 40 cfs (1,131 l/s), Figure 24 was used to determine
the chamber diameter (D2> of 24 ft (7-3 m>« Following this.
Figure 25 was used to check the estimated 90> settleable
solids removal efficiency.
The remainder of the dependent dimensions are shown
with the necessary design equations on Figure 26. However, It
should be noted that It was necessary to alter some of
the radii ( R-j R-j and R4 ) on the gutter lines so that a
smooth curve 'resulted from scale-up of the unit. The axis
87
-------
:•
U£f '••*
jftrc
I
-------
>&**&&&&$
£g»
-------
.
.
-------
for Rj was offset to fit as shown In Figure 26. One omission
from this figure In the original publication Is the equation
for determining the dependent dimension, b£. After consultation
with LaSalle Hydraulic Laboratories, this equation was
determined to be b2 - Dj/6. Table 10 shows the dependent
dimension equations and the final values used for the
Lancaster Swirl Project:
TABLE 10
ACTUAL LANCASTER SWIRL DIMENSIONS*
ft, --3 ft (0.91 m) b, - 1/18 (D2) - 1.3 (0.41 m)
°2 = 24 ft <3 m> b2 = 1/6 (D.,) = 0.5 ft (0.15 m)
H, - 1/4 (D2) = 6 ft (1.8 m) R] , 7/18 (D^J^Q ft (3.05 m)
D3 " 2/3 (P2)- 16 ft (4.9 m) Rg = 1/4 (D^ s 6 ft (1.8 m)
D4 • 5/9 (D2)= 13.3 ft (4.1 m) R3 » 5/48 (D2)= 2.67 ft (0.8 m)
hl - 1/2 (Dj) - 1.5 ft (0.46 m) R4 3 3/16 (Dg)- 4 ft (1.2 m)
h2 = 1/3 (Dj) =1 ft (0.305 m) Rg = 11/18 (D2)»14.67 ft (4.5 m)
* These are actual dimensions and may slightly differ from
the values determined from the aquations.
Additional dimensions of the fI eatables trap, accompanying
spoilers and weir plate were finalized following discussions
with the U.S.EPA and LaSalle Laboratories. Following this basic
hydraulic sizing of the swirl unit more In-depth design of
the unit and the entire flow scheme was completed.
7.5. FINAL SWIRL CONCENTRATOR DESIGN CONSIDERATIONS
The final design Included the preparation of plans and
specifications for all unit processes comprising the
Lancaster Svlrl Project. These unit processes consist of
the Swirl Regulator/Concentrator, the Swirl Degrltter, the
diversion box and the 01scostralner. Final design of the
swirl unit Included a pressurized flushing system for the
purpose of washing down the unit following a storm event.
The previous hydraulic studies demonstrated that as the storm
flow receded, some residue would be left on the walls and floor
of the Swirl because of diminished flow velocity. A flushing
system was Installed around the Swirl walls and under the
overflow weir plate to Insure proper cleanup following an
91
-------
event, washwater generated was discharged to the Stevens
Avenue Pumping Station for conveyance to the City's South
Treatment Plant.
There «ere a number of considerations for the
final design of the total flow scheme, as shown on Figure
3, Chapter 1. The storm f I ov enters the diversion box
through a 60 In. (1.5 m) diameter sewer. The flow passes
through a manually-cleaned bar screen with 2-1/2 In. (6.4 cm)
clear space between bars and Into a 36 In. (0.91 m) diameter
line through the control building and Into the Swirl unit.
At a storm flow of 129 cfs (3,534 l/s), the Swirl
Concentrator Is driven at Its maximum capacity and the water
surface elevation In the diversion box Is 6 In. (15.2 cm) from
the top. At this point a level sensor located In the
diversion box actuates a sluice gate connected to the 60
In. (1.5 m) line to divert the flow In excess of 125 cfs
(3,534 l/s) directly to the Conestoga River. Flow from the
central and enorgency overflows of the Swirl unit Join
this 60 In. (1.5 m) line Into the river. The foul outlet
flow Is directed through a 12 In. (0.3 m) line to the swirl
degrltter which separates the grit and feeds the degrltted
effluent to the sanitary pump station.
The emergency overflow weir was Included since the 40
cfs (1,131 l/s) design flow would be exceeded several times per
year. This overflow can prevent severe overloading of the
swirl by diverting the excess flow to the river. The
emergency overflow was designed as a "window" In the Swirl
walls so that flow would pass through the window and Into
a sloping trough. A weir plate In front of the window
allows adjustment to vary the amount of head over the
central overflow before any flow Is directed to the
emergency overflow. The range of Swirl Concentrator flows
before the emergency weir flow Is Initiated can oe
selected from 53 cfs (1555 l/s) to 85 cfs (2403 l/s) The window
weir consists of one quarter of the Swirl unit's perimeter.
During dry weather, domestic sewage Is contained In the
."wlrl's primary gutter and Is diverted directly to the
Stevent Avenue Pumping Station without any flow going to the
Swirl Degrltter. During a storm event the rising water level
In the Swirl unit Ir detected by a level sensor which
closes the by-pass valve and opens the line to the Swirl
DegrItter.
Wet weather flow to the Swirl Degrltter was initially
controlled by a 12 In. (30 cm) pinch valve. The valve constantly
malfunctioned and was then replaced with a Hydro-Brake. The
92
-------
Hydro-Brake Is a non-clogging flow restricting device with no
moving parts (see sac-Men 7.8).
7.6 SKIRL OE6RITTER DESIGN DETAILS
It was decided during the design phase to size the
Swirl Degrltter to handle the dry weather flow If the City
of Lancaster chose to do so. Therefore, the design flow for
the Degrltter was established as 3 cfs (84.8 l/s) from dry
weather flow records. However, the flow Is presently
controlled at 1.2 cfs (3.4 l/s). The Swirl Degrltter separates
the grit which Is then conveyed by a screw-feed/conveyor belt
combination to a dumpster located on a scale to measure the
amount of grit collected. The degrltter effluent Is cycled
back to the Stevens Avenue Pumping Station for conveyance to
the treatment plant. The Swirl Degrltter design was performed
following U.S. EPA design procedure (5).
The basic design parameters for the Swirl Degrltter are
an Inlet, D*,, of 1 ft (3 m) design flow of 3 cfs (84.8 l/s)
which represents the peak dry weather flow and a 901
settleable solids removal efficiency. (5) Figure 27 gives tht
general Jeslgn dimensions and equations of the Swirl
Degrltter. From Figure 28 a chamber diameter, 0*2 °* 8 ** (2«4 •)
was specified (5). It should be noted that the D*3 equation
on Figure 27 has been revised to be D*3 ° 2/3 (D*2) , This
correction was made by LaSalle following the review of their
design procedures. However, this revised D*3 equation was not
available until the design was complete. The general dimensions
of the swirl degrltter are shown In Table 11.
IAfi!L£ 11
ACTUAL SWIRL DE6RITTER DIMENSIONS
D^-1 ft (0.305 m) H^-20,* "2 ft (0.61 m)
D2*-8 ft (7.4 m) H2* - 1/4 D^ ' 0.25 ft (0.08 m)
03*-4 D,*-* ft (1.2 m) H3*» D^'-t ft (0.31 m)
(old equation-actual dimension)
03* • 2/3 CL* » 5.3 ft (1.6 si) (new equation - proper
dimension)
7.7 OTHER FACILITY DESIGN FACTORS
Pre and post chlorlnatlon capabilities were required.
Chlorine lines were Installed to tho diversion chamber and to
the central swirl overflow. Flow proportional chlorine feed
93
-------
t-- *Jfci ••"•
\", >" T* *• '':•'• *
^B
•TSJSr-
.••~ • :* r
-------
•
• V -r&* .-*r «k.
I
-
53S1
•
•
!
^;;.W.
.
.vigf ri>
*" !'•'
-
*&~
I
-
.
***!£
• TOiTW«*3'KPC':;:.»*;-» '^-^saiS: 1
«
• m
*
^ ; __Ji _
* .
ii ^_: I life
-------
facilities were provided to permit chlorlnatlon at a dosage
proportions! to flow rate. A Dlscostralner unit was also
provldad to treat up to 150 gpn (573 I/IB) of the Swirl
Concentrator foul outlet flow.
The 26ft x 32 ft (7.9 n x 9.7 to) control building
for the project was also a major consideration due to the
required area for the Swirl Degrltter, the Dlscostralner,
various sampling equipment, electrical controls and laboratory
facll(ties.
7.8 HYDRO-BRAKE
A Hydro-Brake Is a patented flow controller made of
stainless steel. It Is self-regulating and has no moving parts.
It requires no power, but uses the static head of
stored water to operate Its own "energy" to retard the
flow. The movement of water through a Hydro-Brake Involves a
swirl action, dissipating energy to control the rate of
discharge. Although the function of a Hydro-Brake Is
somewhat similar to an orifice. It has certain Important
advantages:
1. It permits a much larger opening for passage of
the same amount of water. This Is particularly
Important where clogging Is a possibility, such
as, for Instance, In catch basins. It is also
Important where sanitary or combined sewage flows are
being regulated.
2. The flow rate of a Hydro-Brake Is not significantly
affected by a variation In head. This Is Important
where It Is desirable to maintain a relatively
large passage for the water, yet also maintain a
fixed maximum rate of flow during peak
conditions.
3. The outflow from a Hydro-Brake does not create a
high velocity Jet stream as an orlflca wl'i, thus
avoiding scouring Inside sewer pipe.
The Hydro-Brake was Invented In Denmark about 15 years
ago and Is marketed In North America by Hydro S+orm Sewage
Corporation of New York. The Hydro-Brake must be designed for
the specific application and flow condition. The design Is
patented and units are available only through the Hydro Storm
Sewage Corporation.
96
-------
The unit was designed to pass a asxisua flea of 1.2 cfs (34
l/s). Tne unit vas Installed on the foul sever line from the
Swirl Concentrator.
97
-------
CHAPTER 8
DESIGN - WEST ROXBURY FACILITY
8.1 Foreword
Site description Is presented In section 8.2.
Ralnfa I I/runoff considerations and selection of design flow
conditions are given In section 8.3. Dimensions of the Swirl
Regulator/Concentrator and Helical Bend Regulator are given In
section 8.4. Miscellaneous design details are given In section
8.5.
8.2 SI ta DaserIptIon
A general discussion of the drainage area tributary to
the demonstration facility was presented In Chapter 6. (see
Figures 18 and 19) A plot plan of the facility was shown In
Figure 4, Chapter 1. The 160 acre (64 ha) moderate Income
residential neighborhood Is served by a completely separated
sewer system, discharging through a 30 In (76.2 cm) drain to a
120 In. (305 cm) conduit connecting to the upper Charles River
approximately 1200 ft (364 m) downstream. The 30 In. (76.2 cm)
RCP drain flows southward along New Haven Street. Just prior to
turning 90 degrees to connect to the Intercepting drain, the 30
In. (76.2 cm) drain becomes 87 In. »221 cm) for about 210 ft (64
m) In length. The area behind the homes on New Haven Street Is an
open area. This open area Is located between a line of trees
separating the residential space and a factory building located
on the opposite side of the InterceptInfg conduit. The actual
site within This open area Is located on a former railroad bed.
The railroad tracks and ties have long since been removed
minimizing construction problems. The 87 In. storm drain to be
used Intercepts a 120-In. (305 cm) drain culvert Just downstream
of the site. The drain culvert Is directly tributary to the upper
Charles River.
8.3 Rain* al I/Runoff-Das I g-n Flow Considerations
Thd original intent of the evaluation program was to
monitor 15 storms over a 6 month evaluation period. A recurrence
Interval of 12 to 13 days (2 weeks) was chosen as a design flow
consideration. In addition, the average maximum hourly flow was
arbitrarily chosen as the statistic to consider In the recurrence
Interval considerations.
A design flow meeting the aforementioned criteria of 12
cfs (340 l/s) was computed from a rainfall/runoff analysis using
the Rational Formula (see sertlon 7.4). The analysis war.
performed using a "C" factor of 0.41 (see Chapter 6), a catchment
98
-------
area of 160 acres (64 ha) and a time series of maximum average
hourly Intensities for each of 361 rainfall events measured at'
the Blue Hills Observatory covering the spring-fell portions of
a 10 year period. The resulting rainfall and runoff distributions
are shown In Figure 29. Design flow for each unit was split
evenly at 6 cfs (170 l/s). Maximum flow to each unit was
established at 12 cfs (680 l/s).
8.4 SnIrI/Hal lea I Band DasIgn DImanstons
The Swirl Concentrator design formulation derived by APWA
was used to size the facility. (4) A settleable solids efficiency
of 60% was chosen to deliberately underslze or "over-drive* the
facility since It would be treating stormwater.(See discussion In
Chapter 6.2) Dimensions of the Swirl Concentrator are given In
Tab Ie 12.
The design guidelines for the Helical Bend Regulator were
consulted In developing the design dimensions (12). An Inlet
diameter of 1.5 ft (46 cm) was chosen by extrapolating the design
curves below 10 cfs (283 l/s). Theoretical settleable solids
removal Is approximately 90%. Actual dimensions of the Helleal
Bend are given In Figure 30.
8.5 Mlseal lanaotis Das lyn Dat-al I s
Although the Swirl and Helical Bend Regulators were both
sized to have 1.5 ft (46 cm) Inlet diameters, 2.0 ft (61 cm)
diameter Inlet and outlet overflow pipe sizes were used to
minimize headloss. Piping configuration and sizes were shown In
Figure 4, Chapter 1.
The right angle 8 In. x 10 In. (20 cm x 25.4 cm) Hydro-
Brakes fabricated by Hydro Storm Sewage Corp. were designed to
provide a constraining discharge of up to 0.18 cfs (5 l/s). This
constraining discharge equals 3f of the design flow for the Swirl
Concentrator and Helical Bend Regulator.
As can be seen In Figure 4, the foul sewer lines from both
unlls discharge Into a foul sewer tank with pumpage Into either
the 120 In. (305 cm) storm drain or Into the 27 In. (68.6) YCP
sanitary trunk sewer. Solids-carrying capacity of this sanitary
sewer and segments further downstream were analyzed by computing
average fluid shear stress conditions at half-pipe conditions.
This condition was assumed since the sanitary sewerage system Is
separated but Is severely Impacted by clearwater Inflow. Shear
stress In the trunk sewer and downstream segments exceeded 0.06
Ib/ff2 (0.0026 kg/m2) which Is acceptable for self-scouring
conditions (32).
99
-------
- 4
Mill iii
30 ' 40 ' 50 ' 60
Cumulative Probability
Percentage
100
-------
The various components of the Installation and problems
encountered affecting the design of the facility are discussed In
Chapter 10. Instrumentation details are provided In Chapter 11.
TABLE 12
Design Dimensions of
Swirl Concentrator
West Roxbury
• ••»«w«MW««M»w««wwa»M«waB9MM«B«v«w««MM»a»««a»««iaMaBM»aB«»««M»v«fl
Dimension (ft) (•) (ft) (a)
DI • 1.50 0.46 RI 4.08 1.24
D2 • 10.50 3.20 R2 2'63 °'80
D3 - 6.93 2.11 R3 1.09 0.33
D4 - 5.83 1.78 R4 1.97 0.60
R5 6.42 1.96
HI • 2.63 .80 bj .58 0.17
b2 .50 0.15
101
-------
Note: 24in. clear
water channel
not shown.
481n.
Section A-A
VAVIAC rrna I Mm fn f\Ain
Typical Section B-B
Figure 30. Helical Bend Dimension*;,
W. Roxbury.
Inlet Diameter
102
-------
CHAPTER 9
CONSTRUCTION - LANCASTER PROJECT
9.1 FOREWORD
A general overview of the Lancaster swirl construction
Is presented In Section 9.2. Unique construction details
are given In section 9.3. Project costs are provided In
section 9.4. Instrumentation details are provided In
Chapter 11.
9.2 GENERAL
Design of the Lancaster Svlrl Project vas
completed by Huth In late 1976. Construction bids were
received In December 1976. Actual construction of the project
began on April 13, 1977. Construction consisted of tvo
contracts, one electrical and one general construction, '>hlch
were essentially completed by June 1978 when the project was
dedicated. Minor modifications wer"» not completed until
October 1978.
Following blasting at the site to facilitate rock
removal, site excavation was completed and the drains for
the central overflow, emergency overflow, foul outlet and
various sampling lines were laid. After backfill and
compaction, layout of the Swirl floor with radial
reinforced steel began, as shown In Figure 31. It should
be noted that the larger pipe In the foreground Is the clear
overflow, the shorter, smaller pipe Is the foul outlet and
the pipe In the background Is the emergency overflow.
Simultaneously, work progressed on the control building
basement and the diversion chamber. Figure 32 shows a vl3w
of the Swirl site during construction and s;ome scenes of
the completed project. Generally, the construction proceeded
smoothly, although there were a number of considerations
which required special attention, as discussed below.
9.3 SPECIAL CONSIDERATIONS
One ef the most difficult portions of the project was
forming the gutters In the floor of the Swirl Concentrator. A
large, level plywood platform was constructed on site and
the necessary shape of the primary and secondary nutters was
drawn on the platform with the assistance of a survey
transit. After this curving shape was established, forms were
constructed to the bet depths of the gutters by cutting
plywood to fit the curves on top and bottom with supports
between. These forms were constructed In sections to ease
placement on the Swirl floor as shown In Figure, 33. The
103
-------
Laying overflow ana drain pipes
cring bwin t loor
floor steel in place
Figure 31. Initial construction.
I Reproduced trom
available copy.
-------
o
in
The site during construction
Completed Swirl Project
Completed Swirl Regulator/Concentrator Top floor of control building
Flgure32. Lancaster Swirl Project.
-------
o
0»
Construction of Initial formwork
Placing gutter forms on radial staei
Swirl site during construction 'lacing concrete In Swirl Moor
Figure 33. Swirl floor and gutter construction.
-------
torn sides were than sealed with sheet metal and laid In
place on the radial floor steel. After the forms were
placed, the concrete «as poured and finished. The cross slope
of the Swirl Concentrator floor was established ar 1/4
In./ft (2.1 cm/a) tc aid In solids removal during flushing
after a CSO event. Figure 34 shows the final configuration
of the forms and the definitive shape of the gutters. A
rectangular gutter cross section was formed to simplify
construction during pl&cement of primary concrete. After forms
were removed, the rectangular gutter was grouted to an
hydraulleally efficient semi-circular shape to reduce shoaling
of sol I ds.
Another consideration during construction was the Inlet
section to the Swirl unit. Flow enters the swirl
tangentially and* therefore, required a suitable transition
shape. In addition, U.S. EPA design specifications (4)
required a square pipe while the Inflow pipe was 36
In.(.91m) In Diameter. Therefore, a cIrcuIar-to-square
transition piece was required. A tree-standing flow
deflector wall was constructed at the entrance to the swirl
to break up any vortex motion In the unit. These Inlet
configurations are shown In Figure 35.
An emergency overflow was required so that
overloading of the Swirl unit could be relieved by diverting
the excess flow directly to the river. The overflow
consisted of a slotted opening, or "window*, In the concrete
wall leading to a trough on the outside of the Swirl and
finally to the 60 In.U.Sti) outlet pipe to the river. Flow
over the emergency overflow weir could be measured at any
time since a small weir and level sensor was placed In
the bottom of the trough. The "window" weir was formed with
a small pillar between the two sections and the total weir
covered one-quarter of the Swirl's perimeter. An adjustable
weir plate was placed on the Inside of the unit to
control the amount of flow over the central overflow.
Figure 36 shows the two windows and the drain trough
during construction.
The prefabricated weir and floatables trap assembly,
proved difficult to Install because It was large and bulky,
and required precise placement for correct performance.
Also, In order to prevent any clogging of floating matter
In weir plate supports, the mounting had to be performed
from above. Four large 3/4 In. (1.9 cm) adjustable bolts
suspended +he assembly from the steel walkway, thereby
eliminating any possible protrusions which could catch any
floating matter. Figure 37 shows a view of this prefabricated
as sent b I y.
107
-------
Finished gutfers before grouting
jrouting floor gutters
Figure 34. View of gutters In Swirl floor.
108
-------
Concrete flow deflector
The circular-to-square transition
Inside view of the tangential transition piece
Figure 35. Swirl inlet configurations.
109
-------
View through the weir window Into Swirl unit
Construction of emergency overflow trough
figure 36. Emergency overflow weir.
no
-------
The prefabricated unit
Following installation
Figure 37. Weir and floatables trap assembly.
in
-------
The remainder of the construction proved to be routine and
presented no particular problems.
The Svlrl Degrltter was placed In the lover level of the
control building before completion of the upper levels due
to the size of the unit. The Oegrltter was Installed In a
lowered floor section to provide for easy observation and
to provloe sufficient head for gravity f I o« of the larger
Svlrl unit's underflow to the Degrltter. A sump vas provided
In this lovered section for drainage. Figure 38 shous the
control building floor, the Svlrl Degrttter and the grit
hand! Ing system.
9.4 CONSTRUCTION COST
total cost of the Lancaster s»lrl concentrator
demonstration project was about $1,100,000 (ENR 3000). In
addition to the construction cost of the Swirl facilities, this
cost Includes Inspection and administration, engineering,
equipment, supplies, laboratory charges, operating salaries,
utilities, reoa'rs, and the project report.
Construction of the Lancaster Swirl Project was
divided Into two contracts. Electrical work was done by
Lancaster Electric Co. at a cost of approximately $123,000.
General construction work was performed by Toews, Ayres &
Huber, Inc., a t^ial contractor, at a final cost of
approximately $369,000. The construction costs of the Swirl
facilities were estimated at about $692,000 of which the Svlrl
Regulator/Concentrator cost was estimated at about $168,000 or
about $6,300/mgd ($l48,000/mVs) of design capacity. The other
construction costs Include the Swirl Oegrltter and grit
handling system, the control building, the Instrumentation
and the disinfection system. No land acquisition costs were
Incurred since the City of Lancaster already owned the project
site. These figures also Include a number of various change
orders Incurred due to first-time design and construction.
Also, some extra costs such as the Dl scostral ner and
extensive lab equipment were Incurred for the latter
evaluation work. Table 13 shows a breakdown of the general
construction costs by contract amount. Table 14 presents a
more condensed elaboration i
-------
Degritter forms In control building floor
Installation of the Swirl Degritter
The Swirl Degritter In operation Grit conveyor belt and receiving dumpster
Figure 38. The Swirl Degrftter.
-------
TABLE '3
LANCASTER SWIRL FACILITY
CONSTRUCTION COST BREAKDOWN*
Work Description
General Construction:
Amount
1. Mobilization and Job Overhead
2. Excavation & Demolition
3. Fencing, Paving, Erosion Control & Landscaping
4. Concrete, Forms, Reinforcement & Masonry
5. Prefabricated Swirl Degrltter & Swirl
Concentrator Parts
6. Roofing, Painting, Doors, Misc. Hardware & Metal
7. Plumbing, Piping, Valves, Heating, Ventilation
and Misc. Accessories
8. Dlscostralner, Samplers, Pumps, Hoists &
Chi or I nation Equipment
9. Screw and Belt Conveyors and Grit Scale
10. Laboratory and Monitoring Equipment
11. Change Orders
$28,000
49,700
17,200
103,600
16,000
60,700
107,300
79,000
39,700
47,800
Sub-Total S 568,945
Electrical Construction:
1. Excavation and Backfill
2. Conduit, Wire, Switches, Receptacles
3. Heaters, Thermostat, Mounting & Connection
4. Flowmeters and Recorders
5. Telemetering Alarm System
6. Generator and Motor Control Center
7. Control Panel and Misc. Panels
8. Miscellaneous
9. Change Orders
Sub-Total
Total Cost
1,837
25,280
2,644
29,151
5,456
26,842
21,749
6,752
3.6flQ
123,391
692,336
1977 costs (ENR = 3000)
114
-------
TABLE 14, SUMMARY OF SWIRL FACILITIES
CONSTRUCTION COSTS, LANCASTERa
Item Cost,f
Swirl Concentrator 168,000
Swirl degrltter and 56,000
grit handling systen
Control building 255,000
Instrumentation 146,000
Disinfection system 67-.000
Total 692.000
a.ENR 3000
9.4.1 Operation atid Ma \ ntanance Cos-ts
The operation and maintenance cost for the first year
of operation of the demonstration facilities Is estimated at
about $58,000. The cost Includes operating personnel
salaries, supplies, utilities, and a budget for equipment
repairs or replacement.
There Is a large potential for startup problems and
facilities debugging efforts with any newly constructed
facility. The annual operation and maintenance budget after the
startup period and the demonstration period Is expected to be
about one-half the first year cost, about $26,000.
115
-------
CHAPTER 10
CONSTRUCTION - WEST ROXBURY PROJECT
Section 10.1 Foreword
An historical overview of the various phases of construction
Is presented In section 10.2. Site preparation and Iniet
structures are described in section 10.3. Swirl fabrication
and Installation details are provided In section 10.4.
Fabrication and Installation details for the Helical Bend
regulator are given In section 10.5. Influent structure
appurtenances are described In section 10.6. Clear water and foul
sewer discharge details are provided In section 10.7. Security
•easures are described In section 10.8. Measurement devices are
discussed In section 10.9. Details of the lift station
constructed to simulate CSO are discussed In section 10.10.
Section 10.2 HIstorlea I Overvlaw
Grant awards were completed by March, 1978. Preliminary
site construction and security measures were Initiated during
the summer of 1978. A control building was placed on site during
this period. Extreme vandalism to the existing security system
fnd control bitiding precluded further site work and fabrication
of the units on-slte. Instead, EDP fabricated the Swirl and
Helical Bend regulators In components In Its facility during the
fall/winter of 1978. Additional security measures were completed
during the spring of 1979 and all site work, piping and
Installation of the units were accomplished during the summer.
During this period U.S. EPA desired to augment the existing
grant to Include appurtenances and housing for a third solids
seperator device, the Teacup, and to render the site Into a more
suitable form for viewing by Installing a catwalk system. Most
of the work at the facility was completed by September, 1979 and
preliminary evaluation work began In October. Vandalism continued
to plague the facility during the fa I I/wInter/spr I ng of 79-80.
Further security measures were taken during this period.
After long delays the Teacup finally arrived from California
In April 1980. Upon Inspection and close technical 'review the
unit was deemea Inappropriate by U.S. EPA and
Installation/evaluation of Teacup was cancelled.
A bypass foul sewer gating system for the Helical Bend
regulator was constructed during the summer of 1980. This
modification was added to minimize accumulation In the Helical
Bend following storm events.
116
-------
IP the fall of 1980 a timber slulca was constructed within
the 87 In. (2.2 m) RCP line connecting on-grade the 120 In. (3m)
RCP line on New Haven Street to the Influent chamber.
Photographs In Figure 39 show the completed facility. For
ease In reference a schematic of the facility was depicted In
Figure 4 In Chapter 1. Photograph A, Figure 39 (Figure 39-A) was
taken from the roof of The control building along the north/south
axis. The He I I ca I Bend regulator appears In the lower left hand
side of Figure 39-A. The 120 In. (3m) RCP outfall which receives
the drainage from the 87 In. (2.2) RCP appears In the upper
right hand side of Figure 39-A. The catwalk system shown In the
middle of Figure 39-A obscures the Swirl Concentrator from
view. Outlines of the foul sewer sump tank, the Teacup housing
and the two Palmer-Bow I us flumes can be seen behind the catwalk
In Figure 39-A.
The view of the site shown In photograph B, Figure 39
(figure 39-B) was taken from on top the 120 In. (3m) conduit In
a southerly direction (see Figure 4). The Swirl Concentrator
appears In the upper left hand side of Figure 39-B. The 24 In.
(0.61m) steel Helical Bend clear water discharge line
connecting to 120 In. (3m) RCP appears In the lowermost
foreground of Figure 39-B. The 8 In. (20.3cm) steel foul sewer
discharge from the Helical Bend together with the Palmer-Bowl us
Flume housing appears to the right of the catwalk stairway In
Figure 39-B. The housing for the Swirl foul sewer PaImer-BowI us
Fiume with ultrasonic liquid level recording head Is shown Just
to the left of the catwalk stairway In Figure 39-B. The steel
tank housing for the Teacup Is just to the of the
aforementioned flume In Figure 39-B. The foul sump tank
containing a 250 gpm (15.8 l/s) trash sump pump and the 3 In.
(7.5cm) pvc discharge appears In front of Teacup housing In
Figure 39-B.
Section 10.3 Site Preparation
The Swirl Helical Bend Regulator/Concentrator treatment
complex covers an area of 120 ft x 120 ft (36.6m x 36.6m). The
site Is located within a natural drainage gully. The westerly
side of the site Is abutted by an abandoned cinder-filled
railroad bed. The access road to the control building within
the complex lies on top of the cinder bed. Grade for the site Is
about 12 ft (3.7m) below natural contour of the gully and at the
sprlngllne of the 120 In. (3m) drain connecting to the Charles
River 1000 ft. (305m) downstream.
117
-------
co
A, PHOTO A TAKEN FROM CONTROL BUILDING
B, ftoro B TAKEN FROM EMBANKMENT
(OPPOSITE DIRECTION AS PHOTO A)
Figure 39. Photographs of completed facilities
-------
Roughly 4000 cubic yards (3100 cubic meters) of earth
required excavation. F:!! material was buit dozed (see photograph
A, Figure 40) to form two traverse embankments cutting across the
gully. The existing 120 In. (3m) drain bordering the east side
of the site was laid at a shallow depth and with 5 ft (1.5m)
of cover formed a natural embankment. A number of springs were
unexpectantIy encountered as excavation began. Ground water
pooled In the site Just at final surface grade (see photograph B,
Figure 40). Soil at grade was silt and clay with pockets of
sand. All yardwork, slabs, and housing were constructed under
severe groundwater conditions as site was a quagmire even during
dry weather conditions.
The Inlet diversion chamber Is a 25 ft (7.6m) x 15 ft (4.6m)
reinforced concrete box w I rh slue weir Inlets. As site
excavating began a 3 ft (,9Tn) ring of high strength reinforced
concrete enclosing the conduit (see photograph A, Figure 41) was
unexpectantly uncovered. A "window" s*s cu+ through the conduit
after several weeks of backhoe j&ck- hammering (see photograph B,
Figure 41). The diversion chamber was formed (see photograph A,
Figure 42) and 12ln. (30.5esi) concrete "•Mis poured.
Two 24 In. (61cm) steel pipe headers from the diversion
chamber leading to the two units were Installed (see photograph
B, Figure 42). Each header had a steel rectangular section (see
photograph B, Figure 42) containing two bottom-acting sluice
gates (see photograph A, Figure 43). The gates were used to
modulate flow into thi two treatment devices. A 24 In. (61cm)
Pa I mer-BowI us flume with a Manning ultrasonic liquid level
sensor was Installed on each header about 15 ft (4.6m) downstream
of the sluice gate chambers to monitor discharge (see photograph
B, Figure 43). Strip chart assemblies were maintained In the
control building. After the two 24 In. (51cm) Pa Imer-BowI us
flumes the Influent headers to both units reduced to 18 In.
(46cm) using steel reducers.
Section 10.4. Swirl Concentrator Fabr leat ton/I nstal la-tlon
Due to the severe vandalism experienced at the site In the
summer of 1978 on-slte fabrication of the unit was not attempted.
The Inner assembly shown In photograph A, Figure 44 was
fabricated at EDP's shop. The Swirl steel shed was also shop-
fabricated. Photograph 8, Figure 44 shows the Swirl shell being
lowered Into position by a crane at tho demonstration site. The
Influent transition section and the primary/secondary gutters
appear at the bottom of the steel tank In photograph Bf Figure
44. The 24 fn. (61cm) clear water downshaft was then field
we'ded and Influenv baffle placed (see photograph A, Figure 45.
The Inner assembly was then placed, leveled and field-welded (see
photograph B, Figure 45). Photograph A, Figure 45 shows a top
119
-------
T^L-) .fl[ • •• • " v'*^ •*^fcT*«H-'
»•> . * T*. ^^SL.
"*"r*^- - • • • * •*.
A, INITIAL Cur
Reproduced from
beit •vailabU copy.
B. Cur ABOUT Six FEET ABOVE FINISHED
SURFACE GRADE
Figure 40. Photographs of site excavation,
-------
A, EXPOSED EXISTING 87iN, (RCP)
W /CONCRETE ENCASEMENT
B, CONCRETE DEMOLITION
Figure 41. Photographs of Influent Diversion Chamber concrete demolition.
1 1n • 2.54 cm
-------
A, FORMING INFLUENT SIDE WALL
SPILLWAY
B, INFLUENT PIPING MANIFOLDS
Figure 42. Photographs of Influent Diversion
Chamber •• sldewall construction details.
-------
A, TOP VIEW, BOTTOM-ACTING SLUICE
GATES, SWIRL INFLUENT
B. TOP VIEW, HELICAL BEND
PALMER-BONUS FLUTE (ULTRASONIC)
Rgure43. Photographs of kifluent Gates & Flumes.
-------
A. INNER ASSETCLY
B. SWIRL STEEL SHED (UNDERSIDE)
Figure 44. Photographs of shop-fabricated
Swirl Solids Separator components, W. Roxhyry.
124
-------
ro
ui
A, SWIRL STEEL SHED INTERIOR
(CLEARWATER £HAFT-LHS)
B, LOWERING PRE-FABBED INNER ASSE>BLY
Figure 45. Photographs of field assembly
- Swirl Solids Separator. M. Roxbury.
-------
en
A, TOP VIEW OF FLOATABLE TRAP
B, SWIRL AT END OF STORM
Figure 46. Photographs of completed Swirl Solids Separator, U. Roxbury,
-------
view of the floatables trap. Note that floatables are
restrained between the side of the tank anJ ti>« scum ring and
then are swept Into the trap by the swirl motion within the
tank. Localized vortex motion draws and stores the floatables
within the trap below the weir plate/collar. Photograph B,
Figure 46 shows a top view of the Swl1*! Immediately after a storm
event.
Section 10.5. Ha I I ca I Band Fabr I cat Ion/ I nstaH tlon
Photograph A, Figure 47 shows a stack of cut plywood
sheets depleting the geometric progression of the Helical
Bend's cross-sectional ribs. Two ft (.61 m) wooden 2 In. (5
cm) x 4 In. (10 cm) studs were cut and beveled at each end
(see Photograph B, Figure 47). A computer program was written to
describe In three dimensions the various bevel cuts for roughly
1200 stud segments. Different beveled cuts were necessary
since the Helical Bend cross section flared In both plan and
profile. The logistics of these cuts were accomplished only
after a great deal of difficulty.
Photographs A and B, Figure 48 depict typical
construction details for attaching wooden studs to the plywood
ribs. Each stud was attached 8 In. (20.3 cm) on centers t" the
plywood sheets with wood glue and several wood screws. Studs
were staggered on each side of the rib as shown In both
photographs In Figure 48. Triangular plywood sections were
attached to the outside of each rib adding overall stability.
Photograph A, Figure 49 depicts the triangular stabilizers.
Photograph B, Figure 49 was taken from the Influent
end and depicts the transitional and straight sections of the
Helical Bend. Photograph A, Figure 50 was taken from the
center line of the straight section within the Helical Bend
and depicts the flared cross section of the transition
section. Photograph B, Figure 50 was taken In reverse
direction and shows the treatment section. The clearwater weir
and backside of the clearwater overflow trough are depicted In
the background of Photograph B, Figure 50 while the wall
nearest the foul sewer trough appears In the right foreground.
The plywood ribbed structure was then sheeted with 0.040
In. (1mm) 6061-T6 aluminum sheeting. Photograph A, Figure 51
shows partial cover of aluminum sheeting on the clearwater
weir while photograph B,Figure 51 shows sheeting on the foul
sewer trough. Photograph A, Figure 52 was taken from the
effluent end of the Helical Bend regulator showing the curved
foul sewer section In the foreground and the clearwater weir In
127
-------
00
A, STACKED PLYWOOD RIBS
B, BEVELING WOODEN STUDS
Figure 47. Photographs of Helical Bend wood fabrication.
-------
vo
A, FORMING CROSS-SECTION
Figure 48. Photographs of Helical Bend Regulator,
wooden fabrication details.
-------
OJ
O
A, STRAIGHT SECTION
B, TRANSITION & STRAIGHT SECTION
(PHOTO TAKEN FROM INFLUENT END)
Figure 49. Photographs of Helical Bend wooden shell (In progress).
-------
A, TRANSITION SECTION (PHOTO TAKEN IN
STRAIGHT SECTION)
B, TREATMENT SECTION (CLEAR WATER WEIR
IN BACKROUND)
Figure 50. Photographs of Helical Bend wooden shell (completed).
-------
U)
ro
A, PARTIAL SHEETING - OVERFLOW WEIR/TRFATMENT SECTION B, PARTIAL SHEETING - R>u_ SEWER OUTLET
Figure 51. Photographs of Helical Bend aluminum sheeting Interior.
-------
A. Aluminum Sheeting - Foul Sewer Channel
8. Aluminum Sheeting > Interior
Figure 52. Photographs of Helical Bend
aluminum sheeting interior.
133
-------
the upper background. As the sheeting was nearlng completion.
Photograph B, Figure 52 was taken from the Influent end and
oepicTs the clearwater weir In the background.
A 12 In. (30.5 cm) reinforced concrete slab was poured at
a 2% grade as shown In photograph A, Figure 53. The four
sections of the Helical Bend were transported from EDP's shop
facility to the site and lowered onto the concrete slab by
crane (see photograph B, Figure 53). The regulator was
fabricated Into four structurally Intact sections
(approximately 15 ft (4.6 m) In length). Plywood ribs were
attached by angle Iron onto two welded steel channel pieces
to form each section of the regulator. Each section had been
fabricated as a structural member capable of withstanding
erection stresses.
The external surfaces of the Helical Bend were primed (see
photograph A, Figure 54) and the Inner surface sanded (see
photograph B, Figure 54) before two coats of epoxy were
applied. Protective plywood sheeting was then fastened to the
outer triangular stabilizer (as shown In photograph A, Figure
39) and then curved stainless steel scum and clearwater weir
baffles were attached (as can be seen In photograph A, Figure
39).
Section 10.6. I nf I uant Strtietura Appurtenances
As previously mentioned In Chapter 8, discharge from the
catchment area Is through a 30 In. (76.2 cm) RCP on New Haven
Street Into a short segment of 87 In. (221 cm) RCP
approximately 250 ft (76.2 m) In length. The Inlet chamber
to the demonstration site Is along the 87 In. (221 cm) pipe
section.
The original plan for diverting flow In the 87 In. (221
cm) Influent storm drain to the facility consisted of 2 In.
x 6 In. (5 cm x 15 cm) creosoted wooden stop-logs placed In
slots constructed In the Influent chamber box forming an
adjustable weir to control sldeway discharge Into the Inlet
CMamber box. After the facility was completed In September, 1979
and during the Initial debugging period* an EDP field crew on
September 30 placed stop-logs at a nominal setting as to
create discharge into the units Just before a storm event. The
Intensity of the storm was severe, creating nearly full pipe
discharge In the 87 In. (221 cm) line. The field crew
attempted to manually remove several planks as to relieve
the backwater effects. The hydrostatic pressure was such
that the planks had to be removed with a chain-fall
involving EOP personnel climbing Intc the 87 In. (221 cm) line
at great personal hazard. To remedy this hazard, stop-logs
134
-------
OJ
01
FIELD ASSEMBLY OF HELICAL BEND SECTIONS
HELICAL BEND CONCRETE PAD
Figure 53. Photographs of Helical Bend on-slte Installation.
-------
A, HELICAL BEND EXTERIOR - OIL PRIMED
B, HELICAL BEND INTERIOR - ALUMINUM
SHEETIMG PREPARED FOR EPOXY COATING
Figure 54. Photographs of final surface preparation for Hellcai Bend.
-------
vere attached In sections to form a segmented bottom-
opening sluice gate operated by y motorized assembly located
on top of the diversion chamber.
As mentioned above a baffle system MBS constructed to raise
the vater level up to the aid-point on the 87 In. (221 cm)
conduit allowing for side discharge Into the treatment
facility. An upward-sloped vooden ramp was also constructed to
move particles If entrained In the 87 In. (221 cm) line up
Into the units. It turned out that during low I ntens I ty/ I ong
duration rainfall events the ramp was Inadequate In
transporting settleable solids Into the units. To rectify this
problem, a vooden flume In the 87 In. (221 cm) conduit
connecting the end of the 30 In. (76.2 cm) RCP line on New
Haven Street to the 2 Inlets at the treatment facility vas
built In fall, 1980. This wooden sluice Is approximately 185
ft. (57 m) long. The flumes rectangular cross section was later
revamped Into a triangular section In spring, 1981. The revised
section permitted average velocities exceeding 3 ft/sec (.91 m/s)
for all depths of flow.
SectlonlO.7
The clear water discharge from the Swirl Is below site
grade and Is a 24 In. (61 cm) steel pipe connected to the 120
' n. (3 m) storm drain about 1.5 ft (.46 m) about Invert. The
Helical Bend clear water discharge Is similar In dimension
and appears on the left side of photograph B, Figure 39. The 8
In. (20.3 cm) steel foul sewer Is again below grade extends
roughly 25 ft (7.6 m) on a flat slope and Connects to a flanged
8 In. x 10 In. (20.3 cm x 25.4 cm) 90 denree Hydro-Brake.
The 10 In. (20.3 cm) foul sewer line contlnuas to a 10 In.
(25.4 cm) Palmer-Bow I us flume with a Manning ultrasonic liquid
level sensing head. The Helical Bend foul sewer discharge Is
again an 8 In. (20.3 cm) steel line and Is above grade (see
photograph B, Figure 39). The 8 In. x 10 In. (20.3 cm x 25.4
cm) right angle flanged Hydro-BraKe for this line can also be
seen In the photograph on Figure 55.
Both foul sewer lines discharge Into the foul discharge
tank 10.5 ft (3.2 m) In diameter and 6 ft (1.8 m) In depth.
This tank appears In the foreground In photograph A, Figure 56.
The foul sewer sump trash pump and piping assembly appears In
Photograph B, Figure 56.
As mentioned above +he foul sewer outlet on the Helical
Bend Is regulated by . a right angle Hydro-Brake, permitting
discharge of about 0.2 cfs (5.7 I/sec) Implying discharge
velocity of 0.7 fps (21.4 cm/sec). The ovei — sized line was
Installed to prevent debris clogging and the Hydro-Brake was
137
-------
CO
Figure 55. Photograph of 8in X lOin 90° flanged Hydro-Brake
(Helical Bend Foul Sewer). 11? - 2.54 cm.
-------
A. Top View of Foul Sewer Pump
6. Trash Pump
Figure 56. Photographs of Foul Sewer Sump.
-------
Included to limit underflow to about 3? of design flow, 6
cfs (170 I/sec). Substantial deposition of settleable and
floatable materials resulted during the early phases of the
evaluation programs. In the lower effluent end of the Helical
Bend treatment section during rainstorms because of the low
exit velocities and the off-set of the 8 In. (23.2 cm) x 10
In. (25.4 cm) Hydro-Brake. This situation requires that the unit
be hand-cleaned after an event. To remedy this situation, an
81 n. (23.2 cm) bypass line with a motorized valve was
constructed with the Inlet at the same elevation of the
existing foul sewer outlet and the outlet connected to the
foul sump tank. The 8 In. (23.2 cm) motorized valve provided a
quick flush of the remaining volume within the unit. The
motorized valve was activated from the control building.
Section 10.8. SaeurIty Heasuros
The demonstration site was Initially surrounded by an 8
ft (2.4 m) high chain-link cyclone fence topped with three
strands of barb-d wire. A 20 ft x 20 ft (6.1 m x 6.1 m) pre-
fabbed modular contro1 building was moved In and erected within
the fenced perimeter. This building had been formerly used In a
water pollution control RAD project In Boston, (see photograph A,
Figure 57). The project site Immediately became the +arget of a
tremendous amount of vandalism Including crushed barbed wire
supports, defacing and burning of the control building,
smashing of existing explosion-proof lighting fixtures, and
destruction of all existing control board circuitry and p»»«!s.
The building was twice burnt and electrical control panels
replaced by EDP (see photograph B, Figure 57).
Additional security measures Included: a second 12 ft (3.8m)
high non-climbing security fence which was topped with razor-
blade ribbon barbed wire (see photograph A, Figure 58), the area
was defollated, overhead I Ights were Installed, on-slte guard
service was used during sensitive construction times, and sensing
alarm systems Installed In the control building and at two
locations on the catwalk sending signals both to a nearby
guard service and to turn on audio alarms at the site.
(Photograph B, Figure 58 shows the catwalk under construction).
The security fence gate was vulnerable to vehicles
traveling at high speed on the access road. Tvlce,
vehicles travelling at high speeds crashed and completely
demolished both sets of gates. EDP fabricated and replaced both
pairs of gates for the two security fences and finally
constructed a steel "crash" fence located 15 ft (4.6 m) In
front of the outer fence constructed of heavy steel "I"
beams with side braces mounted on 6 In. (15 cm) heavy wall
steel pipes, filled wltn ojncret? and embedded In substantial
-------
A, CONTROL BUILDING (SECURITY GUARD NEXT*
TO BUILDING)
B, RESTORING VANDALIZED ELECTRICAL
CONTROL PANELS
Figure 57. Photographs of control building, U. Roxbury.
-------
ro
A, SECURITY FENCES
B, CATWALK CONSTRUCTION
Figure 58. Photographs of sltework.
-------
concrete footings. Major vandalism was stopped but rock missies
thrown from outside the area perimeter constantly demolished
control building windows.
Section 10.9. Measurement Dev!ees
Monitoring of the performance of the two control devices
required two types of measurements, both flow and quality,
at several polnfs along each regulating device. Flow was
monitored at the Influent, clear water effluent and foul
sewer discharge utilizing Palmer-BowI us flumes and ultrasonic
meters. Quality measurements were taken dt the drain outlet
box (common Input to both treatment devices), clear water
discharges and effluent streams by use of both flow
compositing and dI sere he automatic sampling devices. The
actual readouts for the flow monitors and controls for
the sampling devices were housed In the building Itself.
Further details are provided In Chapter 11.
Section 10.10. SImuIated CombIned Sewer LI ft StatIon
Part of the Swirl and Helical Bend evaluation program
entailed testing the efficiency of the two units using as
Influent simulated combined sewer overflows. As shown on Figure
4, Chapter 1 a lift station was Installed to pump up to 2 cfs
(56.6 l/s) of sanitary sewage from the 27 In. (69 cm) VCP trunk
sewer (adjacent to the 120 In. (305 cm) RCP storm drain) during
rainfall events and mix this waste with Incoming storm drainage
from the catchment area. Dry weather flow In the 27 In. (69 cm)
line was gaged for several months In 1978. Average flow equalled
0.5 cfs (14.2 l/s). The tributary sanitary system was severly
Impacted by clearwater Inflow and flows quickly rose typically to
9 cfs (255 l/s) during several monitored rainfall events.
U.S. EPA provided a used 6 In. (15 cm) pump capable of
discharging 2 cfs (56.6 l/s). The pump was Installed In summer,
1979. The first time It was turned on the pump seals burst. EOF
disassembled the pump, repaired the seals and replaced the pump
during 1980. The pump was again turned on during an evaluation
event during spring, 1980 (4/4/80) and again the seals burst and
the Impellers were damaged. Since pump components had to be hand-
carried down the eubankment In pieces and then assemblled In an
extremely space-constrained area. It was deemed Infeaslble to
again repair this pump. This portion of the evaluation project
was not performed.
143
-------
CHAPTER 11
Eva Iuatton Methods
11,1 Foreword
This chapter details the methods used In monitoring the
treatment units for both projects. An overview of the
Instrumentation methods for the Lancaster facility Is presented
In section 11.2. Sampling schemes at Lancaster are discussed In
section 11.2.1. Problems with the Initial 1978-1979 Huth
monitoring program are presented In section 11.2.2. Specialized
samplers which were fabricated and Installed In Lancaster by EOP
Technologies, Inc. are described In section 11.2.3. Problems with
the flow monitoring devices In Lancaster and calibration efforts
are reviewed In section 11.2.4.
Section 11.3 Includes an Instrumentation overview of the
West Roxbury facility. Sampling schemes are discussed Including
related problems and solutions and flow monitoring devices are
reviewed. Section 11.4 details the design, and Implementation of
the motorized bl-dlrectlonal rotating settling column, along *lth
a description of the settling column used on-slte at Lancaster.
Section 11.3 reviews all the analytical procedures used In the
analysis. Section 11.6 describes various calulatlon procedures
for noting performance of the two demonstration facilities.
Section 11.7 describes various methods for presenting settling
column results.
11.2 Evaluat1 on Program OvervI aw - Lancaster Swirl Project
Initial EDP Rev lew and Project Startup
On July 19, 1979 EDP was Invited to visit the Lancaster
Swirl facility to assess the numerous and persistent monitoring
program difficulties. The site visit began with a review of
sampling difficulties encountered to date. Inspection of the
existing sampling system yielded numerous system faults and
recommendations as outlined In section 11.2.1. Based on this
visit, EDP was contracted In July, 1980 to oversee the remainder
of the evaluation program.
During the early stages of EDP's Involvement with the
project, the resident engineer provided EDP with written records
documenting project activities starting 7/25/78 through 1979.
These records Included dally log books, storm e/ent rain records
and operational records depleting flow rates and pollutant
concentratIons.
144
-------
Review of this material provided the needed Insight for
propei program formulation for future activities. The dally logs
were studied to determine typical operational procedures,
equipment reliability, repairs and changes ana the actual
adequacy of the reports themselves.
A listing of rain events was made Including the necessary
Information to determine typical frequency, duration and
Intensity of storm events. Combined with the dally reports and
sampled event data, estimates as to the types of storms worthy of
evaluation were determined. Strip-chart records for numerous
storm events were reviewed to estimate typical flow response to
rain events and determine the quality of flow data. Based on this
review equipment repairs were recommended and optimal level
sensor adjustments were determined.
As of July 1980 the evaluation project had been Inoperative
for many months. During th!
-------
First, the facility layout has normal dry weather flow
passing through the unit. Storm event monitoring Is Initiated
when the depth within the swirl Is 18 In. (46 cm). Calculations
show that at the time tha swirl lave! reaches IS In. (46 cm),
flow through the Degrltter bypass pipe Is 8 cfs (226 l/s).
Considering the normal dry weather flew of 3 cfs (85 l/s) much of
the Initial storm flow will have passed through the swirl before
wet weather operation Is Initiated. During thl- time no samples
are collected, thereby missing a portion of the "first flush". At
this time, the Swirl Is filling at amuch slower rate than would
occur If the maximum bypass rate was closer to the actual dry
weather flow rate. Two corrective measures were considered. First
a f low-I Ira 111ng device could be Inserted In the Degrltter bypass
line. Second, the wet weather activation level could be reduced
belov 18 In.(46 ca).The problem with each of either of these
approaches Is the tendency to clog the Degrltter with organic
materials. Either of the above measures would cause occasional
unit activation and valve-opening to the Degrltter during
abnormally high dry weather flow and the high organic solids
content of this flow would cause continual problems with the
degrltter. This problem was not resolved during the evaluation
program.
Second, the storm event monitoring Is terminated when the
Swirl level returns tc 18 In. (46 cm) at which point the
Degrltter bypass Is opened. With the Lancaster setup the flow
then also bypassed the foul effluent flow meter and sampling
port. Evaluation of the Swirl at West Roxbury has Indicated that
much of +he solids matter Is not continually removed from the
unit but accumulates until the end and passes as a "final flush*.
This Implies that at the Lancaster facility much of the solids
may be discharged from the unit after sampling ends.
Third, the Lancaster facility had a contlnuence of equipment
malfunctions. One Influent flow meter was not operating and a
backup meter periodically malfunctioned. In addition, various
level sensors and recorders were of questionable accuracy since
It was discovered that dimensions used In calibrating many of the
strip chart recorders were unknown. In some cases actual settings
were unreasonable. A low sotting on the emergency weir level
meter created off-scale fiow readings and the diversion chamber
settings greatly limited resolution.
Field Trips
In addition to preliminary site evaluation, EDP personnel
visited the Lancaster Swirl sife In summer, 1980 to discuss the
scope of the evaluation program after a complete review of the
facility h'story and operation and an Inspection of all equipment
had been made. After four storm events had occurred In the fall
146
-------
1980 and tie data assessed, additional meetings were held to
discuss the further recommendations and required changes. Due to
the continual problems with flow monitoring equipment on-slte
visits were deemed necessary.
During these visits the following actions were accomplished*
1) Accuracy and calibration check of Swirl ultrasonic
IIquld Ievel meter.
2) Installed Instrumentation necessary to obtain a second
Swirl level record to provide an expanded chart span
at the height of the Swirl weir.
3) Reset spans of various other level sensors to optimize
strip chart records. This work Included expanding the
emergency overflow weir record span to prevent
additional "off-chart" events and reducing *he
diversion chamber span to yield more reasonable
resolution. Communications with Huth Engineering and
City of Lancaster I ndIcatedthat for many of the
earlier strip- charts the referencing dimensions were
unknown. These were determined as necessary.
4) Performed hydraulic tests upon the Inoperative 475
Doppler Influent flow meter (accompanied by the Leeds
and Northrup representatIves).The meter proved
unrepairable In the field and was removed for factory
Inspection and repair.
5) Performed hydralIcs tests upon the malfunctioning
Mapco ultrasonic meter.
6) Measured and recorded level of Swirl clear overflow
weir. Attempts to level weir by adjusting turnbuckle
supports resulted In a level weir at the four points of
support contact, but non-level conditions In
mId-sectIonsdue to plate warpage. Final level
measurements were taken at numerous points to assist In
using the circular weir equation for flow measurements.
However,It was assumed that continual plate warpage
due to differential thermal heating would render these
values useless In a short time.
7) Inspected the 36 in. (91.4 cm) Influent pipe Including
the area selected for Installation of thc> dye Injection
system and the area of the ultrasonic transducer ports.
Debris was found In the transducer ports leading to the
recommenoatlon that a cleaning and Inspection routine
be developed.
147
-------
In a parallel effort described In section 11.2.3 EDP
Technologies, Inc. fabricated, shop-tested and Installed two new
sampling devices on the Influent conduits to the Swirl
Concentrator and Swirl Degrltter. These devices were designed to
take a liquid sample from the entire cross-sectional pipe area
over a short Instant of time. The purpose of the devices was to
ensure representative sampling of flow/solids stratification
characteristics, that Is, bedioad suspended load and wash load.
11.2.1 Data I Is of Samp I Ing LoeatIons at Lancaster
The Initial sampling program at Lancaster consisted of a
series of Manning 6000 samplers located In the laboratory area on
the second floor of the facility. Operation of these units
required drawing the samples from various distances within the
basement area ranging one to two stories below the units. In the
first year of Huth evaluation M978-1979) this measurment system
proved to be Inadequate In providing representative grit
components In the samples. EDP recommended two series of
adjustments.
The first set of adjustments Involved re-pos111 on Ing the
Manning samplers and revising the Intake systems to take maximal
advantage of existing equipment. The second set of adjustments
Involved Installation of new state-of-the-art sampling equipment
by EDP Technologies, Inc. Two sampling schemes were used during
the EDP (1980-1981) performance evaluation. First, the Manning
samplers wore located to provide the best possible samples under
the given circumstances. Tnese positions are Indicated In Figure
59. Swirl Influent was sampled In the 36 In. (91.4 cm) line Just
prior to discharge Into the Swirl Concentrator. The Manning
sampler was located In an enclosed structure at ground level and
coupled to an Intake consisting of three copper tubes placed
horizontally across the 36 In. pipe. Holes drilled In the tubes
were sized to provide optimum Intake velocity, good cross
sectional representation and allow catchment of maximum diameter
particles. The Swirl clear overflow samples fere collected using
a Manning sampler located on the catwalk. A copper Intake tube
was positioned horizontally along the Inner side of the overflow
weir.
Swirl foul flow was sampled via a port located tangentially
to the Invert of the swirl degrltter Inlet transition section. A
probe was Inserted through this port and positioned to sample the
center pipe flow. An Intake similar to that on the Swirl clear
weir was used for degrltter overflow samples. Both the Swirl foul
flow and Degrltter clear flow samplers were Manning units located
on the basement floor.
148
-------
SAMPLER KEY
$• MANNIIM SEQUENTIAL AUTOMATIC SAMPLER*
A SPECIAL Of TECHNOLOGIES CROSS
SECTIONAL SAMPLERS
\
SWIRL OE6RITTER EFFLUENT
TO STEVENS AVE
PUMPING STATION
DRY WEATHER BYPASS-
CONCENTRATED (FOUL) UNDERFLOW-
• COMBINED SEWER
DIVERSION CHAMBER
SWIRL CONCENTRATOR
EMERGENCY OVERFLOW WEIR
EMERGENCY OVERFLOW TROUGH
SWIRL CONCENTRATOR EFFLUENT
SLUICE GATE
SWiRL UWT BYPASS
HYDROBRAKE
CONTROL BUILDING
SETTLING COLUMN SAMPLES
FLOW METER
HYOROBRAKE
COP FOUL
SWIRL UNIT INFLUENT
MANNING INFLUtNT
MANNING CLEARWATER
TO CONESTOGA RIVER
Figure 59. Lancaster Swirl Project sampling locations.
149
-------
The second sampling scheme utilized the cross-sectional
samplers Installed by EDP Technologies, Inc. The 36 In. (91.4 cm)
EOF Technologies Influent sampler was located on-line within the
building roughly one-fifth the distance from the diversion
chamber to the Swirl unit. The new EDP Swirl foul sewer sampler
was Installed on the 12 In. (30*5 cm) line after the dry weather
bypass connection and before the flow meter an Hydro-Brake.
Operation of this unit required dismantling the Manning sampler
previously used for foul flow sampling. All other samplers used
during the first phase were utilized thereafter. Influent sample
collection using both the Manning and EDP unl+i allowed for
comparison of the effectiveness of sampling.
Settling column samples were collected manually from both
the Swirl Concentrator and Swirl Degrttter. Clear samples were
collected at the center of the overflow pipe In each unit.
Influent samples for the Swirl were taken from various points In
the Influent channel. Degrltter Influent samples were taken from
a 2 In. (5 cm) valve at the Invert of the 12 In. \30.5 cm)
Influent pipe to the degrltter. Just downstream of the Hydro-
Brake.
Section 11.2.2. Initial Man I tor Ing ProbIams
In 1979 EDP was Invited to visit the Lancaster Swirl
facility to assist In upgrading the facility monitoring devices.
The site visit began with a review of sampling difficulties
encountered to date at the swirl facility.
Flow was measured at the Influents to the Swirl Concentrator
and the Swirl Degrltter via various pressurized pipe monitoring
devices. Quality measurements are taken at four locations
Including Influent and clear water effluent on the Swirl
Concentrator and Influent and clarified effluent from the Swirl
Degrltter.
Inspection of the existing sampling system yielded numerous
system faults. The piping was far too complicated and lengthy to
deliver "representative" samples to the Intakers of the Manning
6000 samplers. Sediment deposition was an obvious problem. More
Important was the fact that some of the sample streams were
subjected to passage through centrifugal pumps prior to actual
sampling. Solids passing through these pumps were subjected to
maseratlon.
The Swirl Degrltter Influent loop as constructed, drew from
a tap-In point after flow measurement and discharged Into the
city main sewer line, effectively forming a b/pass of all
subsequent operations. Due to the large size of the lines and
150
-------
pumps Involved a significant portion of the Influent to the Swirl
Degrltter would not reach the unit. Performance measurements
taken, that Is, total amount of grit produced, were "biased.
Flow measurement to the Swirl Degrltter was accomplished by*
means of a magnetic flow meter on the Influent I I no. The Hydro-
Brake used to control the Degrltter flow was located upstream of
the magnetic flow meter. Since the Hydro-Brake by design Induces
a radial flow pattern from >ts discharge, and since magnetic
flowmeters are designed for reasonably laminar flow. It was
unlikely that the flow readings measured were correct.
The Influent flow to the Swirl Concentrator Is gaged by two
primary flow meters on tl.e 36 In. (91.4 cm) Influent line. The
original Mapco unit has a problem of "zeroing" on high flov
regimes, undoubtedly due to excessive Influent flow velocities
transposing the sonic waves downstream of the receiver causing a
zero response. As a consequence an additional Leeds and Northrup
Doppler unit was Installed and It malfunctioned.
11 .2.3 SpaeI a I Izad EDP Automatle InfIuant Samp I Ing DevIeas
Implementation of the first set of adjustments recommended
by EDP was not predicted to completely alleviate the problems
with non-representative sampling. To obtain better sampling EDP
Technologies, Inc. fabricated and shop-tested two new sampling
devices to be placed on the Influent conduits to the swirl and
swirl degrltter. The purpose of the devices was to ensure
representative sampling of flow/solids stratification
characteristics, that Is, bedload, suspended load and wash load.
The devices were complete flanged assemblies.
•
Operation of Slide Samp I Ing Devlea
The slide sampling devices constructed for Installation on
both the 36 In. (91.4 cm) Swirl Concentrator and 12 In. (30.5 cm)
Swirl Degrltter Influent IInes took a cross-sectional si Ice of
the InMuent discharge during a sampling sequence. The sampler
consists of three main parts: the containment shell, the slide
cell and the sample bottle cassette. Figure 60 shows the details
of the 36 In. (91.4 cm) Influent sampler used for the lnflu»nt to
the Swirl Concentrator. The sampler for the Swirl Degrltter was
fabricated In exactly the same fashion.
ContaInmant She I I
The containment shell consists of a rectangular box
approximately three times as long as the given pipe diameter and
sllgh'ily larger than the pipe diameter In height. Welded to the
151
-------
iui* earro* V
«*
Figure 60. Diagram of 36 in. Influent Sampling Device I typical).
-------
rectangular containment box are a pair of pipe nipples and
flanges allotting for bolt-In Installation of i he unit.
Samp I i.-. i Cal I
The sampling slide cell Is a rectangular box that fits
inside of the containment shell sealed on four sides by a
neoprene face seal. Two rings of the pipe are welded Into the
si I de as shown on Figure 60. The Innermost ring Is open on Loth
sides allowing for Influent flow-through, wltn no restriction
during Intervals between samplings. The outermost ring or
sampling cell Is enclosed on born sides with the exception of a
slit extend Ing vertically on the upstream or InfIuent face one
sixth the width of the pipe diameter. The sampling cell Is moved
Inward and outtard by means of an air-cylinder, as shown. Samples
collected Jn the sampling cell will drain to the appropriate
sample bottl<» by means of a l-ln. (2.54 cm) pvc pipe line.
Samp I a BottIQ Cassette
The sample botlIe cassette will hold a total of 24 1-1
sample bottles, stacked on two rotary tiers. Each sampling cycle
will cause the rotary cassette to be indexed to the next bottle
position. Drainage to the sample bottles Is by gravity through
the 1-Inch (2.54 cm) sample line.
Photographs of installation are shown In Figures 61
through 64. EDP Technologies, Inc. personnel first removed the
bar screen In the diversion chamber and Inserted an Inflatlble
dam In the Swirl Influent line (see photograph A, Figure 61)
since there was no means for shutting flow to the Swirl. The 36
In. (?i.4 cm) Influent passed through the basement of the control
building near celling level. Photograph B, Figure 61 showsthe
cutting operation on the 36 In. (91.4 cm) Influent line. Insert
from page 9.
Photograph A, Figure 62 Illustrates Ine 36 In. (91.4 cm)
EDP Technologies sampler suspended over the Swirl Degrltter as It
was lowered through an opening In the laboratory floor to the
basement floor. The containment shell, flanged ends and air
cylinder are Included. The Swirl Degrltter Is 8 ft (2.4 m) In
diameter Indicating the magnitude of the sampler size.
Photographs in Figure 63 depict various views of the 36 In. (91.4
cm) sampler be'ng Installed.
PhotographA, Figure 64 illustrates the completed sampler
assembly for tl-e 12 In. (30.5 cm) pipe with the two flanged
connections. Attached to the containment shell Is the control
panel as seen on the right of the photo. Controls ailow selection
of sampling mode and delay Intervals and Include Indicator lights
153
-------
A. Inserting Inflatlble Dam tn Swirl Influent
B. Cutting Segment of 36 In. Swirl Influent
for placing New Sampler.
Figure 61. Photograph of EDP Technologies
Cross-Sectional Sampler installations.
154
-------
Reproduced from
betl available copy.
tn
en
A, EDP CROSS-SECTIONAL SWIRL INFLUENT
SEWER SAMPLER - LOWERED INTO BASEMENT
FLOOR (SWIRL DEGRITTER APPEARS BELOW)
B, EDP CROSS-SECTIONAL SWIRL
FOUL SEWER SAMPLER INSTALLED
Figure 62. Photographs of EDP Technologies, Inc.
Cross-Sectional Samplers.
-------
LT1
B. Partial View of 36 In. EDP Technologies Cross
Sectional Sampler
A. Partial View of 36 in. EDP Technologies Cross Sectional
Sampler (Control Assembly behind Engineer)
1 in - 2.54 cm
Figure 63. Photographs of EDP Technologies
Cross-Sectional Samplers.
-------
A, EDP CROSS-SECTIONAL SWIRL B. INNER ASSEMBLY EDP CROSS-SECTIONAL
FOUL SEWER SAMPLER - COMPLETED SWIRL FOUL SEWER SAMPLER
Figure 64. Photographs of EOF Technologies, Inc.
Cross Sectional Samplers.
-------
and pressure gages for operational setup and monitoring. The
containment shell of the 12 In. (30 cm) pipe sampler with the
cover plate removed is shown In photograph B, Figure 64. The
rectangular piece section Is the containment shell which encloses
the ovdi sampie cell slide. The sample cell Is the rectangular
piece within the oval section. The driving arm from the air
cylinder would be attached to the nut seen on the right.
Photograph B, Figure 62 displays the Swirl Degrltter
sampler, the 1.5 cfs (42.5 l/s) Hydro-Brake, the Dl scostralner
feed pump and pvc feed line and the Fisher-Porter magnetic flow
meter. Wastewater pases through the sampler, then the flow meter
followed by the Hydro-Brake and then Into the Degrltter (not
shown). 01scostralner pump feed Is taken from a port upstream of
the sampler. All components of the complete sampler are visible
In this photo Including the air cylinder attached to the left of
the containment shell, the control panel on top and the sample
cell drains at the base of the shell. The relative size of the 12
In. (30.5 cm) pipe sampler can be determined by reference to EUP
personnel In the three photos.
Section 11.2.4. Lancaster Swirl Evaluation Project Flow Heter
CalIbratton
METER PROBLEMS
Flow meters were Installed on the 3 ft (91.4 cm) diameter
Influent pipe to the Swirl Concentrator and on the 12 In. (30.5
cm) foul underflow leading to Swirl Degrltter. Flow Into the
facility Is determined by the combined sewage and stormwater
runoff volume of the drainage area while flow through the 12 In.
(30.5) line Is limited by a Hydro-Brake with a maximum rate of
1 .2 cfs (34 l/s).
The foul underflow meter Is a Fisher-Porter Magnetic
Meter which was laboratory-calibrated by the Alden Research
Laboratory. Field results Indicate the maximum flow rate through
the foul line to be that for which the Hydro-Brake was designed.
It was assumed that the meter was operating and recording
properly and field calibration was not required.
The 3 ft (91.4 cm) Influent pipe Is monitored by two flow
meters. One Is a Mapco (Modei BOOO) Nusonlcs Flowmeter which
relies on sound velocity measurement for flow rate determination.
For reasons unexplalnable by the manufacturer, this meter falls
to record flow until approximately 15 minutes after the
tranducers ara submerged. This period also corresponds to a time
of high flow reclines. It Is possible that excessive Influent flow
velocities transpose the sonic waves downstream o« the receiver
causing a zero response (although the manufacturer disagrees with
158
-------
this theory). The meter operational mode Is Indicated by flashing
light. However after a complete record of the light sequence by
DP and Lancaster personnel, the manufacturer could still not
trace the problem. As a consequence an additional Leeds and
Northrup, 475 Doppler Flo*meter was also Installed In early 1979.
This device seldom operated properly since Initial Installation
and was malfunctioned during all but one storm event during EDP's
evaluation. This meter was Inspected by the manufacturer's
representative In October 1980. It was removed* repaired and
reinstated In late November, 1980. Short term full-pipe low level
flow could be produced by backing up the swirl. Under this
condition the repaired unit appeared to operate properly.
During a storm event on Apr II 2, 1981 the Swirl level sensor
worked properly and both Influent meters operated. Due to the
historical problems with the flow meters EDP scrutinized the
chart ouputs and Influent flowmeter discrepancies were
discovered. It was noted that although the two meters disagree
sharply In absolute values, the flow patterns were somewhat
similar. Assuming that the Mapco meter was correct, a plot of
flow rates Indicated that the Doppler meter was probably
recording flows between 12 to 92 cfs (.42 to 3.25 m3/sec).
Doppler flow rate data based on an assumed span of 12-92 cfs are
plotted In Figures 65 and 66. The revised Doppler flow rates
agree with the Mapco flow rates, though It was still deemed
essential to check the calibration of both meters.
EDP performed + wo studies to rectify problems with the
Influent meters. Flr«.i, the Leeds and Northrup Doppler flow meter
calibration was to be checked by dye tracer studies. It was
believed that manufacturers' calibrations were not reliable.
Second, diversion chamber and swirl water levels were related to
flow rate through the 3 f1 (91 .4 cm) pipe.
During the next storm event on May 10, 1981, the L >ler
Influent meter which had operated for the first time during the
last event (Apr!1 1, 1981) failed again, and the Mapco unit
exhibited the typical delayed response period. During the next
event on the 15th of May, 1981 neither Influent meter operated.
As per the manufacturer's request the results of EDP's
observations were forwarded to Mapco with the hope of alleviating
further meter problems. Based on the continual failure of these
meters EDP developed a method for estimating Influent flow rates
based on the rate of Swirl tank fill and relative Swirl and
diversion chamber levels. This method was used whenever necessary
to roughly estimate flow data.
159
-------
N
B
cfs = 28.3 1/s
&
o
i
/\
n
3
14. 5
RECORD STARTING
ENDING
4 X 1 X 81
4/1/81
LAN POPPLER METER 0-150 CFS *
L&N OOPPLER METER 12-92 CFS >**
MAPCO METER 0-150 CFS
MAPCO METER FAILS TO OPERATE DURING
F\ftST 15 MINUTES OF EVERY EVENT
(FAILURE LASTED 30 MINUTES THIS EVENT)
15 15. 5
TIME OF DAY
•*•
-l-
18 ' 16.5
<24 HOUR CLOCK)
17
Figure 65. Influent flow record, Lancaster. 4/1/81.
-------
1 cfs - 28.3 1/s RECORD STARTING 4 / 1 / 81
ENDING 4/1/81
O
n
•n
en
\/
L&N OOPPLER METER 0-150 CFS*
L&N DOPPLER METER 12- 92 CFS**
NAPCO METER 0- 1 SO CFS
Actual recording (span 0-150 cfs)
Arbitrary meter span adjustment to
12-92 cfs range
•*•
10. 5
17 17. 5
TIME OF DAY
18
<24 HOUR CLOCK>
18. 5
1Q
Figure 66. Influent flow record 4/1/81 (continuation), Lancaster,
-------
EMPIRICAL RELATIONSHIPS
An Initial attempt was made to relate Swirl Influent flow to
clear and foul flow rates. Accurate determination of the clear
water flow rate proved to be unreliable due to:
I) The large diameter of the Swirl clear water weir,
allowing a substantial change in flow with a slight change In
depth over the weir.
2) The Swirl weir Is not level all around. Attempts to
level the weir were not completely successful. Although final
position measurements were taken after adjustments It was
expected that warpage would continually occur rendering the use
of the circular weir equation of questionable accuracy.
3) The flow over the clear weir does not directly relate
to the Instantaneous Inflow since there Is some lag time as the
tank level Increases. Influent rate Is related to the change In
tank depth as well as the actual level.
The next attempt was to assume the system consisted of
two tanks, that Is, the diversion chamber and the swirl connected
by a pipe acting as a syphon. Basic hydraulics would dictate that
the flow between the two units would be directly related to the
differences In water levels.
A plot of measured flow versus Swirl levels (as shown
In Figure 67) Indicated a range of flow for any given Swirl
level. This range Is probably due to the fact that the liquid
within the t.ank Is not quiescent but flowing freely. Inspection
of Figure 67 Indicates that the head differential between the
diversion chamber and Swirl liquid levels accounts for much of
the variation cited above. The data for this plot were taken from
the liquid love' records and the Mapco meter Influent flow record
for the storm of 10/2/80.
Multiple I I near regression analysis of these data (29
points) taken directly as flow, Q, (cfs) swirl level (S.L.) (ft)
and diversion chamber level (O.L.) (ft) resulted In the equatlont
Q • 11.234 x S.L.+ 34.162 x D.L. - 302
The coefficient of determination R2 Is 0.983. The
correlation between actual Mapco flow and the calculated flows Is
good. There are several limitations as to the application of this
equation. First, the equation Is functional only for times whan
the Swirl level Is above Hie clear overflow weir level of 6.2 ft
(189 cm). Below this level The equation Is Inaccurate, however
tank rate fill calculations can be used. Second, no Mapco flow
162
-------
LANCASTER SWIRL INFLUENT FLOW EVALUATION
IO-2~8O DATA
TO
60
fl? 50
O
0
u.
o
2
20
K>
^••MM
— ^— .
J
HUM
Kn
•ORE
t^1
ERS
rEEN
rw
«•!
X
KMC/I
IYER)
CONS'
4"
I
re HE
ION
TOT
1
^
^
^-*
GMT
«AMI
DTFF!
X
ia
^
HFFE
EM A
RENT
I
1
>
^
^-^
tENTI
O M
wn
^x"
4^
"f
VFL
Wl
0.
in.
WWN
I
x^
^^^
W E
ML4
e
^
2
/
3UAT1
^
i
X
3N 1
-------
data was available for flow rates less than 4.5 cfs (127.4 l/s)
and the resul+^pt equation tends to inaccurately Indicate flows
below this level. Since most of the time storm flow Is above 5
cfs (141.6 l/s) this created no problem. Any flows calculated to
be below this level using the equation were manually reviewed
using tabular data. The raw data were selected from a portion of
the storm level of 10/2/81 when the two level sensors and the
Mapco meter were all operating. During this time the diversion
chamber level varied by 2 ft (61 cm) and the Swirl level varied
by 0.7 ft (21.3 cm).
It was deslreable to determine a relationship between the
Swirl and diversion chamber levels which can be used to determine
the Influent flow rate for all storm events. Data from a second
storm even? (11 points) were evaluated In this regard. For the
event of 5/15/81 the equation developed was:
Q = 15.3xS.L. + SIxD.L. - 230.5 R2 = .981
Although a different equation resulted, as expected* the
correlation was again high Indicating a reasonable relationship
between Swirl and diversion chamber levels and the flow rate.
Figure 68 shows comparison of measured and predicted flows for
the 5/15/81 event.
Although good correlation between Swirl levels diversion
chamber levels Influent flow rates exist as seen by the events of
10/2/81 and 5/15/81, the equation developed for either data set
could not represent the flow rates during the other event. This
difference Is shown In Figure 69 wherein the equation generated
using the 5/15/81 data was used to estimate flow levels for the
10/2/80 event. A positive overestimate of 25 cfs (708 l/s)
results. The relationships are dependent upon Swirl and diversion
chamber levels as Indicated on the level sensor records. These
records are subject to three possible errors: a) the field
calculation of the sensing unit the possibility of a slight
mI sadJustment of the zero or span settings); b) strip chart
recorder; and c) Interpretation of the strip chart lines.
The discrepancy In flow values cited above Is probably due
to the IB Isrecordl ng of liquid levels rather than In the equations
developed. The following analysis Indicates thai this Is the
case. First, a review of data for the two events was made to
compare data points with similar values for the Swirl level and
actual flow rate. Results were as follows:
164
-------
LANCASTER SWIRL INFLUENT FLOW EVALUATION
9-15-81 DATA
TO
60
e
o
»
5
Ul
o
u. 40
a
30
20
K>
/
7
/
X
o
/
E
. 0
7
UIVA
/
EMCi
7
—
UHE
x
i cfs » 2&.3 i/s Mapco Flow CFS (no data btlow I8CFS)
Figure 68. Plot of Mapco meter flow vs estimated flow (5/15/81).
165
-------
Storm of
10/2/80
Storm of
5/15/81
Q (cfs) Swirl Dlv. Cham.
Level (ft) Level (ft)
Q (cfs) Swirl Dlv. Cham.
Level (ft) ~ Level (ft)
39.6
36.7
19.6
6.6
6.6
6.4
7.9
7.8
7.3
41.0
36.1
18.6
6.6
6.6
6.4
7.3
7.2
6.8
The difference In diversion chamber values for near
equivalent swirl levels and flow rates was consistently around
0.5 ft (15.2 cm). The bias noted above for Figure 69 disappears If
the diversion chamber levels of the 10/2/80 storm event were
reduced by 0.5 feet (15.2 cm) and then used In the second
(5/15/81 data) equation to estimate flow values. It Is clear that
with this correction flow values calculated agree with the actual
values quite closely.
It appears that the probable error Is In the diversion
chamber level record. Of the three error sources cited above,
setting of the recorder within the olverslon chamber Is most
likely. The recorded data were reviewed several times to minimize
any manual Interpretation error and 0.5 ft (15 cm) Is quite large
to be a fieid calculation error. As previously staged, the strip
chart resolution Is quite poor with an Incremental value of 0.4
ft (12.2 cm) per line. A minor adjustment error would easily
produce a 0.5 ft (15 cm) discrepancy. In summary, the method of
determining flow -ates from relative swirl and diversion chamber
liquid levels was reasonably effective. This empirical procedure
was roughly ca'lbrated using dye/tracer methods.
Dye GalIbratIon of the SwlrI Flow Monitor Ing System
Dye studies were performed during five events at the
Lancaster Swirl site. Evaluation consisted of adding dye slurry
to the Swirl Influent at the diversion chamber. The concentration
and fI or rate of the dye slurry were known and samples were
collected as the flow entered the Swirl. The dye Injection system
contained a circular ring with eight Injection ports designed to
allow the added dye to penetrate and mix with the entire waste
stream. With the additional mixing In the Influent pipe It was
assumed that complete mix would be accomplished. Dye
concentrations within the Influent samples were determined and
continuity considerations were used to calculate instantaneous
flow rates.
Using this system dye calibration studies were performed
during five storm events Including, July 2nd, 20th am ana pm,
28th and August 3rd, 1981. Neither the Mapco flow meter nor the
166
-------
LANCASTER SWIRL INFLUENT FLOW EVALUATION
A - raw 5/1 5/80 data
10-2-80 DATA* • - adjusted 5/15/80 data
•
P «/v
fe 50
•
III
J *°
il
0
30
20-^
10-
Figure
*
f
/y
— w
«0
^
q
/
/ A
*
— -
/
/
^
*
A
e<
7
EQUI\
4
k/«
ALENi
/
A
EU*
/
* K> 20 30 40 50 60
Mapco Flow CF5
69. Plot of adjusted Mapco meter flow vs estimated flow (10/2/80). 1
167
-------
Leeds and Northrup flow meter were operational during any of
these events. Results of flow rate calculations were compared to
results of 11 ow *st I ma+es h=»sed OP relative swir! ano diversion
chamber liquid levels. Flow estimates were based on the re-cults
of earlier analysis comparing these liquid levels to Mapco meter
flow records.
During this period the primary emphasis during a storm event
was to manually collect sam pies for sett Iedb!I Ity analysis and
since the dya study system also required manual sampling,
dye/tracer experiments were typically not begun until late In an
event. At that point flows had usually diminished and the Swirl
and diversion chambar liquid levels ware low. Although a series
of samples were collectad over a period of time the change In
flow rate was small and relatively slmlla^ for each event
(probably due to the backwater and downdraw characteristics of
the drainage area). The net result was that only d small range of
flow could be calibrated.
3ost.d on these calibrations t!ia flow rates as estimated by
the liquid level method (and the Mapco meter readings on which
they were originally based) were within a reasonable range of the
actual flow as measured by dye experiments. For example,the How
estimates ranged from 6.2 to 9.5 cfs (175 to 269 l/s) fur a
seventeen minute period on the July 28, 1981 event while the
calibrated flow changed from 5.4 to 9.6 cfs (153 1o ""72 :/s).
Considering the errors entailed In strip chart evaluation .*d .• -e
addition/concentration analysis techniques, rhese values -.r-a
vIrtual ly equal.
Summary
Influent flow records at Lancaster were reviewed foi
accuracy by three methods.
1) Mapco flow records were usually available, but en Iy
after the first 15 minutes of each event. This period usually
covered tiie time of high flow. For one rtorm event ftow
records were produced by the Doppler me*3r. Flow data could be
made to agree using arbitrary adjustment assuming the Mapco
values to be correct.
2) ?4apco flow values were compared to relative swirl and
diversion chamber liquid level and to overflow based on the
circular wulr equation. The circular weir equation failed to
demonstrate agreement with Mapco values. Fo>- any given Swtrl
liquid level the weir equation yields a constant value while
•ihe diversion chamber level and Map^o values varied.
168
-------
For one event an equation was developed which
accurately calculated Mapco flow values based on Swirl and
diversion chamber levels. For other events It was necessary to
adjust the equation to produce agreement wl+h Mapco values.
This adjustment consisted of a slight 'change ?n the recorded
diversion chamber level. For each event the amount of
charqe varied typically up to 0.5 ft (15.2 cm). The
need for this change was due mainly to the poor resolution of
the diversion chamber level chart where each minor scale
diversion represented 0.4 ft (12.2 cm). As with method 1 above
the repeatability of data comparison Indicates good precision
but does not ensure the accuracy.
3. To determine the accuracy of flow records a dye
Injection system was used to determine flow rate. The Intent
was to use the system at several flow rates during an event
and compare results to Mapco values. Even a single point would
have sufficed to determine accuracy. Unfortunately the Mapco
meter was non-operational during dye study testing. In
addition, dye study results, although Including many samples,
were all within a narrow range of flow. Adjustments could be
made to liquid levels to yield agreement at the one flow point
but no additional points were obtained.
The precision of the Mapco flow records seems reliable
based on the one-time agreement with adjusted Doppler values.
The accuracy was never adequately tested. In sum, extensive
efforts were expended trying without success to make operational
the flow meters and at the same time develop reasonable flow
calibrations. Unfortunately, all Influent flow values cited In
Chapter 12 must be viewed as reasonable "ball-park" estimates.
Section 11.3 Instrumentation Overview - West Roxbury
Data I I s of Samp I ! nq Loeat-l OTIS
The Initial sampling program at H. Roxbury site consisted of
six hand-sampling locations noted on Figure 70. Samples were
collected during storm events throughout the evaluation period
and analyzed for solids concentration and COD.
i
Further Into the evaluation period Manning automatic
samplers 6000 series were used to collect foul samples from both
units. (See Figure 70 for sampler locations). These samplers
were Installed to monitor the foul flow before It entered the
discharge line where sedimentation could take place.
169
-------
-V-
CONTROL H
BUILDINGN
*
A
KEY
SAMPLE SIGHT (HAND)
MANNING AUTOMATIC SAMPLERS
HAND SAMPLE LATER REPLACED
BY MANNING
SWIRL
SWIRL CLEA
-HYDROBRAKE
BYPASS
FOUL FLOk
SUMP TAW
1 In. = 2.54 cm
-*-
-*-
27
-------
Samp la Col I act Ion Methods
During the sampling program both automated and manual
techniques were employed. Since EDP field crews were able to be
on-slte during most of the runoff events, grab sampling
techniques were used for those events.
Initially Influent samples were taken manually In one I
wide-mouth plastic bottles for both the Swirl Concentratcr and
Helical Bend units. Manning samplers were later Installed for
Influent sample collection but were found to obtain the same
quality of sample as manual procedures so they were removed.
Typl'cally, samples were collected from the start of the event
until unit shut down In 4 minute Intervals.
All clear water samples from both the Swirl Concentrator ana
the Helical Bend units were taken manually during each event.
Samples were collected at the same 4 minute Intervals until unit
shut down when flow over the clear overflow weirs had ceased.
Foul sewer samples originally taken by hand from the units
at the outlet port showed unusually low solids concentrations
which prompted the Installation of automatic Manning samplers
prior to the discharge line. Concentrations of solids from these
samples were much higher Indicating that a considerable amount of
solids were settling within the discharge line. The Manning
samplers were set at 3.75 minute Intervals and were Initiated at
the start <•»? the event and continued sampling after unit shutdown
throughout- the drawdown period. During the first minutes of tank
drawdown the samplers were triggered manually at shorter time
Intervals.
Several tank samples were collected during storm events from
each unit. A specially designed sample bottle holding device was
constructed to obtain cross-sectional or single depth samples.
Initially, a sample bottle was loaded In the device and the
trigger released tightly sealing the bottle. The bottle was
lowered In the tank to the appropriate depths(s) and the trigger
mechanism pulled, allowing the bottle to fill. Release of the
trigger mechanism would close off the sample bottle.
Settling column samples were collected In four gal (15.1 I)
buckets (8 for each test) from Influent, clear and foul flows.
The majority of the samples were collected at the start of the
storm event. Influent samples were taken In the Inlet channel
downstream of the flow monitoring transducer and prior to the
Inlet Into the units. Clear water samples were collected In the
overflow pipe so as to achieve an equal mix of flow passing over
the entire clear overflow. Foul sewer samples were taken during
171
-------
events early In the evaluation period but were discontinued when
It was discovered that foul solids were settling In the discharge
I Ines of the units.
Monitoring of precipitation was conducted using an automated
recording rain gauge located on-slte.
11.4 Proeadura for Patarm Inat I on of SattIaafaI I Ity Character I sties
A major aspect of the Swirl evaluation was comparison of
solids concentrations and settling velocity characteristics of
the material suspended In the Influent and clearwater
wastestreams. Various types of settling column designs and
operational procedures were received by EDP In a previous study
to develop a procedure for determining CSO and stormwater
settleabllIty characteristics. (33) Pertinent details of that
report and final results are Included In section 11.4.1.An on-
slte settling column was used In a' limited fashion In Lancaster
and Is described In section 11.4.2.
11 .4.1 . Sett I I ng Col umn Das Igti
Three design parameters emerged from EDP's study (33) as the
prime design determinants for the new envisioned settling
column/procedure* These parameters Included settling column
dimensions* mixing capabilities and sampling Intake systems*
The first design consideration was the size of the settling
column diameter relative to the quantity of sample drawn from the
column at each prescribed time Interval. A related consideration
Is the time necessary to fill the drawn s am pie. The I ssu& centers
on the comparison of rate of settling column water surface
drawdown relative to settling particle velocities.
The second design parameter Involved standardizing column
mixing mechanisms and procedures. Typically, column mixing has
been accomplished by manual efforts such as hand-operated
rotation and stirring or by use of compressed air. These methods
create undeslreable characteristics Including solid degradation,
floculatlon, agglomeration. Incomplete mixing and generation of
eddies and currents. The settling column design therefore focused
on a method of consistent controlled mixing aimed at minimizing
the aforementioned problems. The design chosen encompassed
steady state, bl-dlrectlonal (axlal/longltudlnal) rotation. Two
electric motors «lth variable speed controllers provided
uniformity and flexibility to the mixing process. Axial rotation
creates a uniform distribution of solids across a cross sectional
area of the column. The longitudinal rotation Is the mechanism by
which the solids would be dispersed throughout tho horizontal
axis of the column.
172
-------
The third design parameter dealt with the sampling oort
mechanisms. A push-rod type sampling Intake device was cl.o'sen to
minimize Interference with settling particles. Nine sampling
ports were Included at six In. (15.24 cm) Intervals, alternating
at 180 degrees. Intake tub'ng with a diameter of 0.40 In. (1.02
cm) was used to assure large solid sampling capabilities and
minimal sampling time. Each of the nine Intake asemblles was
designed to draw a sample from the center of the column and then
be withdrawn to a flush seal on the side of the column. The
discharge side of the Intake was attached to a permanent threaded
cap acting as a positive fastening mechanism for the sample
bottles.
Sett I I ny Co-i uffi.n
Front and side mechanical views of the settling column are
presented In Figures 71-A and 71-B. A photograph of the unit Is
shown In Figure 72-A. The base cylinder Is 0.26 In. (0.66 cm)
thick-walled tempered acrylic measuring 6 ft (1.83 cm) length
with an Inside diameter of 11 In. (27.94 cm). Constructions
performed on the base column Included nine sliding push-rod
sapling ports, three mixing stators, and a base with a drain
valve.
Each sampling port consists of 0.43 In. (1.09 cm) diameter
Intake tube, a water-tight flush Intake seat on the Inside of the
column and the threaded cap attached to the discharge end of the
tube to allow boltles to be positioned prior to the required
sampling time. The bottom plate was also constructed of acrylic
and fltt9d with a 2-ln. (5.08 cm) drain valve to facilitate
sample removal. The top seal was fabricated by fitting a piece of
compressible rubber between two aluminum plates designed to be
hand-tightened together, thus forming a compression seal around
the perimeter (see Figure 73-A).
To support the axial rotation assembly (see Figure 73-B) an
lnn«i framework was constructed of light gauge aluminum, capable
of supporting one drive motor. This frame then was fastened Into
a support structure which besides elevating the settling column,
supported the IongtItudInaI rotary drive motor and provided a
mounting for the speed controllers. The two drl«e motors selected
were capable of supplying a maximum of 5 and 7 rpm for
longitudinal and axial rotation, respectively. A photograph of
the two speed controllers are shown In Figure 72-B.
173
-------
Front wit* «f iU« coluM.
i
b. FroM/lHi *1wt of ttttllng M!MM.
Figurt 71. Front and side views of motorized settling column.
1 In. * 2.54':cm.
-------
A. Settling column
Figure 72. Photographs of motorized settling column.
B. Axial and longitudinal motorized assembly
controllers.
-------
k. /
^^••fr -1_ * -" "f • ^T SJPlk. «,; '
A. Hand-tightened top water seal
B. Lower portion of axial rotation assembly.
Figure 73. Photographs of settling column mechanical details
176
-------
Sett I 1 ng Col umn Oparat lona I Data I I s
Air-tight sample centalners were com pietely filled la the
field and capped carefully so as to minimize air entrapment. The
reason for this procedure vas to reduce the potential for
breakdown of large organic solids. All containers collected at
the West Roxbury site were then packed tightly together In the
transport vehicle to prevent excessive movement. Containers from
the Lancaster project were packed In cartons and alr-frelghted to
Boston for testing. Quality was an unknown factor.
Once the samples were poured Into the settling column a
uniform procedure was developed to assure complete mixing and to
minimize solids breakdown. The axial motion was first activated
and set at a rotational speed of 5 rpm. Longitudinal rotation was
set at a speed of 3 rpm. Mixing was continued for three full
minutes. During the mixing procedure, all sampling port Intakes
were*' thdrawn from the column Interior so as not to Interfere
with the suspended solids.
Immediately following cessation of co'umn rotation, the
first three samples were taken. Sam pies were drawn by sliding the
Intake apparatus Into the center of the column, allowing the
sample bottles to fill and withdrawing tjie Intake following
completion (see Figure 74). Samples were drawn Into uniform 350
ml bottles and were capped Immediately following removal from the
settling column. All botMes were labeled consecutively. After
each set of samples were taken, the next three bottles were
screwed Into the sampling ports and the liquid level In the
column was recorded for future use In computing settling
velocities.
The column sampl Ing Intervals were determined to begin at
time zero, followed by samples taken after 30 seconds, 1, 2, 5,
15, 30, 60 and 120 minutes from time zero. Samples were always
taken at the one ft (30.5 cm), three ft (91.5 cm) and the five ft
(153 cm) column elevation.
No attempt was made IP any of the settl Ing column tests to
distinguish between rise Cloatables) and fall (settIeabIes)
particle velocity distributions. Recent Investigations at the
Water Research Centre at Edinburgh, Scotland (34) devised methods
and apparatus for separating particles Into different rise or
fall velocli-y groups. The settling column results presented In
Chapters 12 and 13 represent gross overall particles settling
characteristics Independent of relarlve distributions of rise vs
fall type particles. If a given sample tested had a significant
fraction of rise (floatable) type materials then the resulting
column test would bias fall velocities for settleable materials.
177
-------
r
00
Figure 74.Photographs of settling column sampling operation.
-------
11.4.2 Lancaster Sattl Ir.g Col limn
The Lancaster settling column was a stationary plexiglass
unit 6 In. (15.2 cm) In diameter 70 In. (178 cm) tall with a
total volume of 8.5 gal (32 I). The unit was not f I I I ed to the top
to avoid spillage. The volume of sample used was 8 gal (30 I).
Sampling pointsvere located at 1 ft (30.5 cm) spaclngs on
the side of the column, flush with the Inner wall and a valve on
the outer tube end. Sam pi Ing consisted of turning the valve one-
quarter turn, allowing drainage collection of about 100 ml of
sample.
Since the settling col urnn was stationary, Initial mixing of
the samp I e wasaccomp I I shed by aeration at the base of the column.
ThIswasaccompl I shed by attaching a diffusion stone to plastic
tubing and lowering the stone Into the column from the top. The
stone =ould be rotated around the base of column by twistinq the
plast'c tube to maximize the mixing operation.
The use of this aeration technique was reviewed In the EOF
report (33) and found to be not as effective as the bi-
directional mechanical mixing technique. The sampling ports,
being flush with the Inner wall, tend to collect samples from
areas subject to wall effects. In addition, the column diameter
affected the relative volume of the unit such that each sampling
sequence lowered the water surface substantially and created
substantial motion within the column.
11.5 Analytical Methodology
Sol Ids Analysis (W. Roxbury and Lancaster)
Samples were analyzed for four solids fractions: suspended
solids, volatile suspended solids, settleable solids and volatile
settleable solids. The methods utilized for suspended and
settleable solids determination were 2080 (Total NonfMterable
Residue Dried) and a revised form of 208F (settleable matter)
from Standiird Methods. (34) The revised procedure consisted of
several stirrings of the sol !ds In an Imhoff cone as opposed to a
single mix at the 45 minute point. Settleable solids were
recorded at the end of the one hour settling time and 200 ml of
supernata.it here siphoned from the center of the cone midway
between the surface of settled sludge and the liquid surface. The
use of this procedure enabled the calculation of the weight of
settleable matter as well as a visual reading of the volume of
settleable solids within each sample. Determination of volatile
suspended and settleable solids were made following Standard
Methods Procedure 208E (Totai Volatile and Fixed Residue at 550
(35).
179
-------
Samples taken from settling column tests were analyzed for
suspended and vc!at!!« suspended soiios according to proceoures
208D and 208E discussed above.
Bacterial levels of stormwater samples at West Roxbury
facility here noted as Tota! Col I form using mI I I I pore filter
technique (35).
Occasional sieve analysis of tank sediment samples and
settling column sediment samples were also performed to determine
grain size distribution of the material as well as organic and
Inorganic content. Each sediment sample so collected was air-
dried, weighed and then a slave analysis conducted. The sieve
series employed and related grain size nomenclature were the
following:
Standard Slave Number S Iza QpanIng
In. mm
8 0.0937 2.36
16 0.0469 1.18
30 0.0234 0.600
50 0.0117 0.300
100 0.0059 0.150
200 0.0029 0.075
Subseqjent to the sieve analysis, each sieved fraction was fired
to conduct a percent volatile analysis.
The COD values of samples collected were determined using
the U.S. EPA approved ampule COO Digestion and Analysis
System.(36) This method requires a 2.5 ml volume of sample to be
digested In the ampule at 150 degrees centigrade for two hours*
The ampule Is placed Into a spectrophometer and analyzed. Each
series of samples ana Iys I s was preceded by the preparation of a
set of standards.
11.6 Elf I eleney CaleuI at Ion Proead iras
This section describes various mecns of computing the
treatment efficiency and removal rates for the Swirl Concentrator
and Helical Bend Regulator units evaluated In this study. A unit
which removes 50% of the Influent solids mass and operates at 10
% foul underflow Is obviously much more efficient than a unit
removing 50% of the Influnnt mass *hlle operating at 30? foul
underflow.
160
-------
First, the removal rate was defined as the percentage of
Influent mass contained In the foul underflow line. Second*
efficiency vas defined as the removal rate minus the percentage
of flow In the foul underflow line. A unit operating at 10 % foul
underflow containing 10$ of the Influent solids mass would have
0$ efficiency since It acts merely as a flow - splitting device.
The rate of removal would be 10$. Removal rates are not flow
related and can be used to match design criteria for total
removals. Efficiency val ues were used to optimize the unit design.
Examples if removal rate and efficiency determination
follow* As can be seen It Is not always necessary to determine
the mass rates. An Influent flow rate of 20 cfs (566.4 l/s) with
10? foul underflow wMI be used In all cases.
Example ^
Given an Influenl solids concentration, I, of 150 mg/l and z
clear solids concentration, C, of 100 mg/l the efficiency equals
fie normalized decrease In solids concentration times the
percentage of clear flow, CF, as follows:
Efficiency, E • (I-O/I) x CF/100
E • (150-100)7150) *90/100 - 30$
Removal rate, RR • efficiency plus percentage foul underflow
rate, FF, that Is, 30? + 10? • 40$.
Example 2,
(Same determination as example I using mass
calculations). Assume 20 cfs (556.4 l/s) Influent and 2 cfs (56.6
l/s) foul flowi
Influent mass IM • 150 x 556.4-56.6 • 84960 mg/s
clear mass CM • 100 x 566.4-56.6 • 50976 mg/s
foul mass FM - IM - CM - 33784 mg/s
RR- FM/IM x 100 • 40$
E • removal rate - foul flow rate » 30$
Example i
Given the Influent solids concentration of 150 mg/l and
a foul concentration of 600 mg/l. Influent mass IM» 84960 mg/s
and foul mass, FM • 56.64 l/s x 600 mg/l » 339C4 mg/s. As before
RR Is 40$ while E • 30$.
181
-------
Examp la £
A simplified version of removal rate determination
using foul flow solids concentrat'on Is possible by defining a
foul flow concentration factor, ffc, as the ratio of the foul
flow concentration, F, to the Influent concentration, I, that Is,
ffc • F/l. Removal rate Is the foul flow concentration factor
times the foul flov percentage, RR«ffc x FF
For the above data ffc - 600/150 -4
RR - 4 x 10 - 40*
and E Is RR minus the percent foul flow or 30*
11.7 Sett laab-l I I ty Curve Man I pu I at t ong
The following two plots are examples of various graphical
ways of presenting settling column results. The example used Is
for the storm event occurring In Lancaster on 7/2/81. Samples
from the Swirl Influent and clear were taken and settling column
tests performed. Figure 75 displays the original data as
concentration remaining as a function of settling velocity.
Figure 76 Is the standard plot of percent remaining as a function
of settling velocity. This plot allows the determination of
fractions of the material In the sample with settling velocities
within a cerlaln range.
In the case of efficiency evaluations of treatment
processes, the efficiency (treatment removal) percentage Is
deslreable for various sett IeablI Ihy ranges. The standard removal
plot requires that the change In percent remaining for eech
sample be determined over the chosen settling velocity range.
These percentages must then be converted to a concentration
basis, compared, and the removal percentage calculated.
From Figure 76 the procedure Is as follows: for a settling
velocity range of 1.0 to 0.1 cm/sec (0.03 - 0.003 ft/s) the
Influent covers the range of 100* to 62*. This 38* difference
corresponds to a change of 67 mg/l based on tho Initial Influent
concentration of 177 mg/l. The clear sample covers the range of
78* to 44* remaining, the difference of 34* corresponding to 51
mg/l based OP the Initial clear concentration of 150 mg/l.
Determination of percent removal next requires subtracting the
relative solids values In this range yielding 16 mg/l and then
dividing by the Influent concentration change within the range
yielding 24* removal In the range or by the Initial concentration
to determine overall removal wlt.iln this range. I.e. 16/177 or
9*.
182
-------
m^.:
H
t-t
Z
o
m- ..
o*
n
H
o ••
3 ••
\ •-
0 •
Influent Ave. Cone. Time (0) 177 mg/1
Clear Ave. Cone. Time (0) 150 mg/1
INFLUENT
« CLEAR
-I-
1 cm/s =0.033 ft/s
•+•
S 0
50
100 150 200
SS CONCENTRATION Cmg/L>
250
Figure 75.
Settling Column results, Lancaster (7/2/81).
-------
00
U)
m
-4
r
M
z
o
m- ..
o1"
n
M
H -r
0 -•
^. INFLUENT
« CLEAR
1 cm/s -O033 ft/s
S0
20
•+•
•+•
->-
-I-
40 60 80 100
SS PERCENT REMAINING
-»-
-I-
120
140
Figure 76.
Suspended solids remaining vs settling velocity. Lancaster Swirl (7/2/81).
-------
This lengthy process can be simplified by plotting the data
tilth the clear % remaining values normalized to the Initial
Influent concentration. All clear percentage valus are determined
as mg/l remaining divided by the Initial Influent concentration.
The resulting plot Is exhibited In Figure 77. The procedure
for determining removals using this plot begins as before by
determining % remaining changes over the desired settling
velocity range. For the range between 1.0 and 0.1 cm/sec the
values are 38£ for the Influent, as before, and (69-36) 29% for
the clear. Since the values are normalized to the Influent
concentration -Wie overall removal Is simply the difference or 9%
as before and the removal within the range Is (9/38) 24$, also as
previously found but much more easily determined. Normalized
plots allow a significantly easier method of efficiency
determination and were utilized for data review.
185
-------
GO
Ot
+
-»-
INFLUENT
CLEAR
(Normalized to Influent)
1 cm/s -OJ033
60 80 100
PERCENT REMAINING
120
•*•
-t
140
Figure 77. Normalized Suspended solids remaining vs settling velocity, Lancaster Swirl (7/2/81).
-------
CHAPTER 12
Lancaster Sw I r I Eva Iuat I on
12.1 Foreword
This chapter details the evaluation results of the
Lancaster Swirl Regulator/Concentrator, Swirl Degrltter and the
DIscostraIner. A chronology of events Is presented In section
12.2. A short description of each sampling event during the EOF
fall, 1980 to summer, 1981 evaluation program Is given In section
12.3. A comparative assessment of Swirl Concentrator/Regulation
Influent suspended solids concentrations obtained for a number of
storm events using the new EDP Technologies Cross Sectional
sampler vs conventional application of Manning sampler Is
presented In section 12.4. These resulrs Indicate that early
conclusions regarding poor performance of the Swirl facility
during the 1979 period (pro EDP Involvement) were Incorrect.
These comparative results also have a bearing on fall, 1980
evaluation program. The new samplers had not been Installed as of
this point. Detailed descriptions of a selected set of 1980-1981
sampling events are presented for the Swirl
Concentrator/Regulation In section 12.5. The comparative Influent
sampler results are used In section 12.6. to reassess several
storms where performance was seemingly poor. Ancillary Swirl
Concentrator sampling results are presented In Section 12.7.
Sett Ieabfifty results for the Swirl Concentrator are presented In
section 12.8. An overview performance summary for the Swirl
Concentrator Is presented In section 12.9. Section 12.10 and
12.11 discuss the operation of the Swirl Degrltrer and
Dlscostralner respectively. The performance of the Hydro-Brake
located on the Swirl foul sewer Is discussed In Section 12.12.
Section 12.2. Summary of Lancaster Evaluation
The following summary highlights the major activity
during the evaluation program conducted by EDP.
1/11/80 Swirl Influent sett IeabI I Ity experiment performed
7/31/80 Short storm event/operation/no sampling
8/2/80 Facility checked out and placed In automatic operation
mode
187
-------
8/10/80 Storm event sampiad. Clear Swirl water sampler non-
operational (manual sampling performed)
8/30/80 Light rain / no operation
9/2/80 Light rain / no operation
9/10/80 Storm event sampled
W14/80 Storm event sampled
10/2/80 Storm event sampled
10/21/80 Nev EDP Tochno logy samplers Installed at site
12/23/80 Second level sensor for Swirl Installed
2/2/81 Storm event / no operation
2/4/81 Site checked out. Non-operation of 2/2 event due to
failure of Swirl level sensor
2/9/81 Swirl Level sensor removed and sant to manufacturer
3/9/81 Swirl Level sensor returned
3/12/81 Swirl Level sensor re-Installed, all set for next storm,
4/1/81 Storm event sampled/sampler problems
4/2/81 Dye study equipment arrived
4/14/81 Light rain/ no sampling
4/22-2V Visit by EDP to adjust samp I ers and set up dye
81 calibration equipment. Discovered problems with
Swirl level sensor explaining results of 4/1
event
4/27/81 Light rain/no samples
4/29/81 Level sensor adjusted
5/6-7/81 Visit by EDH personnel to adjust samplers. Found
electronic controls burnt-out and heavy solids
blocking EDP sampler feed lines.
5/10/81 Storm event sampled
ECP Influent sampler plugged
EDP foul sampler partial plugged
188
-------
5/15/81
5/27/81
6/1/81
6/2/81
Clear sampler not operational/no samples
Dcppiur merer not operational
Light rain/ no samples/ Hapco meter observed
Backwash system recommended by EOF on 5/7
Installed on Swirl Dagrltter
Doppler meter recal'brated
Site operator called Hapco with 5/15 results
Hapco reporrs no Idea as to cause of problems
Storm Event - sampled
EDP Influent sampler clogged/ no samples
EOF foul sampler functioned correctly
Hannlng Clear samples 8 empty-low flow
Ooppler unit not operational
reported
6/17/81 EOF calls Hapco a.id L&N about flow meters
6/19/81 L&N rep. at site to repair motor
6/22/81 New Operator prepares site for event
Storm event
EOF samplers worked well
Hannlrg clear sampler malfunctions
6/25/81 EDP conversation with Mapco rep/ repair requisition
given to Hapco.
7/1/81 Storm event sampled Swirl Oegrltter Inoperative
7/20/81 Two storm events sampled. Basement flooded during first
storm "shorting" DegrItter sampler
7/28/81 Storm event sampled for only sett IeabI I Ity
cnaracterIstlcs
Section 12.3. Description of Each Sampling Event
The following section Includes an overview summary of
each sampling event and analysis data. Performance data Is
briefly described. The four events monitored In 1980 occurred
af+er Instituting the first set of Instrumentation changes EOF
recommended after observing the cross-site In 1979. The second
set of changes, that Is, new EDP Technologies Cross-Sectional
samplers were Installed In early 1981. A detailed discussion of
a selected set of events Is Included In section 12.5.
189
-------
Lancaster SK I r I 8/1 Q/8Q Storm Event Summary
The first storm event monitored after the facility
was shutdown In early I960 ocurred on 8/10/80. The storm was
characterized by a moderate thunder shower followed by
steady light rain. The antecedent period was hot and dry
for about one week.
Samples were collected at four locations during the
event Including the Swirl Influent, clear effluent and the foul
effluent (which acted as the Degrltter Influent) and at the
degrltter effluent. All samples were tested for TSS, VSS,
settleable solids* volatile settleable solids, ammonia, COO and
TOC.
Results Indicated clear overflow concentrations to be
much higher than those from any other sampling station.
Indicating that accumulation of materials within the sampler
Intake and Intake line had occurred during facility shut down
period. This posslbllty led to the standard procedure of
rinsing all equipment routinely and prior to pending storm
events. Available data was Insufficient to perform conclusive
performance analysis.
Lancaster Swlr I 9/1Q/8Q Storm Event /Summary
The storm event was characterized by moderate
showers lasting 45 minutes with a tot.jl accumulation of
0.18 In. (0.45 cm) of rain. The antecedent dry period was
10 days.
Liquid samples were collected at four locations (as
In 8/10/80) during the event, and one sample was taken from
1he grit hopper. A total of 23 liquid samples were
collected at 7.5 minute Intervals.
Although there were problems with the flow meters,
flow record analyses Indicate Influent flow to have been
approximately 18 cfs (509.76 l/s) at the start of the storm
and drop to 1 cfs (28.3 l/s) by the end of sampling, with
1.2 cfs (34 l/s) flowing to the Degrltter.
AM liquid samples were tested for TSS, VSS,
settleable solids, volatile settleable solids, ammonia, TOC and
COD. The grit sample was tested for volatile content.
Results Indicate good removals within the Swirl unit.
Details are Included In section 12.3.
190
-------
Lancaster Swlr I 9/1 4/60 Stgrffi Fven+
The storm event consisted of a brief heavy shower
lasting only 30 minutes with a total accumulation of 0.26
In. (0.66 cm) of rain. The antecedent dry period was 3 days.
Influent flow rates ranged from 40 to 6 cfs (1133 to 170 l/s)
during a 30 minute evaluation period. Liquid samples were
collected at four locations (as in 8/10/80) during the event
and two samples were taken from the grit hopper. A total of
20 liquid samples wero collected at 7.5 mlnule Intervals.
Difficulties continued with the Influent meters and weir
overflow rates were used to estimate flow volumes.
AM liquid samples were tested for TSS, VSS
settleable solids and volatile settleable solids. Grit
samples were tested for volatile content. Results Indicate a
much higher solids concentration In the flrs+ clear
overflow sample than within the Influent or foul flows. It Is
hypothesized that the first flush effect resulted In a
targe amount of f I eatables being Introduced to the unit.
This material would rise to the 'unit surface and not be
Included as part of the foul flow or samples. As the unit
fills* these floatables would accumulate on the surface and
be released with the first clear overflow In a concentration
higher than the Influent level. Unfortunately, no visual
observations were made during the storm and this phenomenon
could not be documented.
Unlike the storm event of 8/10/80 a very short
time had passed since the sample unit was last utilized
limiting The likelihood of accumulations bu M dup In the
sampler Intake and Intake line as assumed for the 8/10/80
event.
Solids concentrations In the remaining clear samples
Indicate Swirl Concentrator removal rates of about 30? at
Inflow rates near 20i-cfs (566.4 l/s) diminishing to minimal
removals at later flow rates of 10 cfs (283.2 l/s) and less.
As with the 9/10/80 event this Is probably due to the
solids characteristics change during the event rather than
being related to flow rate. Additional Swirl discussion of
this event Is Included In section 12.3. The Swirl Degrltter
performance during this event Is detailed In section 12.10.
Lancaster Swirl 1 0/2/BQ Storm Event Summary
The storm event was characterized oy *hree distinct
downpours resulting in 0.02 In. (0.05 cm), 9.08 In. (0.2 cm)
and 0.20 In. (0.5 cm) of rain, respectively. The antecedent dry
period was one week.
191
-------
Sampling began during the first rain period and
continue* through the second and third downpours. The
first downpour lasted about 5 minutes, followed one hour
later by a second event lasting about 10 minutes and
after 15 more minutes the heaviest rain fell over a 30 minute
period. A total of 46 liquid samples were collected and 2
grit samples were taken. Due to the Intermittent nature of
the storm the sample collection Intervals were Irregular.
Samples were tested for TSS, VSS, settleable solids,
volatile settleable solids and COO. Grit samples were tested
for volatile content. Results Indicate good removals during
the event. Swirl details are Included In section 12.3.
Lancaster SwIr1 4/1/81 Storm Event Summary
The storm event was characterized by two distinct
downpours. The first downpour lasted 75 minutes and
resulted In an accumulation of 0.26 In. (0.66 cm). The second
occurred 45 minutes later, lasted 1 hour and resulted In an
accumulation of 0.23 In. (0.58 cm). Combined with periods* of
drizzle the total accumulation during the event was 0.58 In.
(1.47 cm). The antecedent period Included a rainfall
accumulation of 0.25 In. (0.635 cm) over a 6 hour period
during the day prior to the evaluated event preceded by
13 dry days.
Sampling began during the first donnpour and
continued for 2 hours. Samples were collected using the EOF
Swirl Influent sampler, the Manning Swirl Influent sampler,
and the Manning clear overflow sampler. Failure of the EDP
foul sampler to produce samples was later found to be due to
debris clogging the discharge mechanism. Observations of the
EDP Influent sampler operation after the first samples had been
obtained showed that samples were possibly contaminated by
Intermixing and no analysis was performed. (Later review of
the actual sampling operation Indicated that the unit operated
properly). No grtt samples were collected and no Indication
of grit weight Increase was recorded on the charts. Both the
Mapco and the Leeds and Northrup Influent flow meters
operated during this event, however, the Mapco unit failed to
operate during the first 15 minutes. This was the first
event during which the Doppler flow meter op3rated.
All past events have demonstrated the trend of
pollutant concentrations balng Initially high, dropping
quickly and then tapering off. During this event, flow rates
were recorded for the first time over the Initial 15 minute
period. Review of the data obtained trom analysis of Initial
192
-------
samples collected by the EDP Technologies swirl Influent sampler
before "Inter-mIxlng" indicates s "first flush" effect.
4/1/81 Storm Data
Analysis of First 30 Minutes of Event
1 cfs » 28.3 l/s
TIME MIN. FLOW INF. SUSPENDED MASS FLOW RATE
cfs SOLIDS mg/l grans/sec
7.5 66.2 1439.8 2697
15 68.2 861.3 1662
22.5 53.6 650.7 987
30 50.5 522.0 746
Analysis of this storm Is not Included In section 12.3 due
to the uncertain nature of the Swirl Influent sampling.
Laneastar SwIrI 5/10/81 Storm Event Summary
The storm event was characterized by light rain
over a three hour period with an accumulation of 0.03 In.
(0.08 cm) followed by one hour of moderate rain and 0.11 In.
(0.28 cm) accumulation. The antecedent period had been dry for
eight days.
The Swirl operation commenced and sampling was
performed during the moderate rain period. Samples were
collected using the Manning Swirl Influenv sampler, the EDP
Swirl foul flow sampler, and the Manning degrltter foul flow
sampler. In addition, the Swirl tank was sampled at the end
of clear overflow to define tank concentrations. The EDP
Swirl Influent sampler clogged after collection of the first
sample The backflush system had not yet been Installed to
alleviate the problem. Manning sampler Influent samples were
therefore used during the analysis. At the time of the
event the Manning clear sampler was not operational and no
clear samples were collected. Neither the Mapco or Doppler
Influent meters operated properly limiting the Influent
deta.
Based on these Irregularities the number of samples
selected for final analysis was limited to five Influent, eight
Swirl foul, five Degrltter foul, and six Swirl tank samples.
193
-------
The eight Swirl foul samples consisted of four sample pairs
Intended to demonstrate the variability of sample
collection. Samples were analyzed for TSS, VSS, settleable
solids and volatile settle?Me solids. Swirl details are given
In section 1 2.3.
Lancaster Sw I rI 6/2/B1 Storm Evant Summary
The on-slte rain gage was not functional during this event.
Local reports of the total rainfall accumulation during the 12
hour porlod was 0.80 In. (2 en). The storm was not considered
Intense In nature.
All samplers simultaneously operated as the Swirl chamber
began to fill. Samples were collected using the Manning Swirl
Influent sampler, the EDP Swirl Influent sampler, the EDP Swirl
foul flow sampler and the Manning Degrltter clear flow
sampler. The EOF Swirl Influent sampler became plugged after
the Initial sample was taken and did not become operatlcnal
again until the latter half of the storn. Manning Swirl
Influent samples were used for Initial Influent analysis.
During the event the clear overflow receded below the
velr level mid-way through the storm resulting In several empty
clear samp I a bottles.
The Mapco Tlowmeter was not operational until approximately
11 minutes Into the storm. The Ooppler flowmeter failed to
operate. Samples selected for final analysis Included twelve
Influent, three clear, seven swirl foul and two degrltter foul.
Samples were analyzed for TSS, VSS, settleable solids and
volatile settleable solids.
Visual observations following the storm event Include,
approximately three buckets of leaves, paper and a few cans
remained on barscreen, very little debris on either the Swirl
or Degrltter weirs and little grit e<~cumul atlon from the Swirl
Degrltter.
Removal data Is I Imlted due to the lack of clear samples
and Influent samples from the EDP Swirl sampler. No conclusions
were drawn from this event.
Lancaster 6/22/B1 Storm Event Summary
The antecedent dry period prior to this event was 15 , .
Two days before the event the area received 0.25 In. (0.64 cm) of
low Intensity/drizzle precipitation. Prior to this the antecedent
dry period was 5 days. The storm was Intense In nature filling
194
-------
and overflowing the Swirl almost Instantly. Samples were
collected using the Manning Swirl |nf;i«en+ sampler, the EDP Swlri
Influent sampler, the EOF Swirl foul fIov sampler and the
Manning Oegrltter foul flow sampler. Grab samples were collected.
All samplers except the Swirl clear overflow vorked
properly during this event. Samples selected for final analysis
Included 18 Swirl Influent samples (12 EOP and 6 Manning), 1
Swirl clear, 12 Swirl foul, 12 Deprltter foul and 6 Swirl tank
samples. Analyses Included total settleable solids, volatile
settleable solids and COD. Results of this event Included a
dramatic comparison of the Manning sampling and the new EOP
device as discussed In section 12.4. Removals within the
Cagrltter are Included In section 12.10.
Laneas^r Sw I r I 7/2/61 Storm Event Summary
This event was characterized by an Intense rain which
lasted 70 minutes during which nearly 2 IP. (5.1 cm) of rain
accumulated. Automatic Initiation of the sampling system began
with the first clear overflow occurlng several minutes later.
Since personnel were on-slte awaiting the event, manual sampling
for settleab111ty analysis commenced simultaneously with this
Initial overflow so as to catch the "first - flush" effect.
SettleablIIty sampling required about 15 minutes after which
time routine site evaluation and the specialized dye
Injection flow calibration system was operated.
Samples were collected using the EDP Swirl Influent
sanpler, the Manning Swirl clear sampler, the Manning
Degrltter clear sampler and the EDP Swirl foul sampler. Swirl
foul samples were limited due to operational problems. In
addition. Swirl tank samples were collected twice during the
operation and a background dry weather flow sample collected
Just prior to the event. Results Indicated good removals and
are Included In section 12.3.
Lancaster Sw! r I 7/20/fll Storm Event Summary
On July 20 two storm events occured. The first began
at 5i45 an, lasted about forty minutes and resulted In 0.18 In.
(0.45 cm) of rain. The second event began at 6.55 pm, lasted
about fifteen minutes and amounted to 0.35 In. (0.89 cm).
During the first event the Swirl Degrltter and
01scostralner were operated and sattIeablI Ity samples were
collected from the Swirl Influent and clear flows. A problem
with the valvlng system In the building basement caused the
Degrltter to overflow and partially flood the basement. As a
195
-------
result the sett I eab I I I ty samples were collected some time
after Initial activation of the site, discrete "•?'•• ""•
minimal, and the EDP foul sampler located near the basement
floor UBS flooded and "shorted-out. "
The second event occured less than 12 hours later. Since
It had been dry for 16 days prior to the m or nlng event and
since the morning event was only of mild '"+ens'+Y ™°
evening event (predicted to have heavy rainfall) was also
monitored.
All samplers operated properly durl ng th I s event w Ith the
exception of the EDP foul sampler which had been flooded In the
morning storm. Settl eabl I I ty sampl es were col lected for Swirl
Influent and clear and from the Degrltter Influent and clear
overf low.
Results of the morning event were limited and no
conclusions were formulated. Sett I eab I II ty, degr I tter -and
dlscostralner results are discussed In sections 12.8, 12.10 and
12 11 respectively. Results of the evening event discrete
sampling demonstrated good removals and are Included In section
12.3. Settleabll I ty samples Indicated poor removals.
I ^neastar Swirl 7/28/81 Storm EYflHt Summary
Based on the poor sett I eab I I I ty data from the 7/20/81
event It was decided to test again the Sw I r I Inf I uent and cl oar
flow settleabl i Ity characteristics. For this final event,
collection of samples at the very start of the storm was stated
Ts the ma." pr.orfty with all other activities of secondary
Importance. Influent samples were Collected In the
diversion chamber. The flow-through apparatus with slide gate
closures was usel for 12 gal (45.4 I) of sample from the
5 version chamber ar which point the device malfunctioned. The
remaining 20 gal Ions (75.7 I) of sample were taken via the 3 In.
(7.6 cm) Influent line tap In the building. Influent samples
were also collected with the EDP Swirl Influent sampler for
comparative analysis. Results again Indicated poor removals as
discussed In section 12.8.
12.4. Com par I son Q± S«±d Influent ftnnrftfltr at I OnS
unit and the £fl£ Tarhno|QOles £rfliS Sec.tlOP.al
During four of the events monitored In 1981 a total of
fourteen simultaneous, or near simultaneous nf I uent sw I r I
samples were collected using the 36 In. (91.8 cji ) EDP
Technologies, Inc. Cross-Sectional Sampler and the Manning
sampler. Suspended, volatile suspended, settleable . .m voiat ,e
settleable solids analysis were performed on each sample.
196
-------
Relative concentrations of each of these parameters ware
compared for simultaneous samples. For near simultaneous samples
time-weighted average concentrations were used for comparison.
In all cases solids within samples from the Cross-
Sectional sampler unit were substantially higher than those from
the Manning sampler. Influent data from the 6/22/81 event are
plotted In Figure 78. Higher concentrations are Illustrated at
all times In the EDP Influent. The ratio of relative suspended
solids concentrations ranged from 6.7:1 to 2.2:1, and In
general decreased with time of operation. A single clear
sample was collected during this event and Is obvious
that whereas the Manning Influent sample concentration Implies
no removal the samples obtained by the Cross-Sectional sampler
Implies good unit performance. This concept Is further
considered at the end of section 12.6.
Relative ratios of Swirl Influent concentrations obtained by
Cross-Sectional sample to Manning sampler ranged from 2.1:1 to
7.6:1 for the four events for suspended solids, with similar
ranges for other so! Ids types. These multipliers roughly relate
to periods within the storm event as shown In Figure 79.
During the first 10 minutes of an event suspended solids
concentrations of samples collected with the Cross-Sectional
sampler had typically 6 to 7.6 times tho suspended solids
concentration noted In samples taken with the Manning unit.
Between 10 and 20 minutes of operation these ratios ranged
from 4 to 1 up to 6 to 1. After 20 minutes of operation the
effectiveness of the EOF sampler yielded about twice-the
solids concentration of the Manning unit.
These corrective factors were applied to two storm events
to place Into perspective seemingly poor Swirl efficiency
performance results. For this purpose one storm was selected
Just prior to Installation of the new sampler (fall 1980)
and the other event prior to EDP Involvement (1978). Results
of these correction factor applications are provided In section
12.6. In both cases results Indicate removals where previously
the data appeared poor and Inconclusive.
12.5 Detailed Event Analysis - Swirl Concentrator
Five storm event data sets were selected for detailed
analysis and are presented In this section. These events were
selected to demonstrate the efficiency of the devices or to
Illustrate other Informative phenomema. These events Include,
In *ie order of review; 9/10/80, 10/2/80, 5/10/81, 7/2/81 and
7/20/81 .
197
-------
vo
CO
Influent Swirl Concentrator
• EDP Technologies Cross Sectional Sampler
— — — Manning Sampler
« CLEAR
-------
0-10 m1n>6:l
10 - 20 m1n>4:l
>20 m1n>2:l
Sample Tine («1n)
0-10
11 - 2C
>20
Figure 79.
EQUIVALENCE UME
Comparison of Swirl
Influent suspended solIds
cone, using EDP Tech-
nologies Cross Sectional
Sampler with Manning
Sampler.
Marring Sampler (mg/1)
199
-------
Summary of RttsuIts 9/10/80
Plots of suspended solids vs time shown on Figure 80
Indicate good removal efficiencies for the Swirl unit. Clear
effluent samples concentrations are substantially lower than
Influent levels Indicating better than 30$ removal throughout
the storm. The Initial clear sample was taken prior to
operation of other sampling units since the unit began
operation at the first electrical Impulse whereas the other
units are delayed by the sampling Interval time. Based on
Influent vs clear overflow concentrations analysis, the
greatest removal efficiency occured during the beginning of the
event when both Influent flow rate and Influent pollutant
concentrations were highest. During the Swirl operation
suspended solids concentration fell from above 2000 mg/l to
less than 200 mg/l. At the same time the clear overflow
concentrations ranged from 600 mg/l to about 100 mg/l.
Approximate flow rates were 18 cfs (509.7 l/s) for 20 minutes, 7
cts (198 l/s) for the next 10 minutes and 1 cfs (26.3 l/s)
for the remainder of the monitoring event.
Assuming an Influent of 18 cfs (509.7 l/s) at the start of
the event, 94$ of the flow passed as clear effluent and
contained only 27} of the original solids content. By the end
of clear sampl Ing, Influent flow was about 6.5 cfs (184 l/s)
the clear overflow accounted for 83$ of the flow and contained
only 53} of the solids content. The removal rate and
efficiency at 18 cfs (509.7 l/s) were calculated to be 70$ and
64$, respectively. At 7 cfs (198 l/s) the respective values
were 50$ and 39$. Although this Indicates better performance
at the higher flow rates, the results are most I Ikely due to the
greater amounts of grit during the "first-flush" period. All
parameters exhibited trends similar to those exhibited by
suspended sol Ids.
Swirl Concentrator Influent, clear and foul concentration vs
time plots of volatile suspended solids, settleable solids and
volatlIe settlzable sol Ids are presented In Figures 81 through 83,
respectively. Swirl Concentrator Influent, clear and foul mass vs
time plots of suspended solids, volatile suspended solids,
settleable solids and volatile settleable solids are presented In
Figures 84 through 87, respectively. Inspection of these plots
Indicates that approximately 65$ of the Influent solids mass Is
settleable Inorganic grit.
Using approximate flow rates, mass determinations
Indicate that at no time during the event would the
effluent solids equal the Influent mass flow Implying solids
build-up within the tank. At the 'ancestor facility, there was
no provision available for monitoring the complete event up to
and including the final draw-down.
200
-------
Note:
Inf. * Influent
INF
CLEAR
FOUL
Note shading between
Influent and clear.
Figure 80. Suspended solids cone, vs time, Lancaster Swirl Concentrator/Regulator (9/10/80),
-------
r-o
O
rv>
s
s
s
GO
Q
a
I
D
•
00
C
in
•
CO
o
r
3
(D
\
r
CD
Q
INF
CLEAR
FOUL
Note shading between
Influent and clear
20 30
TIME
40
-I—
50
—»
60
Figure 81. Volatile suspended solids cone, vs time, Lancaster Swirl Concentrator/Regulator (9/10/80).
-------
ro
o
u>
TIME minufe««
INF
CLEAR
FOUL
Note shading between
Influent and clear
60
Figure 82. Settleable solids cone, vs time, Lancaster Swirl Concentrator/Regulator (9/10/80).
-------
ro
O
INF
CLE^P
FOUL
Note shading between
Influent and clear
0
60
Figure 83. Volatile settleable solids cone, vs time, Lancaster Swirl Concentrator/Regulator (9/10/80).
-------
ro
o
in
INF
CLEAR
FOUL
Note shading between
influent and clear
40
60
Figure 84. Suspended solids mass vs time,Lancaster Swirl Concentrator/Regulator (9/10/80).
-------
Q
Q
IS)
Q
D
r
is
o
3
\
3
INF
CLEAR
FOUL
Note shading between
Influent and clear
40
—i-
TIME —
Figure 85. Volatile Suspended solids mass vs time, Lancaster Swirl Concentrator/Regulator (9/10/80).
-------
ro
o
(/) *
niu
HQ
• s
Q
if;
O
-------
t\J
o
00
INF
CLEAR
FOUL
Note shading between
Influent and clear
40
TIME -
Figure 87. Volatile settleable solids mass vs time, Lancaster Swirl Concentrator/Regulator (9/10/80).
-------
Summary of ResuIts 1Q/2/BQ
Swirl Concentrator Influent, clear and foul suspended and
settleable solids concentration vs time plots are presented In
Figures 88 and 89 for the 10/2/80 event. Periods of positive and
negative efficiency relative to the Swirl Influent and clear
solids levels are highlighted. Overall the sampling results from
this event follow the trend noted throughout the program In
that the highest efficiencies are at the storm's beginning. Poor
efficiencies noted at later times may be attributable either to
solids buildup within the swirl unit and/or to less concentrated*
toore dilute Influent flow during these periods.
Based on estimated values, the flow pattern of this event
started at an Initial rate of 45 cfs (1274 l/s), decreased within
4 minutes to 17 cfs (481 l/s) and within 11 minutes to 7 cfs (198
l/s). After 20 minutes the flow was 2 cfs (56.6 l/s) for the
remainder of the event. The Influent solids concentration
declined steadily from 1400 to 200 mg/l.
Clear overflow samples concentrations Indicate efficiency
rates of 20 to 50* with efficiency rate decreasing with flow and
concentration. After 30 minutes the Influent concentration
continued to decrease. With flow at an approximate rate of 2 cfs
(56.6 l/s) clear overflow was minimal. Clear solids
concentrations were representative of tank liquid conditions so
that comparison with Influent values Indicates negative removals.
Although this may be due to the possible non-representative
influent samples, low Influent flow conditions Is the probable
reason.
During the latter portion of this evert clear and foul
sample concentrations were similar. With the ?low too low to
drive the swirling action required for trear-nant, the Swirl
tended to act as a small detention tank, "if nee the clear
solids concentrations did not drop In comparison to foul
concentrations during this period, It might be '.iferred that the
Swirl was of Insufficient size to act as a der-.ntlon tank even
under low flow conditions. It should be noted that application
of the suspended solids multipliers derived In suction 12.4 would
Indicate greater overall efficiency.
Removals based on foul samples are estimated to be 12.3?
at 17 cfs. (481 l/s) using a foul flow concentration (ffc)
factor see Section 11.6 of 1.4. At 4 cfs (113.3 l/s) the ffc
of 1.65 Indicates 58% removal. At 2 cfs (56.6 l/s) the
average ffc of 1.27 indicates 95$ removal.
209
-------
S
S
U)
s
s
o
U)
rs
s
s
s
\
INF
CLEAR
FOUL
Note shading between
Influent and clear
Negative
efficiency
—I
50
10
20
TIME
30
Figure 88. Suspended solids cone, vs time, Lancaster Swirl Concentrator/Regulator (10/2/80),
-------
Q
Q
S
0)
S
s
s
m
-i
3
(Q
\
r
§\
0
INF
CLEAR
FOUL
•
Note shad ings between
Influent and clear
Negative efficiency
TIME
Figure 39. Settleable solids concentration vs time, Lancaster Swirl Concentrator/Regulator (10/2/80).
-------
Results of samples tested for COD (see Figure 90) Indicate
similar efficiency rates as suspended solids samples. Based on
Influent suspended solids (SS) concentrations the Influent CCD
can be related to the influent SS by the equation:
COO- -.5 + .61 SS
with 9 correlation coefficient of 0.93 (N=11). These COO
estimates can be adjusted using the multiplier factors
derived In section 12.4. from the comparison of EDP Technologies
Cross Sectional and Manning Samplers. Application of these
factors would increase the Initial Influent COD concentration to
a level greater than the Initial clear value. Data as collected
indicate COD removal rates between 25 to 54? with efficiencies up
to 46?. The unadjusted results Indicate reasonable Swirl
efficiency and adjusting Influent values rould Indicate even
greater effectiveness.
Summary of RasuIts 7/2/B1
During this event the foul sampler failed to operate and
analysis Is based on clear samples only. Flow rate estimates
are Questionable and no mass rates were determined. In general
the flow began at about 20 cfs (566.4 l/s), rose to about 90
cfs (2549 l/s) at 6 minutes, fell to about 20 cfs (566.4 l/s) by
20 minutes and then remained near this level until the end of
clear sample collection.
Influent and clear concentration vs time plots for suspended
solids, volatile suspended solids, settleable solids and volatile
settleable solids are shown In Figures 91-94, respectively.
Inspection of these plots Indicate that most of the Influent
suspended solids was settleable Inorganic material. Influent
suspended solids concentrations were high during the first five
minutes and efficiency rates exceeded 60%. During the time the
flow was decreasing, the Influent solids concentration decreased
sharply and efficiency averaged 70%. As the flow stabilized the
Influent concentration Increased slightly and efficiency rate
Increased to an average of 80$.
Summary g_f Results 5/10/B1
Lack of clear sample collection during this evenl
precluded the standard analysis. Suspended solids and settleable
concentrations vs time plots for the Swirl Concentrator/Regulator
are given in Figures 95 and 96 for this event. Foul samples
collected In pairs by the EDP Technologies sampler
demonstrated a concentrating effect of the Influent solids.
212
-------
INFLUENT
CLEAR
FOUL
Figure 90. COD Concentration vs time, Lancaster Swirl Concentrator/Regulator (10/2/80)t
-------
ro
INF
CLEAR
Note shading between
Influent and deer
50
60
Figure 91. Suspended solids concentration vstlrae, Lancaster Swirl Concentrator/Regulator (7/2/81).
-------
LANCASTER SWIRL PROJECT 7-2-81
INF
CLEAR
DECRITTER
PO
(--
Ol
Note shading between
Influent and clear
TIME
Figure 92. Volatile suspended solids concentration vs time
Lancaster Swirl Concentrator/Regulator (7/2/01).
-------
ro
LANCASTER SWIRL PROJECT 7-2-81
INF
CLEAR
DEGRITTER
Note shading between
Influent and clear
20
30
TIME
40
50
80
Figure 93. Settleable solids concentration vs time
Lancaster Swirl Concentrator/Regulator (7/2/81J
-------
ro
LANCASTER SWIRL PROJECT 7-2-81
INF
CLEAR
DECRITTER
L
Note shading between
Influent and clear
TIME
Figure 94. Volatile settleable solids concentration vs ti.e
Lancaster Swirl Concentrator/Regulator (7/2/81).
00
-------
SUSPENDED SOLIDS VS TIME
INF
FOUL
FOUL DUPLICATES
TIME -
40
Figure 95. Suspended solids concentration vs time
Lancaster Swirl Concentrator/Regulator (5/10/81).
-------
SETTLEABLE SOLIDS VS TIME
INF
FOUL
FOUL DUPLICATES
10
TIME -
Figure 96. Settleable solids concentration vs time
Lancaster Swirl Concentrator/Regulator (5/10/81)
-------
Foul sample pairs consisted of samples taken one Immediately
after another with pairs separated by the standard sampling
Interval.
Comparison of the first two foul pairs Indicates a 6%
concentration dlscrepanc . This slight discrepancy Is
reasonable. At the 20 minute point In the event It was noted
that some leakage occurred through the sampling unit between
sampling Intervals.
For the purpose of evaluating this event the second
sample of each pair (lower curve In Figures 95 and 96 labelled
"foul duplicate") was used during evaluations to eliminate any
bias due to leakage. Foul sample suspended solids
concentrations averaged 1.63 times the Influent concentration
during ten minutes of foul sampling. During the next ten
minutes the foul concentrating effect decreased to 1.30.
Approximate Influent flow rates were 20 cfs (566.4 l/s) at
10 minutes, 8 cfs (226.3 l/s) at 20 minutes and 6 cfs (170 l/s)
at 30 minutes. At the ten minute mark with an approximate
Influent flow rate of 20 cfs (566.4 l/s), a foul flow rate of
1.5 cfs (42.5 1/sJ and a concentrating factor of 1.65, the
removal rate Is estimated as 12.3?. Similarly removal rates at
20 and 30 minutes are 22.5 and 35?.
In addition ^o the standard discrete sample analysis the
Swirl tank was hand-sampled at the end of the event. Six
samples were collected from wlthtn the Swirl tank at the end
of unit operation. Three of the samples were representative of
solids concentrations In the top 3 ft (91.4 cm) of the tank at
various positions around the tank. Three additional samples
were representative of the solids concentrations from the
surface to the tank bottom. Visual observation of the clear
plexiglass sampling tube Indicated that a 10 In. (25.4 cm) sludge
blanket existed at the tank bottom. These samples were
collected at positions midway between the tank outer wall and
the scum baffle. Although this sludge blanket somewhat
verified the notion of solids buildup during unit operation
the samples were taken 20 minutes after clear overflow ceased
and sampling ended. No mass balance was possible due to lack
of Interim samples, and the degree of settling of tank solids
could not be determined.
JLumm.flx:¥ fli fi£2JUl±2 1L2.Q.LS.L
Figures 97-100 present concentrations vs time plots of
suspended solids, volatile suspended solids, settleable sol'ds
and volatile settleable solids for the Swirl
Concentrator/Regulator. Inspection of these plots Indicate that
220
-------
l\)
Q
Q
S
LANCASTER SWIRL 7-20-81 PM
r-o
ro
0)-
«a
ffis'
CM-
m
•v). -
s •
rn
en GO
Pi;
D
0)
U}
\
r
^
s-
s
s
10
H 1—I 1 1-
20
30
40
-•—l—»—l—I—h
TIME
50
—l
80
Figure 97. Suspended solids concentrations vs time
Lancaster Swirl Concentrator/Regulator (7/20/81).
-------
r\j
ro
LANCASTER SWIRL 7-20-81 PM
60
TIME
Figure 98. Volatile suspended solids concentrations vs time
Lancaster Swirl Concentrator/Regulator (7/20/81).
-------
ro
r\>
GJ
LANCASTER SWIRL 7-20-81 PM
Figure 99. Settleable solids cone, vs time, Lancaster Swirl Concent,-itor/Regul ator (7/20/81),
-------
LANCASTER SWIRL 7-20-81 PM
TIME mir,Ljt«.
40
50
60
Figure 100. Volatile settleable solids cone, vs time, Lancaster Swirl Concentrator/Regulator (7/20/ai)
-------
most of the suspended sol Ids Is Inorganic settleable material.
Clear samples concentration were consistently much less than
Influent values during this event. Influent flow estlnatos
Include « low flow at first with 4.5 cfs (127 l/s) at 3
minutes, rising to 37 cfs (1048 l/s) at 11 minutes and up to
a peak of 56 cfs (1586 l/s) at 27 minutes. Estimated flows
remained high for the duration of sampling decreasing to only 30
cfs (850 l/s) at the 50 minute mark.
At the beginning of the event the removal rate was 41$
with an efficiency of 11$ based on the low level flow. Clear
solids concentration dropped quickly while the Influent
Increased resulting In an average removal and efficiency rates of
83$ and 80$ during the 20 minute clear sampling period at flows
above 35 cfs. (991 l/s)
Section 12.6. REASSESSMENT OF EARLY PROGRAM DATA
The results of the storm events just presented Indicate
that during those events the Swirl operated as an effective
solids separation device. During many of the ear I I or events In
1978, 1979 and even for several storms In.the fall of 1980, the
results were not as conclusive. lil fact, the opposite conclusion
was.true - the Twirl Concentrator did not seem to orovide any
treatment. ' .
The analysis presented In this section Is Intended to
provide Inferential evidence that the Swirl
Concentrator/Regulator was In fact providing reasonable solids
treatment removal during those periods. The Influent suspended
solids multipliers derived from sampling equipment differences
are used to reassess Influent concentration levels to the Swirl.
Two of the four 1980 events monitored by EDP were
previously discussed In Section 12.5 and had demonstrated good
efficiencies. The 8/10/80 event was noted to Include samples of
questionable quality and was dropped fron consideration. The
fourth event occurred on 9/14/80. Although low treatment
efficiency was noted, review of the raw performance data In
light of the potential sampling bias Indicates higher removal
rates. Numerous events were monitored during 1978 and 1979 and
few indicated positive efficiency results. The 6/21/78 storm was
deemed representative of these events and reviewed again.-Revised
results indicated solids treatment.
Section 12.4 discussed In detail the relative
efficiencies of the Manning sampler and the now EDP
Technologies Cross-Sectional sampling unit during collection of
Influent samples. It was shown that EDP unit samples were
225
-------
several times more concentrated and the relative" concentrations
were related to the time of sampling. Date from the 9/14/80
event and the 6/21/78 event were reviewed considering these
concepts. Table 13 Includes the original Influent data and
Influent data adjusted by appropriate concentration factors.
Original and adjusted removal rates and efficiencies were
determined for the 9/14/80 event. Originally the efficiency
rate was zero for the first sample and/ranged up to 30f.
Suspended solids efficiency rates exceeded 38jt. Overall, the
adjusted data Indicates effective swirl operation throughout
the event.
Similarly, the original 6/21/78 event data shown In Table
15 Indicated poor performance. Most sample sets Included
clear sample concentrations much higher than found In the
Influent. Adjusting the actual Influent concentrations by the
factors developed earlier yields values higher than clear
sample values. Removal rat'js which were low or negative based on
original values were calculated to be 20% to 56$ on an adjusted
solids concentrations basis.
Table 15
9/14/81 Analytical Data
TIME FLOW INFLUENT CLEAR OVERFLOW % REM JEFF
(MIN.) cfs (mg/l) (mg/l)
0 22 550 (3300)* 1280 - (74) - (57)
7.5 43 500 (3000) 350 32 (88) 29 (85)
15 24 350 (1440) 250 33 (83) 27 (77)
22.5 11.6 220 (440) 200 24 (60) 11 (47)
30.0 6.5 110 (220) 110 23 (61) 0 (38)
* adjusted values In parenthesis ( )
6/21/78 Analytical Data
Time Influent Clear Overflow % Rem**
Mtn. mg/l mg/l
7.5 47 ( 329) 822 - -
15.0 261 (1255) 913 - (27)
22.5 179 ( 537) 386 - (28)
30.0 104 ( 312) 249 - (20)
37.5 92 ( 276) 163 - (41)
45.0 74 ( 222) 124 - (44)
52.5 77 ( 154) 68 12 (56)
60.0 56 ( 112) 70 - (38)
** no flow date available and efficiencies not calculated
226
-------
12.7. AneI I Iary Samp I Ing Cons IdaratIons Swirl Concentrator Tank Samp I as
Tank samples were collected fron. within the Swirl
chamber at the end of unit operation during four events :
5/10/81, 6/22/81. 7/2/81 and 7/20/81. Three samples were
taken within the upper third of the tank at various
positions around the tank while three additional samples were
taken In the lower third. Samples were collected at positions
midway between the tank outer wall and the scum baffle.
Suspended solids concentrations of tank samples
from the 6/22/81, 7/2/81 and 7/20/81 storm events approximated
solids levels froa clear samples Indicating that there was
little solids accumulation within the unit during Its
operation. Samples from the 5/10/81 event had a higher
suspended solids concentration In the lower depth sample
stations. This concentration effect may have been due to a
delay of 20 minutes In starting hand sampling. During this 20
minute period after unit shutdown, solids began to settle In
the tank causing the formation of a sludge blanket on the
unit floor.
First Flush Mass
Most of the events monitored at Lancaster exhibited
Initial high concentrations occurlng at the start of
the event followed by decreased levels with time of
operation. Visual observations of pre-storm conditions
Indicated that the diversion chamber and 5 ft (1.5 m) pipe
leading to It acteti as a settling basin during dry
weather flow. This was largely due to debris caught on the
bar rack creating backwater In this area. The question
arises as to how much of the "first-flush" materials seen In
sampling event are actually due to this accumulation as opposed
to the runoff materials typical of the area.
For two of the storm events exhibiting high
"first-flush" effects the mass of solids entering the Swirl
during the unit was calculated. From this value and the
diversion chamber dimensions the depth of solids accumulated
within the chamber prior to the event required to produce
such a mass flow was calculated. A solids density of 75
lb/ft.3 (1.2 g/cc) (sg * 1.2) ind a moisture content of 90$
were assumed for materials accumulated In the diversion
chamber.
Calculations Indicate that +he so I Ids-I IquId mixture had
to have been 13 to 19 ft (396 cm to 579 cm) In depth In the
diversion chamber area ahead of the bar rack or 5 to 7 ft
227
-------
(152 to 213 cm) deep In the entire chamber to hs/e accounted *or
the entire "first-flush" mass. These numbers could be reduced
somewhat considering accumulations In the 5 ft (152 cm)
Influent pipe to the chamber but In any case It Is clear
that most of the material noted during "first-flush" was not
diversion chamber accumulations.
12.8 SattIaabI IIty ExparIments- SwIrI Concentrator Reou1ator
Before the EDP evaluation program had commenced, a
flow-through hand sampling device with trigger-activated slide-
down gates was for* ded along with sample vessels ana sampling
Instructions to Lancaster to collect a set of Influent samples. A
sett IeabI I Ity experiment was completed In January 1980. Eight
additional experiments were conducted during the evaluation
program covering the sett Ieab11 Ity characteristics of the
Influent and clear overflow. All mixing and settling was
performed using the state-of-.he-art sett IeabI I Ity column
detailed In Chapter 11.
Analysis of Swirl Concentrator/Regulator settleablI Ity
sample results Is reported In two parts. In section 12.8.1 all
sett IeabI I Ity data are compared by simultaneous plotting to
demonstrate the effects of Initial solids concentration and
sampling technique on sett IeablI Ity characteristics. In section
12.8.2 each event was reviewed where feasible to determine the
overall removal and the degree of removal at various
settling velocities.
12.8.1 Comparative Summary of Swirl Inf iuant SettlaabI IIty
Character IstIes wIth DIserete Samp Ier ResuIts
A comparison of all obtained suspended solids sett IeablI Ity
curves and Swirl Influent suspended solids concentrations noted
during first 20 minutes using discrete sampling Is shown In
Figure 101. The settleab11 Ity plots were derived from settling
column testing and evaluation of 30 gal (113.5 I) samples
collected as early as possible during an event. Suspended solids
concentration ranges are those noted by discrete automatic
sampling during the first 20 minutes of each event. For each
event the highest obtained discrete concentration occured within
the first sample and the least concentration within the 20
• Inute sample. This Is due to the first sample being taken
after the Swirl tank began to fill and the "first-flush" In
progress. Twenty mInutes generally allowed sufficient time for
the "first-flush" to pass and dilute stormwater to flow.
In Figure 101 the data as shown Indicate tha storm dates
with matchlrrg symbols used to label each data set. Influent
suspended solids concentrations ranges obtained by discrete
228
-------
RANGE OF DISCRETE SAMPLES
TAKEN DURING fST 20 MINUTES
Initial Concentrations
SUSPENDED SOLIDS CONCENTRATION mo/1
Figure 101 Lancaster Swirl Concentrator/Regulator,
influent, Comparison of settling column analysis
with discrete sampler results.
229
-------
samplers are included for all sampled events. Sett IeablI Ity data
or w included for Influent f!o« sain pies for five storm events.
Significant aspects of sett Ieabi I I ty samples as demonstrated by
simultaneous plotting are the range of Initial concentrations
and the varying percentages of solids fractions within
settling velocity ranges.
The settling curve plots In Figure 101 Illustrate that those
curves with the highest Initial concentrations have a greater
percentage of solids with high settling velocities and
contain solids fractions within all ranges. Comparison of the
sett IeabI I Ity data to discrete sampling data In Figure 101
Indicates that In most Instances the Initial sett Ieab11 Ity
sample concentrations were less than the range of solids
encountered during the first twenty minutes of discrete
samplIng.
For those storm events where clear overflow sett IeabM Ity
analysis (not shown In Figure 101) were performed, the range of
Initial concentrations was narrow (60 to 140 mg/l) while
simultaneously collected Influent samples ranged in Initial
concentration fro«-> 8 to 200 mg/l. Clear sett IeablI Ity samples
were typically c!< to clear discrete sample values than the
corresponding so. f Influent sett IeabI I Ity and discrete
samples.
The significance of Figure 101 Is that those Influent
sett)eab I I I ty samples which were similar to discrete Influent
concentrations also resulted In the best overall removals (In
comparison to clear overflow settling curves) and that removal
rates Increased as sett IeabI I I ty samples approached discrete
sample concentration levels. Thus, suspended solids treatment
efficiency could only be demonstrated by sett IeablI Ity samples
representative of "early-event" conditions. In reality some
of the Influent samples were not representative due either
to sampling late In the event or to Inadequate sampling
techniques or both.
The initial range of Influent and clear discrete sampler
data and the settling column results for the 7/2/81 event are
depicted InFlgure 102 and Illustrate the problem. The Initial
clear sett IeabI I Ity sample concentration !s similar to the
discrete sample concentration within the first twenty
minutes. Sett IeabI I Ity samples were collected over a 25
minute period starting several minutes after Initial discrete
sampling. The Influent settIeabI I Ity sample which was
collected simultaneously with the clear sample, did not
contain solids concentrations similar to discrete Influent
samples collected during this time. This discrepancy Is
probably due to the sampling techniques used. Clear samples
230
-------
10
20jnin. influent
Clear
Ornin Time
Clear Overflow
Settling Curve
20 min.
1
.01
.001
0 min. Time
Influent Settling Curve
Figure 102. Comparison of influent/
clear suspended solids
settling curves with
discrete sampler results,
Lancaster Swirl Conc-
entrator/Regulator (7/2/811
""""' l '' iobo r~
Suspended solids concentration (mg/1)
10,'OQO
231
-------
were collected by allowing a complete slug of water to flow
Into a bucket within the central clear down pipe. The bucket
was removed as soon as It filled. Thlstechnlqueellmlnated
the possibility of material bypassing the bucket.
Several Influent sampling techniques were attempted and
It appears that none could adequately deliver a large
representative sample In comparison to discrete sampling
results. This problem had been noted earlier and and led to
the design, manufacture and Installation of the EOF Technologies
Cross Sectional sampler. This unit was designed for small
volume samples (II) and the number of sampling cycles
required to obtain an adequate settleablllty sample, 30 gal
(104 I) during the 5 to 10 minutes In which the "first-flush"
usually passed, was mechanically prohibitive.
In sum, adequate and representative samples were obtained by
the Influent/ clear discrete sampling techniques and the clear
settleablllty sampling technique. Influent settleablllty results
are questionable. Analysis of Swirl solids treatment efficiency
based on Influent and clear sett I e&b I I I ty results are
questionable and on the low side. Conclusions reached In this
evaluation regarding Swirl treatment performance are based on
discrete sampling results.
The proposed "grit and organic" curves shown on Figure 6,
Chapter 4, Indicate that most of the solids to be removed had
settling velocities between 1 to 10 cm/sec (.032 - .33 ft/sec).
The settling curve plots In Figure 101 Indicate little materials
In this range. Discrete sampler results on Figure 101 would seem
to Indicate more material actually present.
12.8.2. £afflJiJCJ.SflJl £.1 Xtaar£±J.£iJ. JtfllJl ACJLUJJ.
Removal
In this section the performance comparing measured Influent
and clear solids settleablllty results are compared with
theoretical removal results as computed using formalisms
developed elsewhere (4).
Liquid flow velocities (from either the mathematical model
or physical model) for a Swirl Concentrator/Regulator of size,
Sf can be scaled to represent the flow In a geometrically
similar concentrator of sUe S2 using Froude number scaling.
Sett I eab I I I ty samples were collected from the Swirl
Influent and clear overflow streams on 7/2/81, during the AM and
PM events on 7/20/81 and on 7/28/81. Previously, the Influent
only had been sampled on 1/11/60.
232
-------
Each of these data sets are graphically displayed In
Figures 103-106 by plotting suspended solids concentration vs
settling velocity. It is obvious That no treatment Is svldent
based on the samples collected during the 7/20/80 am/pm and
7/28/81 events. The Initial concentration Is defined as the
average concentration of three samples taken from the column
Immediately after complete mix. In these three sample sets
the Initial clear concentration was substantially higher
than the Influent value. This could be due to sampling
methods or relative sampling times or both.
Sampling during the 7/2/81 event required 15 minutes
during which time the flow ranged from 20 to 90 cfs (566.4 -
2549 l/s) and averaged 45 cfs (1274 l/s). Based on the
constituents of the Influent flow as seen In the settleabl 11 ty
analysis and the APMA evaluation technique (4), the theoretical
efficiency rate would be 30$ at 5 cfs (141.6 l/s% and 12* at
40 cfs (1133 l/s). Actual efficiency was found to be 24$ and was
similar throughout the range of settling velocities. Since the
greatest degree of treatment would be expected to occur at the
higher settling velocities the cause of treatment remains
questionable. Possible causes of Irregular results Include
clear water sample collection at a later time than Influent
sampling, biased Influent sampling In which non-representative
amounts of heavy material were collected and the actual flow
containing much greater quantities of light material such that
even a low fractional degree of removal would result In high
overalI removal.
Hand sampling for this event consisting of alternately
collecting 8 gal (30.3 I) samples from the Influent and clear
flow streams, precluded error associated with time sequencing. Data
review Indicates that the mass of solids was not high for the
small sized solids eliminating the third possible source of
error. The second possibility Is quite likely considering the
sampl Ing techniques used. During this event the 3 In. (7.6 cm)
sample line located at the Influent pipe Invert was used for
sampling. Sample line velocity may have been Insufficient to
capture heavier particles. At the time when the clearwater
discrete sampler obtained samples having solids concentrations
similar to thGoi noted at the start of the settleablltv
analysis, the two discrete Influent samplers/(EDP Cross-Sect-
ional and the Manning) obtained samples averaging 545 mg/1. If
the ir.fnjent sampling technique for methods would have obtained
samples with solids concentration equalling 545 mg/1, the overall
observed removal would have been 74% at an average flow of 45 cfs
233
-------
IV)
CJ
-Pi
B:
HIM::
H
M
2
O
m- 1
o*
n
M
H 4-
0 •-
3 •
\ --
A
0 •
S 0
M
Figure 102
^ INFLUENT Initial Cone. 184 mg/1
* CLEAR Initial Cone. 140 mg/1
50
-f-
•+•
H-
100 150
SS CONCENTRATION
-------
W
z -•
n
TM
a
n
M
H
o
3
\
I
A
0 •
+ INFLUENT
Init. cone. 32 mg/1
^ CLEAR
Init. cone. 48.mg/1
-*-
40 60
SUSPENDED SOLIDS Cmg/L>
60 100
1 cm/s • 0.033 ft/s
Figure 104. Influent and clear suspended solids settling column results,
Lancaster Swirl Concentrator/Regulator,(7/20/81 am).
-------
ro
CJ
o»
m
H
H
r
M
z
o
m •
TM
o ;
n
H -•
/\
0
3
e
ID '
o •
^ 8
19
H. INFLUENT Initial Cone. 79 mg/1
^ CLEAR Initial Cone. 127 mg/1
-f-
-*-
-•-
20 40 60 80 100
SUSPENDED SOLIDS
-*•
120 140
Cwg/L>
1613
-t
1S0
OT/S = 0.033 ft/s
Figure 105.
Influent and clear suspended solids settling column results.
Lancaster Swirl Concentrator/Regulator (7/20/81 pin).
-------
8;:
ro
CO
H
r
M
z
o
m« ..
oK
o
H ••
o
3
\
ft
A
0 •
INFLUENT
Inlt. Cone. 97 mg/1
CLEAR
Inlt. Cone. 100 mg/1
20
i i
I
> - 1
1— *-
60
SUSPENDED
80
SOLIDS
I I
100
I I
-I
120
Figure 106.
Influent and clear suspended solids settling column results.
Lancaster Swirl Concentrator/Regulator (7/28/81).
1 cm/s = 0.033 ft/s
-------
(1274 1/s) which is reasonable considering the design parameter of
90* removal at design flow of 40 cfs (1133 l/r). The
theoretical removals cited above are based on the actual
(possibly biased) Influent sample.
Section 12. 9 OvervIBM Summary - SwlrI Concentrator Performance
A summary of the Swirl Concentrator/Regulator suspended
solids performance for 5 events monitored during the EDP
evaluation program and reviewed In section 10.5 Is presented In
Table 16 > Estimated removal and efficiency rates associated with
the Indicated estimated flows are presented In Table 16. These
results were derived primarily by Inspection of raw data. Overall
flow-weighted removal and efficiency calculations are not
presented due to the approximate nature of flow measurements. It
appears that the Swirl Concentrator/Regulator provided
significant suspended solids removal near design flow conditions.
Treatment seemed to deteriorate with lower flow rates. With the
exception of the 5/10/81 event (removal and efficiency based on
Influent and foul sampling) efficiency treatment rates exceed 60?
for flows In excess of 20 cfs (565.5 l/s). It should be noted
that most of the suspended solids noted were settleable Inorganic
grit matter.
The Swirl Concentrator/ReguIator appears to provide
approximately the degree of treatment that It was designed for,
that Is, 90% removal of settleable material at design flow of 40
cfs (1131 l/s). It appears that the performance data Is
Inconclusive to note any relationship between Influent rate and
efficiency of solids removal.
238
-------
Table 16
Summary of Swirl Concontrator Performance
Date of Event
9/10/80
10/2/80
5/10/81
7/2/81
7/2C/81 pm
Estimated*
Flo* Rate
(cfs)
18
7
50
7
2
20
8
6
55
20
5
46
Removal **
%
70
50
52
46
80
12
23
35
76
86
41
83
Eff lclency«»
62
39
47
24
5
5
5
10
73
78
11
80
Notes
a
a
a
a
a
b
b
b
c,d
c,d
c
c
** Definitions provided In Chapter 11.
Notes
a - Influent samples taken using Manning sampler
(correctly aligned)
b - based on Manning Influent and foul EDP Cross-
Sectional sampler results
c - EDP Influent Cross-Sectlonal/Sarnpler operative
efficiency based on sett Ieab I I I ty results Is 24%,
however, efficiency would have been 74$ If Influent
discrete sampling concentration results were used.
12.10 PagrIttar OparatIon Summary
Table 17 details the results of the Swirl Degrltter
operation during the evaluation period. Minimal amounts of
grit were collected during most of the events. The Degrltter
operation was plagued with constant solids "bridging" problems
due to Its structural design and lack of Internal wash down
unclogglng systems. Available data aro generally estimated values
of volume or weight. The grit weighing scale and r3corder
239
-------
mechanism frequently failed to ooerate and combined s!th ths
low grit quantities provided no rate of accumulation data.
Just prior to the 7/20/81 PM event the scale and
rocorder were tested and adjusted for proper recording of
moderate grit accumulations. Although the Degrltter appeared to
operate properly, the grit hopper
high winds and eliminated possible
In weight. Grit accumulated on
later plled-up, measured and
representative sample contained 67%
average of 56%) and 51$ percent
average of 38.7$).
Table 17
blew off the scale during
scale readings by the change
the scale Itself and was
sampled for analysis. A
moisture, (higher than the
volatile* (higher than the
Date
Swirl Degrltter Operation Summary
Total Grit
Time of Operation
Minutes
Discrete Degrltter
Sampling Operation
8-10-80
9-10-80
9-14-80
10-2-80
6-2-81
6-22-81
105
45
30
261
41
33
20 Ib
15-20 Ib
7-2-81
7-20-81
AM
PM
HOPPER
DEGRITTER
54
1.5 en ft
0.4 cu ft
MID
END
32
25
70
47
53
72
B
14.6
6.7
59
29
43
30
(sample taken falling from conveyor)
0.7 cu ft 82.5 76
73
50
160
5.0
.22
cu ft
cu ft
67
A • % Moisture
B * Volatile
Average Moisture Content
Average % VolatlIe - 56$
51.5
1 .7
38.7$
240
-------
The moisture content Is largely dependent upon the time
between unit operation and time of sampling as the vater
content tends to drain to the bottom of the grit pile and out
of the hopper. On 7/2/81 the grit was collected as It
fell from the conveyor and contained 72$ moisture. The
volatile percentage Is related to the ratio of organic
material and Inorganic grit. During the 6/22/81 event a mid
storm grit sample contained 29% volatile material and an end
of storm sample wes 43$ volatile, Indicating a decrease In the
organic grit content of the waste stream with storm duration.
Early; mid and late storm samples from an event In
July, 1980 (not shown In Table 17) were 23$ 38$ and 49$ volatile,
respectively. In addition, the ptrtlcle size distribution for
each sample was determined by sieve analysis followed by
volatile determination of each fraction. Sieve results for this
storm are Included In Table 18. The early-event sample contained
the greatest distribution by particle size. Particles within
the mid-storm sample were generally larger and the end-of-storm
sample was nearly entirely composed of large particles.
Early- event sample fractions *ere typically 20 to 30$
volatile Indicating mostly Inorganic material. Mid-storm
sample fractions were more volatile especially among the larger
particles. End-of-storm sample fractions were the most
volatile averaging 49$.
TABLE 18
SIava Ana lysis of Gr11 From Sw1rI Degr ttter
Portion of Storm
EARLY MID LATE
$ Moisture 43 56 61
$ Volatile 23 38 49
Steve Percentage ^Volatile •
No. Size EARLY MID LATE EARLY MID LATE
8 37 45 56 25 47 49
16 20 24 23 23 41 51
30 16 14 13 26 38 60
50 13 8 5 17 18 46
100 11 6 2 6 9 12
200 221 758
PAN I 10 6 13 0
241
-------
A sample of the accumulated material from the 7/20/81
pm event was determined to have a density, as sampled Including
air and moisture, of 57.9 1fc/ft.- (0.9 g/cc). Based on the
moisture content of 61%, a volatile fraction of 51.5? and
assumed specific gravities of 1.2 for volatile material and
2.65 for grit, the removed grit constitutes 56% air, 29.5%
water, 7.5? volatile material and 7.05$ Inorganics on a volume
basis. These values can be used to yield a rough Indication of
the transport volume required for removal of solid vaste
measured by weight. During the 7/20/81 am event, 0.7 cu ft
(198 I) of grit accumulation was estimated. The grit scale
recorded approximately 37 Ib (16798 g)of material from which the
density could be estimated as 53.lb/ft.3 (0.85 g/cc).
It vas difficult to de-l ermine the overall efficiency
of the Degrltter on an accumulated grit basis. The quantity
of material entering the unit was unknown due to the problems
with sample col I ect I on and the long'term Degrltter operation
during Swirl draw down. It could not be d.sumed that «||
material entering the Swirl not seen in the clear
overflow actually entered the Degrltter. It was observed that
some degree of accumulation of material occurs In the Swirl
during operation. Since the final minutes of Swirl tank drawdown
bypass the degrltter, an unknown quantity of material never
enters the unit preventing mass balance determination.
Efficiency evaluation of the Degrltter was attempted
on a short term basis by comparison of Influent and clear
overflow pollutant concentrations. This comparison was
performed In two ways during the evaluation. First, discrete
samples were collected from the Degrltter Inflow and clear
overflow during tde period of Swirl operation. Second, during
one event short t^rm, large volume samples were collected for
sett)eablI Ity analysis.
Degrltter Operation 6/22/B1
On 6/22/81 the Degrltter operated throughout the
storm event from the time the Swirl liquid level reached a
depth of 18 In. (45.7 cm) through the sampling period and
until the Swirl drained down to 18 In.(45.7 cm) again.
During tne first 33 minutes of this period, (see
Figure 107) samples were taken using the new EDP Technologies
foul flow Cross-Sectional sampler and the Manning degrltter
clear overflow sampler. Relative suspended solids
concentrations of these samples Indicated removal efficiencies
between 19$ to 75% with an average efficiency of 59%. Since the
entire liquid flow passes over the clear weir the degrltter
efficiency and removal rate are equivalent as defined In
242
-------
CJ
CD
Q"
Ql
0)
CQ
w
u
m
z
/\
OQ-\\
c . \\
D
U)
r
Q..
•. \
\\
_ INFLUENT
.. CLEAR
i I h
10
t l
I I I 1 t-
20
TIME mit-io-boo
30
40
Figure 107. Influent and clear suspended solids cone, vs time,
Lancaster Swirl Degritter (6/22/81).
-------
this report. The total mass'of solids accumulated during this
event was not recorded, nor was rne total time of operation
since the unit automatically shut down.
Accumulations over the 33 minute period of operation
«ere calculated to be 100 Ib (45.4 kg). Mid-storm and end-
of-storm samples were calculated to be 29$ and 43} volatile,
respectively. Based on these measurements and an assumed
moisture content of 50$ the volume of accumulation during the
first 33 mlrutes could be roughly estimated as 25 ft^ (0.7 m^).
Dagrlttar Operation 7/20/61
Sett IeabI I Ity analysis was performed upon the Swirl
Degrltter Influent and clear overflow streams collected during
the 7/20/81 am evaluation. Due to the difficulty In sampling the
Degrltter clear flow stream the sample volume was limited and
an 8 gal (30.3 I) settling column at the Swirl site was used*
Procedures for use of this column are detailed In Chapter 11.
For the bake of consistency the Influent Degrltter sample was
also tested using the column at Lancaster. Meterlal at the base
of the column at the end of the Influent sample test was
removed for sieve analysis. The results of the sieve analysis
were then used to aid the overall evaluation of the degrltter
performance.
Visual review of the 7/20/81 an data plot (see Figure 108),
Indicate little or no removals. TheDegritter clear overflow con-
centration of 130 mg/1 is slightly higher than the influent con-
centration of 103 mg/1 at the time of initial mix. This phenonmenon
was also indicated within sieve analysis results of the material
in the base of tne column at the end of each test. There were
greater amounts of material in the clear sample than in the in-
fluent with settling velocities in low velocity ranges.
Review of the sampling procedures used provide a
reasonble explanation. The procedure was to hand-sample the
clear overflow with a small bucket a sufficient number of
times to fill a 4 gal (15 I) container, requiring about 14
minutes per container. Then a 4 gal (15 I) container was
filled within 3 seconds with Influent sample via a 2 In. (5
cm) valve at the Invert of the Influent line.
Although this time differential appears minimal the
effects on relative concentrations can be substantial. The
detention time within the 250 ft3 (7080 I) Degrltter at a
flow rate of 1.2 cfs (34 l/s) Is 167 soconds. Combined with
the 1 minute time of sampling the clear samples were
actually collected nearly 4 minutes prior to the time of
influent flow through. Review of discrete clear samples
244
-------
ro
*»
in
3
H
r
INFLUENT initial Cone. 103 mg/1
CLEAR Initial Cone. 130 mg/1
60 eta 100 120
SUSPENDED SOLIDS Cmg/L>
140 150
1 cm/s = 0.033 ft/s
Figure 108.
Influent and clear settling column results,
Lancaster Swirl Degrttter (7/20/81).
-------
Indicates a concentration change from 180 to 120 mg/l over 7.5
minutes. The 60 mg/l change over 7.5 minutes Implies a
32 mg/l change over the 4 minute do lay. This vcluc agrees
veil with Initial concentration discrepancy seen.
On this basis It appears that the unit provided
no treatment, however the larger particles revealed during
the sieve analysis negates this conclusion. At the end of the
Influent settling column analysis* the column liquid
suspended solids averaged 55 mg/l and 11.9 grams of material
were removed from the base of the unit for sieve analysis.
As stated the settling column at Lancaster has limited mixing
potential. Had the solids fraction above been completely
suspended during the Initial mix, the Influent would have been
-.52 mg/l. On this basis the removal efficiency of the unit
could be estimated using the clear concentration of 130
mg/l to be 71%* This value Is based almost entirely on
material too large to be captured by the sett IeabI I Ity
analysis. Considering the time lag calculated above, actual
removals could have been higher had removal of smaller
particles actually occured.
This high value of removal and the large quantity of
large particle grit led to further review of the sampling
techniques. Influent samples were collected from a 2 In. (5
cm) port at the Invert of the 12 In. (30.5 cm) pipe, one foot
downstream of the Hydro-Brake. Although the Hydro-Brake should
create complete mix, calculations Indicate that up to 6.6
grams of the sieved 11.9 grams could have been non-
representative and biased the sample. Based on this the
actual Influent concentration could be estimated at 187 mg/l.
Indicating an overall removal of 30$.
Evaluation results for the Swirl Oegrltter are not
conclusive as to Its performance. Photographs of grit removed are
shown In Figure 109.
12.11 LANCASTER 01SCOSTRAINER EVALUATION
The Ofscostralner Is located In the mezzanine area
above the basement area. The unit Is set up to treat
either Swirl Influent or foul flow with the selected waste
stream pumped from the waste pipe Into the DIscostra Iner.
Flow rate Is controlled by throttling a valve downstream
of the pump.
The Intent of the D I scostra Iner is to remove fine
particles from the waste stream. The waste enters a central
basin wherein a circular screen assembly rotates on each
side. Liquid passes through the screens by gravity flow
246
-------
S3^3vr*KL
v -^r w£? '-J1 **x *S •' v"«*i
AflMjfeSl
b- MBPT
Figure 109. Photographs of grit removed by Swirl Degritter, Lancaster.
247
-------
leaving partlculate matter larger than tne screen mesh size
on the screen. The screen rotation allows t',e under* «+er
portion to rise out of the waste stream ar-'J >.Tr!« builds up
between the screens and eventually overflows Into the
screw conveyor from the degrltter.
Operation of the DIscostra I ner has demonstrated
that the unit falls to effectively remove the material as
predicted. Host observations of the unit operation Indicate
that this Is more due to the method of application than to
actual unit efficiency. In the period of 1978-1979, the
unit was operated using foul Swirl flow. Due to poor operation
of the backwash system the screens became totally clogged.
They could not be cleaned and were replaced.
Operation In 1981 Indicated that the backwash system
continued to clog the spray Jets. The backwash system was
designed to use tap water rather than treated DIscostraIner
waste and the clogged Jets are the result of fine
particles In the waste.
Accumulations within the central assembly of the
Clscostralner were minimal for a test conducted on 7/20/81.
Only 4 cu In. (65.5 cc) of material were collected during tho
morning event and on July 28, 1981 only 16 cu In. (262 cc)
were removed.
Since the feed to th" system Is throttled It Is
likely that only minimal amounts of material were In the
Influent flow. The material present would have been
macerated while passing through the pump. With the pump
located on the bottom floor, the pipe rising to the
mezzanine undoubtedly settled out much of the larger
mater tal.
The system as presently operated, falls to
demonstrate reasonable removal of material. Proper operation
for reliable testing would require setting the unit to
receive gravity Inflow with no vertical pipe sections for
settling and changing the screen backwash system to uMIIze
tap water.
12.12 Hydro-Brake
The Lancaster swirl degrltter was designed to treat
the entire foul flow from the swirl concentrator. At a 3$
foul underflow rate this would Imply 1.2 cfs (34 l/s) of the
40 cfs (1133 l/s) Swirl design flow. A 12 In. (30.5 cm)
248
-------
underflow line connects the two units and flow contro1 Is
required along the line. Initially, a pinch valve was uMllzed
coupled to a flow asetar. Sccsd on operational problems and
the desire to Investigate new technologies, the pinch valve
was rep1 ced by a" Hydro-Brake.
Flow through the 12 In. (30.5 cm) line was recorded
using a Flshei — Porter magnetic flow meter, calibrated at a
hydraulic testing laboratory prior to Installation. A flow
record taken during one event with the Hydro-Brake In place
Is pictured In Figure 110. With only slight variation the unit
delivered the design flow throughout the event. At the same
ime the flow to the swirl considerably varied and the height of
backwaterlng varied substantially as shown In the records In
Figure 110. •
The Hydro-Brake pro/ed to yield consistent flow
throughout each .event and required no maintenance during the
project. Oegrltter solids waste often Included tin soda cans.
There was never a clogging problem with the Hydro-Brake.
-------
I—I—I—1—I—I—I—I
I—f—
10 "fflf
Swirl Level
Swirl Inflov
l I I I I 1 I I I I
ritter Inf
•f Hydro-brake Design
§How 1.2 cfs
1 cfs * 28.3 1/s
1 ft - 30.5 cm
Figure 11C. Stripcharts, Hydro-Brake Evaluation.
2SO
-------
CHAPTER 13
West Roxbury Evaluation
13.1 Foreword
This chapter details the results and analysis of the
Swirl and Helical Bend Regulator/Concentrator operations at
West ^oxbury. Mass. Section 13.2 contains a chronology of
events during the evaluation and details operational problems
and solutions and field/analytical data for each event. Section
13.3 presents Individual event results which demonstrate the
effectiveness of the units for stormwater treatment. Results of
several sett IeabI I Ity analysis are described In section 13.4.
Summary performance results are presented In section 13.5.
Section 13.6 presents the results of various dye studies
performed during the evaluation period.
13.2 Eva IuatI an OvervIew
The following chronological summary highlights
major activities during the West Roxbury evaluation.
the
DATE
UNIT
ACTIVITY
11/26/79 Swirl
Helleal
4/4/80 Swirl
4/10/80 Swirl
5/8/80
6/29/80 Swirl
Hel leal
7/8/80
7/16/80
7/29/80
8, 1 1/80
Swirl
Sw Ir I
Hel leal
Samples Collected Solids, COD Tested
Samples Collected Solids, COD Tested
Samples Collected Solids, COD Tested
Samples Collected Solids, COD Tested
Monitoring Initiated Insufficient
runoff for sampling
Samples Collected Solids, COD, TP 'iested
Samples Collected Solids, COD, TP Tested
Influent Column Test - Sett IeabI I Ity
Insufficient runoff to operate unit
Monitoring Initiated Insufficient Runoff
Samples Collected Solids, COD Tested
Samples Collected Solids, COD Tested
Monitoring Initiated Insufficient Runuff
251
-------
Swirl
Helleal
8/15/80
6/30/80
9/16/80
9/26/80
10/3/80 am
10-3-80 pm
10/18/80
10/25/80 Swirl
Helleal
11/18/80
11/28/80
12/10/80
12/14/80
5/12/81
5/14/81
5/20/81
5/28/81
6/9/81
6/22/81
Swirl '
Helical
Sw Irl
Helical
Monitoring Initiated Insufficient Runoff
Begin pipe sleeve Insrallatlon
Complete pipe sleeve Installation
Monitoring Initiated Insufficient Runoff
Monitoring Initiated Insufficient Runoff
Samples Collected Solids, COD Tested
Samples Collected Solids, COO Tested
Influent Column Test - Sett IeablI Ity
Monitoring Initiated Insufficient
Runoff
Samples Collected Solids, COD Tested
Samples Collected Solids, COD Tested
Influent & Foul Effluent Column Test-
SettleabllIty
Snow
Monitoring Initiated Insufficient Runoff
Snow
Snow
Monitoring Initiated, Insufflcent runoff
for samp I Ing
Begin pipe sleeve alterations; V-Shape
End pipe sleeve alterations
Monitoring Initiated, Insufficient
runoff for samp I Ing
Samples Collected Solids, COD,
Bacteria Tested
Samples Collected Solids, COD
Tested
Swirl Influent & Clear Settling Column Test
Samples Collected, Solids, COD Tested
Samples Collected Solids, COD Tested
Swirl Influent & Clear, Helical Clear
Settl ing Colum.i Test
252
-------
6/25/81 Monitoring Initiated, insufficient
runoff for sampling
7/29/81 Monitoring Initiated, minimal runoff,
limited sampling
8/4/81 Swirl Samples Collected Solids Tested
Helical Samples Collected Solids Tested
Swirl Influent & Clear, Helical Clear
Settling Column Test
Summary of Events Data
The following Is a description of the slorm events
monitored atthe West Roxbury site. Performance data Is briefly
described. Those events demonstrating good performance and
facility operation are discussed in greater detail in section
13.3.
Event! 1 1/26/79
The pre-storm event consisted of light rain during
most of the early morning hours with a total accumulation of
0.12 In. (3mm) over a 12 hour period. The actual sampled event
was characterized by moderate to heavy downpour resulting In
records Indicated an antecedent dry period of over a week with
no measurable amounts of rainfall.
The Swirl and Helical Bend Concentrator/Regulator were
operated simultaneously. Samples were collected manually at the
Influents, clear overflows and foul underflows. Flow In the
Swirl ranged from 0.18 to 1.19 cfs (5.1 to 33.7 l/s). In the
Helical Bend flow varied from 1.20 to 1.53 cfs. (33.9 to 33.7
l/s).
Swirl samples selected for analysis Included 7
Influent, 6 clear overflow and four foul effluents. Parameters
tested were TSS, VSS, settleable sol ids, volatile settleable
solids and COD. The Helical Bend samples Included 3 Influent, 2
clear overflows and 2 foul effluent. Samples were analyzed for
TSS, VSS, settleable solids, volatile settleable solids and COD.
Event; 4/4/80
This storm event was characterized by moderate rain
over a 70 minute period. The antecedent period consisted of a
dry day, one day of light rain followed by a dry day
Immediately prior to the storm event evaluated.
253
-------
Twenty six samples Mere collected from Swirl Influent,
20 from clear overflow and 29 from the foul underflow. Flow
during sampling ranged from 0.23 to 2.3 cfs ( 6.5 to 65 l/s).
Samples were analyzed for TSS, VSS, settleable solids volatile
settleable solids and COD. A dye study was performed during this
event to observe flow patterns within the Swirl.
Event! 4/10/80
This event was characterized by medium rain for a 2
hour period resulting In a total accumulation of 0.2 In. (5
mm). The antecedent period consisted of 3 dry days and 2 days
of trace precipitation Immediately prior to the event.
Samples were collected manually from the Swirl at 5
minute Intervals. A total of 79 samples were taken: 27 Influent,
20 clear and 32 foul samples. Flow through the unit during
sampling ranged from .55 to 1.1 cfs (15.5 to 31.1 l/s). Samples
were analyzed for TSS, VSS, settleable solids, volatile
settleable solids and COD. Removal results are discussed In
detail In section 13.3.
Eventi6/29/80
The storm event was characterized by 0.16 In.(4 mm) of
rain In a 50 minute period. The antecedent period consisted of
7 dry days and one day of trace precipitation prior to the
event.
Samples were collected manually from the Swirl
Influent/ clear/ foul and Helical Bend Influent at Intervals of
approximately 5 minutes. 43 samples fro: the Swirl Included,
14 Influent, 9 clear, 19 foul and 1 floatables sample. 44
samples from the Helical Bend Included, 14 Influent, 8 clear and
21 foul with one special floatables sample taken from the
unit.
Influent settling coluirn samples were collected
Immediately following the start of rainfall within the first 15
minutes if the sampling program.
Flow In each unit during the event averaged .88 to 3.5
cfs (24.9 to 9 l/s). Samples were analyzed for TSS, VSS,
settle&ble solids and volatile settleable solids.
Event; 7/79/BO
This storm event was characterized by a period of
moderate rain, resulting In a total of 0.43 In. (11 mm) over a 36
hour period. The aniecad3nt dry period was 5 days.
254
-------
Sampling began on the Swirl using both manual grab
sampling and Manning automatic samplers. Automatic samples were
taken at Swirl Influent and foul flow at 3.75 minute
Intervals. Manual sampling of Swirl clear. Helical Bend clear.
Swirl Influent, Helical Bend foul and Swirl foul were coordinated
with automatic samplers. Ftoatables samples were also taken
early In the storm event.
Flow Into the Swirl unit ranged from .44 to 4.7 cfs
(12.5 to 133 l/s) during sampling. A total of 52 samples were
collected from both units. Including II Swirl Influent (3
automatic and 8 manual), 5 Swirl clear (manual), 11 Swirl foul
(9 automatic and 2 grab) along with 2 foul drawdown samples
taken after unit shut down. Five Helical Bend clear samples
were taken manually and 18 foul samples were collected
automatically, 11 of which were drawndown samples. A dye
study was conducted on the Swirl to note flow patterns.
Sediment had collected at the bottom of the Helical
Bend covering an area approximately 7 In. (18 cm) wFde by
0.125 In. (0.6 cm) deep throughout the length of the unit.
Samples were taken and analyzed for TSS, VSS, settleable solids
and volatile settleable solids.
Event; 1Q/3/BQ
The storm event was characterized by light rain over a 20
hour per I o<1 prior to the sampled event. The antecedent dry
period was 4 days. Heavy rain began at 12:30 pm resulting In a
total accumulation of 0.16 In. (4 mm). The Swirl operation began
with sampling of Swirl Influent, foul effluent and clear
overflows.
Manning automatic samplers were used for Swirl Influent
and foul effluent at 3.75 minute Intervals. Clear overflow
samples were taken manually In 1-1 bottles. Manual grab
samples were also taken of Swirl Influent. Flow Into the Swirl
varied from 0.01 to 6.3 cfs (0.3 to 178 l/s) during t.Hs
event. After two minutes Into the sampling event the Swirl foul
became barkwatered and the Influent gates were shut down. At this
rime the Helical Bend was turned on. Operation of the Helical
Bend was possible due to the elevated position of the Helical
Bend foul discharge over tho swirl foul discharge. Helical
Bend samples were taken manually every 3 minutes for
Influe"1-, clear and foul. The Helical Bend was shut down at
2:35 pn. and the Swirl was then operated for the collection of
discrete *.nd settling column samples. During sampling flow
levels In the unit ranged from 3.76 cfs to 5.2 cfs (106.4 l/s
to 147.2 l/s)
255
-------
Fifteen Influent, 22 clear and 14 foul samples
were taken from the Swirl unit. Twelve Influent, 20 clear and 12
foul samples were collec+ed from the Helical Bend. Samples were
analyzed tor TSS, VSS, settleable solids, volatile settleable
solids and COD. Influent settling column samples were
collected and settling column test was performed for this event.
Event; 10/25/80
The storm event was characterized by 3.5 hours of
moderate rainfall with a m-xlmum 15 minute Intensity of 0.48
In./hr (12 mm/hr) and a total accumulation of 0.54 In. (14
mm). Antecedent dry period was 6 days. Approximately 1 hour
after rain began, flow entered the Swirl and Helical Bend
units. Influent and foul flow sampling was then Initiated.
After one-half hour, tha units were full and clear overflow
sampling commenced.
Mechanical problems limited the pumpage from the foul
sump tank and the Srlrl operation was temporarily ceased. The
Helical Bend unit operation continued for another 45 minutes, at
which point Influent to the Helical Bend was shut off and the
Swirl placed on-line. The Swirl unit was then operated for one-
half hour.
A total of 59 samples were collected during the Swirl
operation and 39 samples were collected during the Helical
Bend operation. Influent, Swirl foul , and Helical Bend foul
effluent samples were collected for sett IeabI I Ity analysis.
During sampling flow ranged from .27 to 4.5 cfs (7.6 to 127 l/s)
In the Swirl and 0.27 to 4.3 cfs (7.6 to 122 l/s) In the
Helical Bend. All samples were analyzed for TSS, VSS, sett IeabI«
solids, volatile settleable solids and. COD. Sett I eab I I 11-/
samples were analyzed for TSS and VSS.
Event; 6/9/81
On June 9, 1981 a light rain occurred throughout the
morning at a rate of 0.1 In./hr (2.5 mm/hr). Between noon and
1 pm 0.25 In. (6.4 mm) of rain fell and the Srlrl and
Helical Bend were operated for an hour. The antecedent dry
period was 2 days. Samples were collected from The Swirl and
Helical Bend Influent, clear flows and foul flows for discrete
ana IysIs.SampIes were also taken from the Influent and clear
flows of both units for sett IeabI I Ity analysis.
Thirteen Influent, 13 clear and 17 foul samples
were collected from the Swirl and 10 Influent, 9 clear and 15
foul samples were collected from the Helical Bend. All samples
256
-------
were analyzed for VSS, TSS, settleable solids, volatile
setleable solids and COD. Several Influent and clear Swirl
samples were analyzed for Total Col:fcrms. No bacterial removal
was noted.
Evant; 6/22/81
On June 22, 1981 a heavy thunderstorm occurred
amounting to 0.4 In. (10 mm) of rain. The antecedent dry
period prior to this event was 6 dry days. The Swirl and
Helical Bend were operated simultaneously until problems arose
with the Swirl operation and the unit was temporarily shut down.
For this event a screen assembly was placed In the Swirl clear
overflow downtube to collect any debris passing over the
clear weir. The clear overflow screen assembly filled with
floatables material to the extent that the screen clogged.
Since It was Impossible to remove the screen with the unit on-
line, the Swirl Influent was stopped and the screen assembly
pumped dry.
During this time the Helical Bend was operated
after a slight Initial delay while evaluating the Swirl
condition. The Helical Bend was successfully operated for 1
hour and Influent, clear overflow and foul underflow samples
were collected along w!1h composite samples for sett IeablI Ity
analysis of the Influent and clear overflow. After terminating
the operation by closing the Influent gates, drawdown samples
were collected In the foul flow for an additional 14 minutes.
The Swirl unit was placed back on line and run
simultaneously with the Helical Bend for 19 minutes. Swirl
operation continued after Helical Bend shutdown for a total
operation time of 33 minutes plus 30 minutes of downdraw.
Influent, clear overflow, foul underflow, clear samples for
settleablllty analysis were collected. Eleven Influent, 10
clear and 12 foul samples were collected from the Swirl unit
and 12 Influent, 12 clear and 12 foul samples were collected
fronr. the Helical Bend during this event. Flow In the Swirl
ranged from 1.5 to 2.6 cfs (42 to 74 l/s) ana 1.5 to 10 cfs
(42 +o 280 l/s) In the Helical Bend. A dye study was performed on
both units to visually observe flow patterns.
Eventi 7/29/81 Storm
Rain showers were moderate to heavy. En route to
the site sho«ers were light with a short period of heavy
downpour. The streets were relatively clean. Upon arrival
at the site the rain had stopped and the ;:nlts were turned on
and began filling. The stormwtler was dark grey and heavily
257
-------
laden with quarry dust. Floatables were minimal and consisted
mainly of leaves, grass and seeds. Grab samples of Swirl
Influent and clear and Helical Bend clear were collected .
Automatic Sw!<~! fou! and Helical B e n u foul samplers each
col lected 20 samples. As the units emptied the bottoms of both
were covered with fine dark grey sedlmenl, presumably quarry
dust.
Summar 6/4/81
This storm event was characterized by a heavy
downpour lasting approximately 35 minutes resulting In a
total accumulation of 0.86 In. (22 mm). The antecedent dry
period was 5 dry days. Sampling began 15 minutes after the
start of the event with Swirl and Helical Bend Influent and
flow samples along with Influent samples for settling column
analysis. Sixteen minutes Into the event clear overflow began
over the units and samples were collected. Samples were
collected manually for Swirl and Helical Bend clear flows and
for Influent and clear settling column samples taken. Automatic
samplers were used for the collection of Swirl Influent and
foul samples and Helical Bend foul samples at 3.75 minute
I nterval s.
Eleven Influent, 12 clear, 20 foul and 3 tank samples
from the Swirl unit were analyzed for TSS, VSS, settleable
solids, volatile settleable solids and COD. 11 clear and 18
foul samples from th^ Helical Bend were analyzed for the same
parameters. Influent samples were not collected from this
unit during this event.
Settling column analysis was performed on samples
taken from the Swirl Influent/ clear and Helical Bend clear
overflow locations. Samples were collected during the Initial
20 minutes of the storm event and were analyzed for TSS and
VSS.
A dye study was performed on the Swirl unit during
the storm event. Flew during the sampling period ranged from 0.7
to 12.3 cfs (19.8 to 348 I / s ) In the Swirl. Flow Increased to
12.3 cfs (348 l/s) during first 10 minutes of operation. Similar
flows were observed In the Helical Bend.
A mass solids analysis was also performed on the
Swirl unl+. A sample of solids remaining on the bottom of the
unit following the event were quantified and compared with
solid concentrations of liquid samples taken during the
event. Tank solids accumulations at various stages of the
storm event were estimated.
258
-------
13.3 Data I Ied Event Analysis
Seven storm event data sets were selected for
detailed analysis to demonstrate the efficiency of the test
devices or display any other event occurances relevant to the
evaluation. The events discussed In detail Include: 6/29/80,
7/29/80, 10/3/83, 10/25/80, 6/9/81, 6/22/81 and 8/4/81.
Removal rate and efficiency results are calculated using methods
outlined In section 11.3.
Although both suspended and settleable solids were
determined during the evaluation program suspended solids Is
considered as the prime parameter to assess removal performance
for the following reasons. In the sett IeabI I I ty test a sample Is
at loved to settle for 1 hour and then the supernate Is
sampled to determine non-settleable solids. Settleable solids
are calculated based on suspended solids and non-sett IeabIe
solids estimates. This one hour test bears no relation to the
Swirl or Helical Bend efficiency since the solids fractions
removed are not comparable. The maximum depth within the sample
settling vessel from which the supernate sample Is
collected was typically four In. (10 cm) . Given a one hour
specified settling time the supernate would be expected to
capture no particles with settling velocities g- eater than
0.01 ft/sec (0.003 cm/sec). Since the Swlr! Is roughiy expected
to remove 100 percent of pertlcles with settling velocities
greater than 0.15 *• jec (5 cm/sec.) and few parrlcles with
settling velocities -.ess than 0.0075 ft/sec (.21 c.i/sec.) the
results of the settleable solids test are not significant.
Pollutant concentration vs time plots are presented for
nearly all events and solids type. Plot of mass removals are not
Included since overall removal and efficiency rates for both
units were not hIgh.
6/29/80 RasuIts Summary
Swirl Influent, clear and foul sewer time vs
concentration plots of suspended solids, volatile suspended
solids, settleable solids and volatile supsended solids are
presented In Figures 111 through 114 respectively, for the
6/29/80 storm event. Shadlngs are noted on each plot to denote
both positive and negative solids treatment removals. Discharge
vs time plots are Included on each figure. Similar time vs
concentration plots of suspended solids, volatile suspended
solids and settleable solids are presented In Figures 115-117
respectively for the Helical Bend Regulator.
Influent solids concentrations fluctuated sharply
creating periods of removal and negatl/e removal in comparison
259
-------
ro
CT>
o
SOLIDS REMOVAL
10
80
Q0
Figure 111. Suspended solids concentrations vs time,
W. Roxbury Swirl Concentrator/Regulator (6/29/80).
-------
SOLIDS REMOVAL
INFLUENT
CLEAR
FOUL
80
O0
Figure 112. Suspended solids concentrations vs time,
W. Roxbury Swirl Concentrator/Regulator (6/29/80),
-------
SOLIDS REMOVAL
Figure 113.
Settleable solids concentrations vs time,
W. Roxbury Swirl Concentrator/Regulator (6/29/80).
-------
3SSSSS SOLIDS REMOVAL
NON-REMOVAL
INFLUENT
80
00
Figure 114. Volatile suspended solids concentrations vs time,
W. Roxbury Swirl Concentrator/Regulator (6/29/80).
-------
•VJ I
O
*
0
-b
r\)
a
in
s
B
to
»&
z
D
5
s
O
r
(n
3
(D
GT
S
SOLIDSi REMOVAL
NON-REMOVAL
INFLUENT
CLEAR
FOUL
1 cfs » 28.2 1/s
I
I
\!
30
40
TIMfc.
50
70
Figure 115.
Suspended solids concentrations vs time,
W. Roxbury Helical Bend Regulator (6/29/80).
B0
-------
ro
cr>
Oi
s
•Q
W
m
n*
HQ
ra
m
>
ou
rw
ma
Q
(A
O
r
•-•IN)
O Q
(OS
100
mg/L
SOLIDS REMOVAL
NON-REMOVAL
___ CLEAR
____ FOUL
1 cfs = 28.2 1/s
1
| \
' I
*
0
-»—
10
_« 1 1—
20 30 40
TIME
50
60
70
80
GO
Figure 116.
Settleable solids concentrations vs time,
W. Roxbury Helical Bend Regulator (6/29/80).
-------
SOLIDS REMOVAL
INFLUENT
CLEAR
Figure 117.
Volatile suspended solids concentrations vs time,
W. Roxbury Helical Bend Regulator (6/29/80).
-------
to Swirl clear overflow sample concentrations. Overall the
data Indicated an average of 2\% renova I rate for the Swirl.
At an influent rate of 2 cfs, (56.6 l/s) the Swirl efficiency
rste was 8.2$.
Helical Bend removal and efficiency rates averaged 23$
and 6% over a 30 minute period with flow between 1 and 2 cfs.
(28.3 and 56.1 l/s)
7/29/80 ResuIts Summary
Plots of time vs pollutant Influent, clear and foul
sewer concentrations of suspended solids, volatile suspended
solids* settleable solids and volatile settleable solids are
presented In Figures 118-121, respectively for the Swirl
concentrator and In Figures 122-125 respectively for the Helical
Bend Regulator.
During this event Influent samples were collected
manually and by automatic samplers, for comparative analysis. The
automatic sampler Intake was located at the Invert of the
Influent pipe. Hand samples were taken throughout the depth of
flow. The first hand sample contained heavy solids,
floatable organlcs. The automatic sample failed to Include
this material. Thereafter the automatic sampler Indicated
Influent solids to be much higher than manual sample values.
Volatile fractions were similar for the two techniques
Indicating tne excess material to be Inorganics. Based on the
location of the Intake It can be assumed that heavy materials
settled and moved along the base of the pipe, biasing the
automatic sampler results to be somewhat higher than manual grab
values. Efficiency shadlngs are rot shown on Figures 118-125 due
to differences In Influent sampling techniques.
Swirl Removal and efficiency rates averaged 35.3$ and
27.5$ respectively during 10 minuter; of clear sampling. Prior
to this negative removal results were Indicated due to a
single high clear solids concentration value* The sample was
highly organic and probably contained floatable material
remaining from the period of high Influent floatable materials.
The Helical Bend demonstrated average removal and
efficiency rates of 36$ and 2$ respectively over a ten minute
period as the flow Increased from 1 to 4 cfs (28.3 to 112 l/s).
1 Q/3/80 ResuIts Summary
Plots of time vs Influent, clear and foul sewer
pollutant concentrations of suspended solids, volatile suspended
267
-------
ro
W
00
AUTO
CLEAR
FOUL
GRAB INF
1 cfs = 28.2 1/s
20
•+•
•+•
-*-
•+•
•4-
30 40
TIME mir.u»t«»«»
50
60
70
118. Suspended solids concentrations vs time,
• W. Roxbury Swirl Concentrator/Regulator (7/29/80).
-------
IV)
tft
VO
•nl/T
•
o
0 "
^
'. ^^^^
\^~~^
01
^ si
-H*
M £•
TQ
m
"
U) 0)
c B-
W Q
TJ
m -
z
O i\)
T7_ •"
m K\"
• • ^y
D O
w •
o
r i-»
•-i Q-
D Q
w
3
(D S
i
i
*• AiiTn TM*7
i
i
i CLEAR
i
« FOUL
i
i
• ' GRAB INF
1
i 1 cfs = 28.2 1/s
i
i
i *
^r^^-sZ^^'' -^
• r ^5J r.^'^ ~~^~"" ~~
'-"""" X ^^- — /
r 21 10 20 30 40 50 60 70
TIME mir-»Ljt«»
Figure 119. Volatile suspended solids concentrations vs time*
U. Roxbury Swirl Concentrator/Regulator (7/29/80).
-------
o •
0
?BT
500 400 300 200 100 0
SETTLEABLlE SOLIDS mg/L
^^
— ^^— ^—— !-B--
' _ AUTO INF
i
\ CLEAR
'i
', FOUL
• i
', GRAB INF
i /
\ ' 1 cfs = 28.2 1/s
0 10 20 30 40 50 60 70
TIME minut«»
Figure 120. Settleable solids concentrations vs tine.
W. Roxbury Swirl Concentrator/Regulator (7/29/80).
-------
- J
2! ur
1 -
0
o
?s;
500 400 300
VOLATILE SETTL
m
OJN
rs-
ms
w
0
r •"•
Us-
D S
w .
3 S.
_^^
AUTO INF
I CLEAR
i
» FOUL
i
' ' _„ GRAB INF
i
. i
1 wo*..
\ 1 cfs = 28.2 1/s
• i
i
i
i
i
^^-^c-.ir ;*~~
^ 0 10 20 30 40 50 30 70
r TIME ml nut, ••
Figure 121. Volatile suspended solids concentrations vs time.
W. Roxbury Swirl Concentrator/Regulator (7/29/80).
-------
ro
-«4
ro
2 ur
0 '
ae
0
^^^^
^s^
^ ^^
Ul
SI
Q
cn
c
TJ Q'
n*i s
z .
o
m
£B
D K-
n
^j
(A ^
r* B"
Q
S
i
i
i
i
i
.1
i
i
i
i >»
' sS
/ >'"' \
^ // r~\ /\ A —
H\ /"A / ^— ' v v/
1 \ // / \ AIITn TWP
• \ / l
\_^< 1 CLEAR
\ / FOUL
0 10 20 30 40 50 60 70
TIME minvj't««
Figure 122. Suspended solids concentrations vs time.
U. Roxbury Helical Bend Regulator (7/29/80).
-------
o •
o
"*> m-
o a
o
r •
r
> *
HQ-
r~
m.
(/) a)
(/)§
TJ
m •
z
D |\)
m Q-
D Q
(/)
O
p ,_,
3
(0 a
\m
./
^^
. ^ •• *
~*~~~"^
'
1
^ AMTn IMF
1
.1 CLEAR
i
i
• FOUL
• i
i
' GRAB INF
i
» 1 cfs = 28.2 1/s
' 7\
1 /
- l>^^x^v-^ ^--^-'' w^^
r 2 10 20 30 40 50 60 70
TIME m 1 r-ii_cfe.«*
Figure 123. Volatile suspended solids concentrations vs time.
W. Roxbury Heli-cal Bend Regulator (7/29/80).
-------
o
c
ro
B0
70
Figure 124. Settleable solids concentrations vs time,
U. Roxbury Helical Bend Regulator.(7/29/80)
-------
23 UI"
0 '
500 40C
• VOLATIL
) 300
E SETTLE
>
U) -
O
a _
(fl S-
r
i
_^
_ _ AUTO INF
i
i % CLEAR
"i FOUL
i
•i GRAB INF
- i 1 cfs = 28. 2 1/s
i
i
• i
i
• \ ^ /\ /x
Z 10 20 30 40 50 G0 70
TIME m i fiufe-A*
:igure 125. Volatile settleable solids concentrations vs time,
U. Roxbury Helical Bend Regulator (7/29/80).
-------
solids, settleable solids and volatile settleable solids ar»
presented In Figures 126-129 respectively for the Swirl
Concentrator and In Figures 130-133 respectively for the Helical
Bend Regulator.
During the first 20 minutes of this event the flow
Into the Swirl Increased from zero to 5 cfs (142 l/s) while the
Influent suspended concentrations decreased from 300 to 180
rag/I. The removal rate averaged 22% over the first ten
minutes* followed by 10 minutes of higher clear overflow
solids concentrations. Between the 20 and the 40 minute point of
this event the flow remained constant at 6 cfs (170 l/s) and
Influent solids concentration rose to 850 mg/l with average
removal and efficiency rates of 36$ and 32$ . During the
final 10 minutes of operation Influent concentrations fell
sharply to 150 mg/l and the removal rate fell to 5$ while
averaging 6% at an efficiency of 7.2$.
The Swirl foul flow concentration factor averaged 1.056
between 10 and 50 minutes Indicating a removal rate of 7$ and
an efficiency of 6.5$ . These values are lower than estimated by
clear flow considerations Implying tank concentration solids
buildup or deposition on the tank floor. Foul flow solids
concentrations rose sharply during final drawdown clearly
Indicating the presence of this material.
The Helical Bend demonstrated nc removals during this
event based on relative Influent and clear sample
concentrations. After 5 minutes of operation the unit
demonstrated reasonable removal based on foul sample analysis.
The foul flow concentration factor averaged 1.375 during
this period at an average flow of 4.5 cfs (127 l/s). Removal
rate and efficiency averages were 16$ and 10$ respectively. In
addition, the flna' drawdown samples contained Increased
solids concentration Indicating additional removal.
10/23/80 Resii I ts Summary
Plots of time vs Influent, clear and foul sewer
pollutant concentrations of suspended solids, volatile suspended
solids, settleable solids and volatile settleable solids are
presented In Figures 134-137 respectively for the Swirl
Concentrator and In Figures 138-141 respectively for the Helical
Bend Regulator.
On the basis of clear sample concentrations the
Swirl unit provided no removal, as clear overflow solids
concentrations were typically higher than Inflow values. Ratios
of foul to Influent solids concentraMcn ranged from 1.07 to
276
-------
SOLIDS REMCV.*.
E222D NON-HEMOVAu.
INFLUENT
CLEAR
FOUL
0
Figure
20
30
40 50
TIME
60
70
80
100
126. Suspended solids concentrations vs time,
W. Roxbury Swirl Concentrator/Regulator (10/3/80).
-------
00
O »
O
r
^0)
,-s
W
in
2
n
o
w
o
r
Q
a
ui
\
0
SOLIDS REMOVAL
INF
CLEAR
FOUL
1 cfs • 28.2 1/s
10
20
30
40 50
TIME
60
70
-l 1 1 i t
30 G0 100
Figure 127. Volatile suspended solids concentrations vs time,
W. Roxbury Swirl Concentrator/Regulator (10/3/80).
-------
SOLIDS REMOVAL
ro
0
-*>
01
a
NON-REMOVAt.
INFLUENT
CLEAR
FOUL
100
Settleable solids concentration vs time,
W. Roxbury Swirl Conct. trator/Regulator (10/3/80).
-------
TI **
O
U)
3>
" Q
,*
IVM\VM Suu-IuS RtXCVAi.
/^~ ^v
f \ 1':->'."'A':1 v" \.-3p-kj.-y A
, S \ INUiN (\trlUV At-
/ \
O
r
* B
(D
r •
m
w a-
O Q
w
D
WN
3 Q
(0
r
INFLUENT
. CLEAR
— FOUL
1 Cfs « 28.2 1/s
0 10 20 30 40 30 60 70 80 ' 90 ' 100
TIME m i r->i_ft««»
Figure 129. Volatile settleable sol
-------
ro
CD
00 100
Figure 130. Suspended solids concentrations vs time,
W. Roxbury Helical Bend Regulator (10/3/80).
-------
ro
oo
ro
0
o^ ;
200 150 100 50 0
, VOLATILE SUSPENDED SOLIDS mg.
ESSEX! SOLIDS REMOVAL.
IvA^J wnM-RPMOVAL
__ INFLUENT / \
/ I
. CLEAR / \
/ \
FOUL / \
• j
1 cfs * 28.2 1/s I \
1 » /l
/ \ ' *
TIME mir-»u«k«»»
Figure 131. Volatile suspended solids concentrations vs time,
W. Roxbury Helical Bend Regulator (10/3/80).
-------
ESSSS SOLIDS REMOVAL
NON-REMOVAL •
INFLUENT
CLEAR
FOUL
1 cfs - 28.2 1/s
Figure 132.
Settleable solids concentrations vs time
W. Roxbury Helical Bend Regulator (10/3/80).
-------
8
* •
•n
(A ISL
200 150 100 50 0
VOLATILLE SETTLEAQLE SOLIDS rr
CWiW SOLIDS REMOVAL
''••''"•'"'^ NON-REMOVAL
1 A
INFLUENT / \
/ »
CLEAR t \
. FOUL / |
/ \ ,
1 cfs = 28.2 1/s / \ A
/ \ / «
9^;^, _., .' V '
^^^. -1* LA* • •. •~*^— • i —f — . — — r- — s . . ,,, t— .. L.
fl 0 10 20 30 40 50 60 70 80 Q0 100
|- TIME mlncrfc,«*
Figure 133. Volatile settleable sol Ids concentrations vs time,
U. Roxbury Helical Bend Regulator (10/3/80).
-------
-------
k\'\\\\Vi
REMOVA
.a
NON-REMOVAL
INFLUENT
1 cfs « 28.2 1/s
I 1 >-»•••> H-H »—H 1 1 <•»•>•» I » I >>>>>(> > t t > 1 1
Figure 135. Volatile suspended solids concentrations'vs time.
W. Roxbury Swirl Concentrator/Regulator (10/25/M).
-------
SOLIDS REMOVAL.
CO
'••'•'•'V'^1 NON-REMOVAL
20
Figure 136.
160
Settleable solids concentrations vs time,
W. Roxbury Swirl Concentrator/Regulator (10/25/80).
-------
r
m
SET
SOLIDS REMOVAL
NGN-REMOVAi
INFLUENT
CLEAR
FOUL
i i i i 1 T
0
FLEABLE SOLIDS mg/L
1!
1 1
/ \ J j 1 cfs - 28.2 1/s
~--AL—--*L^ fi
3 20 40 80 B0 100 120 140 1 Old
TIME m i r-it_j-b«*
Flgyre 137. Volatile settleable solids concentrations vs time
W. Roxbury Swirl Concentrator/keoulator <10/25/80)
-------
SOLIDS REMOVAL
CD
80 821
TIME m
120
140
180
Figure 138. Suspended solids concentrations vs time,
W. Roxbury Helical Bend Regulator (10/25/80).
-------
o
KYKWfl SOLIDS REMOVAL
£22 NON-REMOVAL
I—i—I—I—I—I—I—(
180
Figure 139. Volatile suspended solids concentrations vs time
W. Roxbury Helical Bend Regulator (10/25/80).
-------
r\i
10
-n Ul
r*
o
* w
• '
0 ,n
t>UI
SOLIDS REMOVAL
NON-REMOVAL
i—i—i i » i i i i i i i i
INFLUENT!
CLEAR
FOUi_
160
Figure 140.
TIME
Settleable solids concentrations vs time,
U. Roxbury Helical Bend Regulator, (10/25/80).
-------
ro
v£>
ro
SOLIDS REMOVAL
NON-REMOVAL
I I I I »—H 1—I 1 1
180
Flgyre 141. Volatile settleable sol Ids concentrations vs time,
W. Roxbury Helical Bend Regulator (10/25/80).
-------
2.7 over an 80 minute period Indicating solids concentrating
effect. The Influent flow ranged from 2 to 4 cfs (56.4-112.8
l/s) and the solids concentration fluctuated between 50 and 150
mg/l. Average suspended solids removal and efficiency rates are
estimated at 29.7% and 21.4$ based on an average flow of 3 cfs
(85 l/s) and foul concentration factor of 2.05.
Average suspended solids removal and efficiency rates
for the He I leal Bend Regulator are estimated at 36.5$ and 26.5%
for an average flow rate of 2.5 cfs (70.5 l/s).
6/9/81 RasuIts Summary
Plots of time vs Influent, clear and foul sewer pollutant
concentrations of suspended solids and settleable solids are
presented In Figures 142 and 143 respectively for the Swirl
Concentrator and In Figures 144 and 145 respectively for the
Helical Bend Regulator.
During this event Swirl foul samples Indicated no removals
as all concentrations were less than Influent values and minimal
final drawdown flush was observed. Heavy solids deposits were
observed In the Swirl tank floor near the foul sewer outlet at
the end of operation. Based on a representative square foot
sample, the total approximate solids debris In the tank was
26.5 Ib (12 kg) of which 25% were volatile solids.
Based on clear overflow sample concentrations, the Swirl
removal rate was 27$ removal rate at 3 cfs (85 l/s) and 34$
at 0.8 cfs (23 l/s) after 30 minutes. Corresponding unit
efficiencies were 21$ and 4$ respectively. These results
demonstrate the effect of Influent rate on unit efficiency.
Although the removal rate Is higher at 0.8 cfs (23 l/s) the
actual efficiency Is greater at the higher flow rate since
less flow-splitting occurs. After 30 minutes the Influent flow
remained below 1 cfs (28.2 l/s) and the Influent solids
concentration decreased continually.
Based on clear overflow samples no removals occurred for the
He) leal Bend during the first 15 minutes. From 15 to 45 minutes
the unit averaged 41$ removal rate at an Influent rate of 0.8
cfs (22.6 l/s). The efficiency equalled 10$ at this flow rate.
However, considering the foul sewer the average foul
concentration factor was 1.12 Indicating a removal rate of 50$
and an effclency of 19$ over 60 minutes followed by a highly
concentrated final flush.
293
-------
f\J
UD
SOLIDS REMOVAL
CLEAR
FOUL
1 cfs - 28.2 1/s
0
Figure
10
-i 1 1- -i 1— I I 1 i 1 1 1 i 1 1
20 30 40 50 80 70 80 Q0 100 110 120
TIME m 1 r~ii_ct«k«»
142. Suspended solids concentrations vs time. West Roxbury Swirl (6/9/81).
-------
IV)
<0
en
INFLUENT
CLEAR
FOUL
100 110 120
0
Hgure
10 20 30 40 50 80 70
TIME mino-b«k«
143. Settleable solids concentrations vs time, West Roxbury Swirl (6/9/81 )t
-------
FOUL
cfs - 28.2 1/s
10 20 30
• I «—I—I—t-
H 1 1
TIME mlnutta*
«•« 144. Suspends, so,1ds concentr.,tons „ ttme> ^ ^^ ^^ (w>
-------
(V)
i*
0
• S.
.
" ,
^
0)
wj*-
H .
r
in -
P i\)
Q-
(0 .
O
r -
(E Q-
\
r
Q
\
,\
\
.V^^
> ^ M
0 10 20
Figure 145. Settles
•
SuuIuS RcMGVAL
NGN-REMOVAL
Figure 145. Settleable solids concentrations vs time, West Roxbury Helical (6/9/81 )t
-------
Solids remaining In the unit following this event were
concentrated at the last 12 ft (3.6? m) of the unit. A
representative sample was taken 5 ft (1.5 m) from the discharge
line approximately 1 ft? (0.09 m2) In area. The total solids
residue In the unit was calculated to be 15.3 Ib (7.5 kg) of
which 35$ Is volatile solids. Photographs of the Helical Bend
Regulator during and after the event are shown In Figure 146.
6/22/81 Removal Results
Plots of tine vs Influent, clear and foul sewer
pollutant concentrations of suspended solids, volatile suspended
solids, settleable solids and volatile settleable solids are
presenled In Figures 147-150 respectively for the Svlrl
Concentrator, and In Figures 151-154 respectively for the Helical
Bend Regulator.
During the first 15 minutes of this event no Swirl
removals are observed by relating Influent to clear
concentrations. Influent flow averaged 2.2 cfs (62 l/s) with an
average foul concentration factor of 1.2. Foul rate was 0.25
cfs (7 l/s). Based on foul flow data an average removal rate of
27 % and efficiency of 16$ were calculated. For the remaining
15 minutes of evaluation clear samples Indicated an average
removal rate of 33$ and efficiency of 21$ at a flow of 2 cfs
(56.6 l/s). Foul samples Indicated no removal although a final
drawdown /flush was observed.
The Helical Bend showed negative removals using clear
sample data. The average foul concentration factor was 1.43
using an average flow rate of 1.8 cfs (51 l/s). Based on foul
flow data, an average removal rate of 35$ and an average
efficiency rate of 21.5$ were calculated.
6/4/61 Removal ResuIts
Plots of time vs Influent, clear and foul sewer
pollutant concentration of suspended solids, volatll* suspended
solids, settleable solids and volatile settleable solids are
presented In Figures 155-158 respectively for the Swirl
Concentrator and In Figures 159-162 respectively for th» Helical
Bend Regulator.
The Swirl average removal and efficiency rates equalled
about 6$ based on clear sample values during the tlire of stable
flow at 12.3 c's (340 l/s). Prior to this point no removals were
Indicated on a clear sample basis. As the flow approached the
design rate of 6 cfs (170 l/s) the removal rate Increased to an
average of 9.5$ while the efficiency was calculated to be 55(
considering the foul sewer data.
29H
-------
ro
10
B. SolIds remaining after draw-down
Reproduced from
best Available copy
A. View from Influent end
Figure 146. Photographs of the Helical Bend Regulator
operation, (6/9/81).
-------
oo
o
o
0
t>
• Q
NON-REMOVAL.
1 cfs - 28.2 1/s
INF
CLEAR
FOUL
30 40
TIME
50
60
0 10
Figure 147. Suspended solids concentrations vs time, West Roxbury Swirl (6/22/81).
e0
-------
SOLIDS REMOVAL
INFLUENT
CLEAR
FOUL
30 40
TIME m
60
70
80
Figure 148. Volatile suspended solids concentrations vs time, West Roxbury Swirl (6/22/81).
-------
CO
s
S'
0
t>
SOi-IDS
1 cfs « 28.2 1/s
30 40
TIME
50
60
70
G0
Figure 149. Settleable solids concentrations vs time, West Roxbury Swirl (6/22/81).
-------
co
o
CO
o
«U1
0
J*
G
<
°GT
»-»
r w
m 01 •
a
i\)
r
m **
> 01"
GO Q
r
w|"
0Q
r
n Ul-
KA
3 i«
in ^
iu
\
vX^**1^^^^^ flsMMsVfl SCt.ICS REMGVAi.
"**^^-' • — -^^^
"~*
| •••.•.<:•. v.V:i hGN-REMCVA,_
1 *l LJ I » 1 » L» ( 1 •fc-' » r 1 —
1 cfs - 28.2 1/s
r\
, . INFI OFNT
\ ^_ CLEAR
1
-. t , FOUL
1'
\''\ "^>*N*» /\S\
^-*^3X. NN KX\N\
^^A-^,, ^^Vj^^OsNi
^^^C^^35£^2§^;iu.
"*"*" -— — "" ""^—
0 10 20 30 40 50 60 70 80
TIME mlrtu^tt*
Figure 150. Volatile settleable solids concentrations vs time, West Roxbury Swirl (6/22/81).
-------
to
2
TI j\)
n 1 •
0
•
3000 '2000 'l000 0
SUSPENDED SOLIDS rwg/L
^^^ RVWiVS SOLIDS
| ••/.»:•.-•.*>'•; 1 WON— RFK
i cfs = 28.2 1/s
1
. INFLUENT jl
II
CLEAR l l
FOUL 1 1
J I
1 l
II
l i
i l
/ 1
l
^^ «^ gff^^^c* '*****, i 'y*i ****(/• 3^ ^**"»*i" ** *iC^ "•^^*«c
. _! — — (. . 1 •-•
fe 10 "20 30 *B S0 8" 7B
TIME minot**
Figure 151 Suspended solids concentrations vs time, Mest Roxbury Helical (6/22/81).
REMOVAL
IOVAL
80
-------
SOLIDS REMOVAL-
3 B
(Q
V
r
0
10
20
30 40
TIME i
50
70
80
Figure 152. Volatile suspended solids concentrations vs time, West Roxbury Helical (6/22/81).
-------
CJ
o
o>
ZL. 5 0 3000 ' 2000 ' 1000 ' • 0
FLOW of. SETTLEASLE SOLIDS -ng/L
-^ RmvM SOLIDS REMOVAL
!'•:• '.*.>••.%'•; 1 MniSl-RFMOVAL
.... IN^l.UFNT 1
II
CLEAR ![
FOUL | j
1 cfs = 28.2 1/s ' •
'• l\
.^£§*% i !
-'{ :-Sv ^f tfV->iV^l*.t^«'.-?i't<^jy^ ^^^v^'W^ ^rrii''iff — ^ — ' uS7^^*f'^ '> i~ «'J "j"^ i' '~ "^* _ f- ^ |
. **• i .1 — —
0 10 20 30 40 50 B0 70
TIME minufc.**
Figure 153. Settleable solids concentrations vs time, West Roxbury Helical (6/22/81).
H
80
-------
oo
O
IV)
• •
(Jl
< 0)
o a
TQ.
H
»-*
r
m o
Q
tn Q
m
H ,
H
r
m
a
in
3
W
\
SOLIDS REMOVAL.
r.-..'.v--.V:l NQN-REMCVAi-
M
INFLUENT
CUEAR
FOUL
1 cfs * 28.2 1/s
30 40 50
TIME minute*
70
80
0 10 20
Figure 154. Volatile settleable solids concentrations vs time. West Roxbury Helica1 (6/22/81),
-------
a
SOLIDS REMOVAL
0 CJ
t> •
• Q
NON-REMOVAL
30 40
TIME mirtot«i<
50
Figure 155. Suspended soMds concentrations vs tlrae, West Roxbury Swirl (8/4/81).
-------
Co
o
?^
o -f-
0
t>
Sui-IDS REMOVAL
1 cfs = 28.2 1/s
Figure 156. Volatile suspended solids concentrations vs time. West Roxbury Swirl 8/4/81.
-------
U)
•—•
o
SGuIuS ReXCVAL
\ ;
INFLUENT
CLEAR
FOUL
Cfs = 28.2 1/s
•
30
TIME
40
50
60
70
Figure 157. Settleable solids concentrations vs time, West Roxbury Swirl (8/4/81)
-------
-nWf
r B
o ••
M. .
0 S
f» >
• Q /
%
>®T
HQl
M T
r
m -r
3S--
HQ
5 1
gs A
r T^
^ 1 » ^
Ilj¥
5n
D
0) 4-
3 B ,
\ ^
r
Figure
ty^wi SOLIDS REMOVAL
,x ^^-
r-iV^-A"! NON-REMOVAL
... INFi UENT
. CLEAR
t J\ 1 FOUL
r/ \^,—^-^~— ^ lcfs-28.21/s
' ^X VVOC *^ •Vv^vT --ss^X ^
^ V v\
^-v/
10 20 30 40 50 60 70
TIME m i r-i i_i -k. • •
•
158. Volatile settleable solids concentrations vs time. West Rexbury Swirl (8/4/81).
-------
o ••
0
t»
•
SOLIDS RE:MOVAL
l-.i-v--.-.'.•.*•.'•; I KinM-RgMQVAL
0)
w i"
(0
D
m •
z
D
a K.
5)
O
r
3 Q
(0 S
r
Q
1 cfs • 28.2 1/s
INFLUENT
CLEAR
FOUL
j^ ^*-'
10
20
30
TIME
50
60
70
Figure 15a. Suspended lolids concentrations vs time, West Roxbury Helical (8/4/81).
-------
CO
I—"
CO
"* ST
o -•
s
(/)
c
(/)
D
m
2
D
m
D
(/)
o
r
a
3
(fl S
r 0
\
Sui_IDS REMOVAL
NON-REMOVAi
1 cfs = 28.2 1/s
INFLUENT
CLEAR
•
FOUL
20
30
TIME
40
50
60
70
Figure 160. Volatile suspended solids concentrations vs time. West Roxbury Helical (8/4/81)<
-------
o •-
-»> Q
u
m s
m Q
H
r
m
m
r
m
U)
o
r
Q
s
D
M-
Q
3 S
(D a
r
—i—
10
REMGVAu
1 cfs = 28.Z 1/s
-»-
30
TIME
40
50
60
70
Flaure 161. Settleable solids concentrations vs time, HesJ; Roxbury Helical (8/4/81),
-------
CO
»—•
en
N +
o Q±
S
s
D 0)
fs
en0
m
Q
B
B
S
l0
SOLIDS REMOVAL
$:w.l NON-REMOVAL
1 cfs - 28.2 1/s
^ H
20
30
TIME
40
50
-H H
60
70
Figure 162. Volatile settleable solids concentrations vs time, We:>t Roxbury Helical (8/4/81).
-------
The Swirl removal rate averaged 6% over the first 10
minutes and averaged 4$ for the last 23 minutes. Foul
concentrations continued to decrease during tank drawdown due
to settling of materials resu I +1 njj ! " a layer of solids on the
tank bottom at the end of cperatlon.
Analysis of clear sample concentrations for the Helical
Bend Indicated an average removal rate of 31$ during the
Initial 10 minute period during which the flow Increased from
0 to 12.3 cfs (341 l/s). Removal rates averaged 10. Of over the
27 minute period of constant flow at 12.3 cfs (341 l/s). Flow
then decreased to 6 cfs (170 l/s) and the removal rate averaged
12.91. Efficiencies during these three periods are 27.0, 8.0,
and 8.9$, respectively.
Analysis of foul sewer data Indicated no removal
dur!ng operation but the drawdown period exhibited high solids
concentrations and materials remained In the unit after
operation.
13.4 Sattlaabt I Ity Experiments
tests similar to those performed at
Lancaster were conducted on samples from storm events.
Initially, Influent storm water settling characteristics
were determined for the 6/29/80 event. The wooden sleeve was
then constructed In the 87 In. (221 cm) pipe and the effects of
this Installation on Influent characteristics were again
determined during the 10/3/80 event.
Swirl foul underflow sample collection for
settleabll Ity analysis was added during the 10/25/80 event to aid
In the determination of removal efficiencies. Results showed
tha-i materials were settling In the Swirl foul discharge line
(sol Ids concentrations were less than In the Influent). In view
of this problem, the discrete foul sampler Intake was re-
positioned to within the swirl tank foul discharge port and
collection of foul settl eab I I I ty samples was discontinued.
(10/25/80 foul settleabll Ity results are not reported).
A further alteration to the wooden pipe sleeve within
the 87 In. (221 cm) was mada after the 10/29/80 event to Improve
Influent flow velocity.
Three Influent and clear settling column experiments
were conducted on 6/9/81, 6/22/81 and 6/4/81 to ascertain
suspended solids efficiency rates. When sett I eab M Ity experiments
316
-------
were simultaneously conducted on both the Swirl and the Helical
Bend (6/22/81 and 8/4/81) samples were taken from Influent waste
streams and uniformly mixed.
Analysis of Influent sett IedbJI Ity results are reported In
section 13.4.1 In which a comparison of all Influent
settleablIIty data Is made. In section 13.4.2 of settling column
analyses of Influent vs. clear are presented for 6/9/81, 6/22/81
and 8/4/81 events. Overall suspended solids efficiency rates are
computed from measured results and are compared with theoretical
efficiency rates computed using the APWA results (4).
Section 13.4.1. Influent SettleabI I Ity Characteristics
Figure 163 displays the Influent suspended solids
settleab 11 Ity curves for the 6 storm events sampled for
settling column analysis at West Roxbury. Significant
characteristics of the sett IeabI I Ity curves demonstrated by
simultaneous plotting are the range of Initial concentrations
(81 to 758 mg/l) and the varying percentages of solid fractions
within settling velocity ranges. Data values for curves shown In
Hgure 163 end other pertinent related Information are given In
Table 19.
The wide range of concentrations seen among the curves
Is the result of four factors. First, average storm event flow
rates varied between 1.2 anvj 12.3 cfs (34 and 348 l/s). Second,
the wooden pipe sleeve was constructed within the 87 In. (221cm)
pipe and then altered. Thirdly, fall events were Impacted by
filtering action of catchbaslns clogged with leaves. Fourth, the
presence of a quarry In the watershed discharging Into the storm
drainage system a variable amount of fine rock dust resulted In
variable background solids concentrations.
Influent flow and Initial settling column suspended
solids from the 6/29/80 event (prior to the sleeve Installation)
yields a mass rate of 1.29 Ib/mln (.58 kg/mln). After the
sleeve Installation a mass rate of 2.4 Ib/mln (1.1 kg/mln) was
recorded on 10/3/80. As the data shows the actual
concentration of solids was less during the second event. The
Increased storm Intensity and flow rate was sufficient to offset
this decrease Implying that the concentration decrease HAS due
to dilution effects. The Increased storm Intensity would
probably have added to the amount of material washed Into the
upstream end of the system and the new sleeve would have
reduced settling In the Influent line. The degree to which the
mass rate Increased due to each of these changes cannot be
determ I ned.
317
-------
to
t—•
00
U)
H ::
r -•
M
z -
o
m' ::
O "
n -•
M
H ••
0 ••
3 ••
\
I
I
0 •
S 21
I I I
1 > t
200
H 1 1 1-
300 400 500
SUSPENDED SOLIDS
700
Figure 163.
600
mg/L
1 cm/s - 0.033 ft/s.
Influent suspended solids settleabillty characteristics, W. Roxbury.
800
-------
TABLE 19
INFLUENT SUSPENDED SOLIDS SETTLEABILITY ANALYSES SUMMARY
Settl Ing
Velocity
cm/see
5.0
0.9
0.5
0.1
0.05
0.01
6/29/80
Inf. Cone.
my/I
.
155.0
148.2
131.0
120.1
93.6
10/3/80
Cone.
rag/ 1
.
104.0
103.0
99.5
94.2
80.3
10/25/80 6/9/81 6/22/81 8/4/81
Cone.
mg/l
.
81 .0
79.8
75.7
73.4
60.3
Cone.
mg/l
758.0
750.0
743.0
685.0
635.0
450.0
Cone.
mg/ 1
544.0
490.0
470.0
375.0
320.0
163.0
Cone.
mg/ 1
283.0
230.0
225.0
155.0
125.0
61.0
Event Key
B
B
Flow (cfs)
Col. Initial
2.2
156
6
107
4.5
81.58
1.2
759
2.2
552
12.3
283
Cone. (mg/L)
Mass Rate
( Ib/mln)
Background
Cone, (mg/l)
Adjus. Inlt.
Cone, (mg/l)
Adjus. Mass
Rate Ib/mln
Sam p. Time
for Sett, (mln)
Di screte Samp.
Cone, (mg/l)
8* Low value due to leaves In catchbaslns filtering flow
88 Based on discrete sample concentrations
888 Based on discrete sample cone.less background cone.
A - Pre Sleeve/ B - Post Sleeve/ C - Post Sleeve Modification
I cm/sec - 0.03 ft/sec"; I cfs = 28.3 l/s, 1 kg/m1n. = 0.454 Ib/m1n.
1.3
93.6
62.4
0.5
0-15
125
2.4
80.3
26.7
0.6
0-20
110
1.4
60.3
21.2
0.4
0-15
60
3.4
450
308
1 .4
45-55
650
4.5 12.7
180 61
372 222
(3(3(359)
3.0 10.0
(3(3(3 (16.1)
0-20 0-7
498 420
319
-------
The next event on 10/25/80 Included a moderate flow
rate yet the mass rate was typical of pre-sleeve Installation
events. The low Influent concentration was deemed attributable
to leaf buildup In the catchbaslns feeding the system. These
accumulations tended to filter out much of the solids cr.tar I ng
the catchbaslr. prior to the Influent line.
After this event the catchbaslns were cleaned and the
wooden pipe sleeve was altered to Increase the flow velocity.
During the next two events on 6/9/81 and 6/22/81, low flow rates
still produced high solids concentrations Indicating good
drainage area washout and minimal settling In the Influent feed
11nes.
Flows were substantially higher during the last
event monitored 8/4/81. The low solids concentration was due to
dilution effects since the mass flow rate Increased. Mass
rates were based on settleab11 Ity sample concentrations which
were typically similar to discrete sample values with the
exception of the 8/4/81 data as discussed below.
The shape of the Influent sett IeablI Ity curves shown
In Figure 163 for the three 1981 events sampled are dissimilar
Indicating that the percentages of solids at the various
settling velocities varied during each of the events. Variations
In total concentrations were caused by the presence of light
weight solids materials causing a variable background
concentration. Background material has been traced to a quarry
dust residue entering the sewer above the fjclllty rrom a
local rock quarry In the West Roxbury area.
This fine material has a settling velocity less than
that accounted for In the two hour settleablIIty test, that Is,
0.003 ft/sec (0.01 cm/sec), yet greater than that roughly defined
by the settleable solids determination test, that Is, 1 x 10~5
ft/sec (0.003cm/sec). ft Is therefore not removed In the
settling column but Is considered settleable solids.
SettleablIIty sample background solids concentrations, loosely
defined as the material with a settling velocity less than 0.003
ft/sec (O.OIcm/sec) were 450mg/l for the 6/9/81 event, 180 mg/l
for the 6/22/81 event and 61 mg/l for the 8/4/81 event,
respectively.
Since the quarry material could be considered non-
representative of typical stormwater loads, the column Initial
concentrations were reduced by these background values
resulting in the adjusted Initial concentrations and mass rates
I r. Table 19. The adjusted Influent solids concentration values
are quite similar and the mass flow rates are nearly
proportional to the flow rate for first two 1981 events. For the
320
-------
8/4/81 event, similar mass rates can be derived by using solids
concentration obtained from the discrete sampling (as opposed to
the settling column sampling technique). As next explained,
solids concentrations obtained by discrete sampling techniques
are probably more reliable*
Comparison of these values to the shapes of the
settleabI I Ity curves Indicates a certain trend. The 6/22/81
sample has the highest Initial concentration. Of the three 1981
events these data also reflected the greatest amount of material
with a settling velocity above 0.03 ft/sec (I.Ocm/sec). The 6/9/
81 event had a lower adjusted concentration and a low mass flow
rate due to the low flow rate. This event also contained ilttle
or no material with a settling velocity above 0.03 ft/sec
(1.0 cm/sec). The lack of material of this type Is probably due
to the effects of low flow on the degree of street washdown and
settling within catchbaslns. In addition, samples were
collected late In the event when most heavy material may have
already past through the system (see table 19).
Data from the 8/4/81 event demonstrate half the
material In the range above 0.03 ft/sec (1.0 cm/sec) as seen In
the 6/22/81 event. Since the flow rate was six times as much the
actual solids mass rates of materials above 0.03 ft/sec (1.0
cm/sec) were much higher. This would be expected considering that
the higher flow rate of the 8/4/81 would produce a greater mass
of solids of al! sizes. In the case of the 8/4/81 data
discrete sample concentrations were 1.5 times the settleablI Ity
sample values. It Is suspected That the methods used for
large volume sett IeabI IIty sampling are limited In the ability
to deliver representative samples. Discussion of sampling
techniques and Implications relating to sett Ieabl I Ity results
are Included In the next section.
It Is evident from Inspection of the suspended solids
settling curves In Figure 163 that the stormwater solids
contaminants from the urbanized watershed In West Roxbury, Boston
generally have lower settling velocities than the "proposed"
solids settling velocity curves used In the formulation of the
Swirl Regulator/Concentration design documents (4) (see Figure 6,
Chapter 4).
Section 13.4.2 £±il£ J.fiJLfi^ deter m 1 nat I on, MS I n.g
Sett I Bab IIIty Results
6/9/B1 Event
Figure 164 displays the Influent and clear Swirl
Concentrator sett IeabI I Ity characteristics for the 6/9/81
event. Overall the 6/9/81 event Indicated a suspended solids
321
-------
ro
ro
m
H
r
»H
z
n
m • ..
o *"
n
M
H -•
0
3
\
8
S
S
*
«
INFLUENT
CLEAR
200 400
SS CONCENTRATIONS
600
1 cm/s.i
800
0.033 ft/s
Figure 164.
Influent and clear suspended solids settleablHty characteristics,
W. Roxbury Swirl Concentrator/Regulator (6/9/81).
-------
removal efficiency of 28.6} (calculations given In Table 20).
Tiie Swirl Concentration efficiency Is 21.4$ considering solids
«lth settling velocities down to 0.015 cm/sec (0.003 ft/sec). The
additional Increment of 7.2$ (Indicated In Table 20) Is
attributable to the 55 mg/l differential associated with the
lower end of the two settling velocity curves shown In Figure
164. The theoretical efficiency 's 8.1$ with calculations In
Table 21. The low overall efficiency Is due to the lack of
removal (as expected) of the heavy background quarry solids
concentrations. Since this material accounted for 60$ of the
Initial solids the 28.6$ actual overall removal could actually be
stated as 71$ removal of the non-background material.
Based on the breakdown of material by settling
velocities seen In Table 20, no removals occured above 0.006
ft/sec (0.02 cm/sec) which Indirectly Implies that there was n«
material with settling velocities greater than this value In
the Influent flow. The fact that the plots In Figure 164 and In
Table 20 Indicate minimal material above this value Is due to the
"bast curve" fitting techniques used. Due to limitations of
sample mixing, sampling techniques and analytical techniques
some variation in resultant solids values Is expected. When
the actual amount of solids In a certain range Is similar to
the typical variation, exact values cannot be determined. In
the case of the 6/9/81 Influent sett IeabI I Ity sample It
appears little or no material with a settling velocity greater
than 0.006 ft/sec (0.02 cm/sec) was present. Possblle causes
of this are discussed at the end of this section.
6/22/8 I Event
Sett Ieab11 Ity tests for the 6/22/81 event Include both Swirl
Concentration and Helical Bend units. Figures 16? through 168
display the sett IeablI Ity characteristics of ihe suspended and
volatile suspended solids for the Swirl and Helical Bend,
respectively.
Swirl data Indicate an overall removal rate of 4.5$.
Efficiency calculations are provided In Table 22. If the two
settling curves shown In Figure 165 coincided at the lower
settling rates (less than 0.01 cm/sec or 0.0003 ft/sec) then the
efficiency would be 22.5$. Based on the Influent flow rate and
solids concentrations the theoretical removal rate Is 23.3$.
(Table 23) For this event 10$ of the Influent solids had a
settling velocity greater than 0.03 ft/sec (0.1 cm/sec) and
It would be expected that the Swirl would remove almost =11 of
this m aterI a I.
Comparison of the Swirl Influent a .1 d clear sample
sett IeabI I Ity curves Indicates several Irregularities. The
323
-------
TABLE 20
SWIRL EFFICIENCY BASED ON SETTLE*3ILITY MEASUREMENTS, 6/9/81
Flow Rate » 1.2 cfs (34 l/s)
Set. Vol. I n f Clear
Region Set.Vel. TSS TSS Inf. Clear
cm/sec Cone. Cone.
Change Change
mg/l mg/l mg/l mg/I * ** i 99
>5.0 758 540
1 5 5 0.65 0.65 0 0
4.0 753 535
2 55 0.65 0.65 0 0
0.51 748 530
3 10 10 1.3 1.3 0 0
0.18 738 520
4 58 40 7.6 5.3 2.3 31.0
0.09 680 480
5 45 20 5.9 2.6 3.3 55.5
O.C5 635 460
6 55 20 7.2 2.6 4.6 63.6
0.03 580 440
7 110 25 14.5 3.3 11.2 77.0
.015 470 415
plus 1.2% Dlfferentral Below 21.4
0.015 en/sec (see Figure 164) 7.2
Actual Efficiency 28.6$
* % Inf. cone, change relative to Inf. starting cone.
**% clear cone, change relative to Influent starting cone.
%% Removed of Initial Influent core.
%%% Removed within set. vel. region.
324
-------
TABLE 21
THEORETICAL SWIRL EFFICIENCY, 6/9/81
flow rate of 1.2 cfs (34 L/s)
Region Sett.Vel. Inf. % of Total Model 'lodel Avg. % % Rent
TSS TSS In Sett. Vet. % Removal In
Cone. Region Rem. In Region Region
cm/sec mg/L cm/sec
7.5 758
0.65
4.5 . 100
100 0.65
4.0 753
0.51 748
0.18 738
0.09 680
0.054 635
0.03 580
.015 470
0.65
1.3
7.6
5.9
7.2
14.5
2.4
0.3
0.1
100
80
45
0.053 23
0.03 15
0.017 <10
0.008 <10
90 0.60
62.5 0.80
34 2.60
19 1.10
12.5 0.90
<10 1.40
1 cm/sec a 0.03 ft/sec
Theoretical Efficiency 8.1$ •
325
-------
ro
en
m M--
H
r
M
2
o
m« ..
o h
n
H ••
o -•
3 ••
\ •-
0
ft
0 •
B
INFLUENT
CLEAR
I I
I I
I I
I I
Figure 165.
200 300 400 500
SUSPENDED SOLIDS mg/L
1 cm/s = 0.033 ft/s
Influent and clear suspended solids settleablllty characterlestlcs,
W. Roxbury Swirl Concentrator/Regulator (6/22/81),
i i
600
-------
CO
ro
W
m
H
H
r
M
z
o
m
o
n
M
H
Q::
0
3
e •
\/
Q-
INFLUENT
CLEAR
20
40 60
VSS CONC.
Cmg/L>
80
1 cm/s
100
0.033 ft/s
120
Figure 166.
Influent and clear Volatile suspended solids settleabillty characteristics,
W. Roxbury Swirl Concentrator/Regulator (6/22/81).
-------
ro
00
H ::
r --
1-1
2 -•
O
m • .
r M:
o ::
n --
/\
o
3
\
t)
fi
0
s
^ INFLUENT
« CLEAR
> I
I l
H 1 1 1-
200 300 400 500 300
SUSPENDED SOLIDS mg/L 1 Cm/s = 0.033 ft/3
Figure 167. Influent and clear suspended solIds settleabillty characteristics,
U. Roxbury Helical Bend Regulator (6/22/81).
-------
«*>
ro
\o
U)
m
H
H
r
w
z
o
m-
TM:
D :
n
H
0
Q0
>-»
Figure 168.
40 80
VSS CONC.
+ INFLUENT
« CLEAR
•4-
80
-l-
1 on/s
100
0.033 ft/s
120
Influent and clear volatile suspended solids* settleablHty characteristics,
W. Roxbury Helical Bend Regulator (6/22/81),
-------
TABLE 22
SWIRL EFFICIENCY BASED ON SETTLEAB IL ITY MEASUREMENTS. 6/22/81
flow rate = 2.2 cfs (62.3 l/s)
Region Set.Vel Inf. Clear Inf. Clear
TSS TSS Cone. Cone.
change change
cm/sec rag/ I mg/l rag/ I rag/ I • «•
1 5 ' 552 530
32 30 5.8 5.4 .4 7
2 2 520 500
25 20 4.5 3.6 .9 20
3 1 495 485
25 20 4.5 3.6 .9 20
4 .5 470 460
95 55 17.2 9.9 7.3 *'2
5 .1 375 405
55 25 9.9 4.5 5.4 54
6 .05 320 380
157 115 4.5 20.8 7.6 32
7 .01 163 265
Differential Remaining 22.5%*
Below. 01 cm/sec -18.0
Actual Efficiency 4.5>
*$ Inf. cone, change relative to Inf. starting cone.
**$ clear '•one. change relative to Inf. starting cone.
&% remove*! of Initial Influent cone.
%% % removed within sett, velocity region
330
-------
TABLE 23
THEORETICAL SWIRL EFFICIENCY, 6/22/81
('Sow rats 2.2 cfs (62.3 L/s)
Region Sett.Vel. Inf. % of Total Model Model Avg. % % Rem
SS SS In Sett. Vel. % Removal !n
Cone. Region Rem. In Region Region
cm/sec mg/L cm/sec
552
2 520
1 495
0.5 470
0.1 375
0.05 320
0.01 1 63
I cm/sec • 0.03 ft/sec
5.8
4.5
4.5
17.2
9.9
28.4
2.8 >100
1.1
.53
94
78
.30 63
.053
.03
10
97.0 5.8
86 .0 3.9
70.5 3.2
38 .0 6.5
11.5 1.08
10.0 2.8
.0053 10
Theoretical Efficiency 23.3*
331
-------
Influent solids tend to have higher settling velocities than
the clear sample material. Although the Influent was
Initially more concentrated, the clear sample had a greater
ass cunt of solids wltnln the lower sett I eeb I I I ty ranges. One
possible explanation Is the breakdown of larger particles within
the unl t.
Volatile fractions of the Swirl Influent and clear
samples were 21$ and 13$ respectively. Implying an overall
removal of 8$ on a volatile basis alone. Given the overall
removal rata of 4.5$ there appears to have been a 3.5$
Increase In non-volatile material. Breakdown of larger particles
would not account for this effect. This Increase could be the
result of solids accumulations within the Swirl unit.
Since there vas a definite decrease In material with
settling velocities greater than .06 ft/sec (2 cm/sec) the
Swirl can be stated to be effective In this range. For this
event 22$ of this material was removed.
Simultaneously the Helical Bend provided an efficiency of
43$ (see Table 24). The total overall efficiency rate noted In
Table 24 Is 42.9$. Above 0.01 cm/sec (0.003 ft/sec) the removals
divided among the designated settling velocity regions on Table
24 totcl 34.2$. Below 0.01/sec settling velocity region the unit
achieved an additional 8.7$ removal efficiency (also see lower
portion of two settling curves on Figure 167). Volatile fractions
In the clear sample were nearly Identical to those within the
Swirl clear sample. Eight $ of the organic material was removed
Implying 35$ removal of Inorganic material. There appeared to be
minimal bulld-i'p of material In the unit during this part of
the event. The highest removals occured In the highest settling
velocity ranges although good removals were demonstrated In each
range.
8/4/81 Evant
Figures 169 and 170 display thn Swirl Concet. rrator and
Helical Bend Influent and clear suspended solids settling curves.
Inspection of these Figures Indicate no removal based on the
sett IeabI I Ity analysis. Initial clear sample, concentrations
for both the Swirl and Helical Bend were higher than the
Influent sample.
Considering the sampling technique: used, the data
shown In Figures 169 and I/O can possibly be explained as
follows. During past events flow rates were between one-
tenth to one-half the 12 cfs (3>8 l/s) which occured during the
8/4/81 event.Sett IeabM Ity samples were collected In the
Influent line just downstream of the Palmer Bowlus flume where
332
-------
TABLE 24
HELICAL BEND EFFICIENCY BASED ON SETTLEABILITY MEASUREMENTS
6/22/81
Region Set.Vel
cm/sec
1 5
2 1
3 .5
4 .1
5 .05
6 .01 *
Inf.
SS Cone.
mg/l
552
495
470
375
320
163
Cl ear
SS Cone.
mg/l
315
310
300
245
210
Inf.
Cone.
change
mg/l
57
25
95
55
157
1 1 5
8.7% Differential
RA 1 AM - _fl 1 ft
Clear
Cone.
change
mg/l »
5 10.3
10 4.5
55 17.2
35 9.9
95 28.4
Remal ns
n / e &f
*** 6 @(j
0.9 9.4 92
1.8 2.7 60
10.0 7.2 42
6.3 3.6 36
17.2 11.2 35
34.2
8.7
Actual Efficiency 42.9*
* $ Inf. cone change relative to Inf. starting cone.
*» % Clear cone, change relative to Inf. starting cone.
i % Removed of Initial Influent cone.
66 % Removed within set. velocity range 1 cm/sec " 0.03 ft/sec
333
-------
CJ
CJ
W
m
-i
H
r
M
2
n
m- .
TM:
o :
n •
H ••
0
3
INFLUENT
CLEAR
0
vx
Q-
200 300 400
SUSPENDED SOLIDSemg/L
•y ^
500 600
1 cm/s = 0.033 ft/s
Figure 169. Influent and clear suspended solids settleablllty characteristics,
W. Roxbury Swirl Concentrator/Regulator (8/4/81).
-------
CO
OJ
en
m
H ::
r -•
z
o
m. j.
o *"
n
H --
o
3
\
8
ID
0 •
^. INFLUENT
« CLEAR
S 2
-f-
200 300 400
SUSPENDED SOLIDS mg/L
500
1 cm/s =
600
0.033 ft/s
Figure 170.
Influent and clear suspended solids settleability characteristics,
W. Roxbury Helical Bend Regulator (8/4/81),
-------
the flow depth was minimal. A four gal. (17.5 I) bucket was
used to collect samples. Under previous storm conditions
Inserting the bucket to the Invert of the Influent line was no
problem. During the a/4/81 event the Increase In flow depth and
velocity appears to have acted to fill the sample bucket before
It could be submerged to the Influent line Invert. Samples
collected would thus not contain representative bed load
materials Discrete samples were not affected by the
Increased flow and velocity as the methods used allowed
complete submergence of l-l sample bottles. (Filling while
raising the I/I bottles tended to yield a more cross- sectional
composite sample).
Comparison (see Table 19) of Initial settling column
suspended solids concentrations with concentrations of samples
taken for discrete analysis (see section 13.3) Indicates that for
previous events discrete Influent sample and settleable sample
concentrations were similar. For the 8/4/81 event the
Influent discrete suspended solids concentration equalled 420
mg/l while the Initial or s1! art Ing concentration of the Influent
settling column test was 283 mg/l, much higher. If the
sett IeablI I ty sample concentration had equaled the discrete
sample concentration of 420 mg/l the Swirl efficiency would have
been cat cul ated to be 11$ while the efficiency for The He I leal
Bend would have approached 20%.
13.5 Performance Summary
Table 25 summarizes the suspended solids performance data
for the Swirl Regulator and Helical Bend units. Suspended solids
efficiency rates are comparable for both units. Efficiency rates
are low for both units and do not seem to be related to flow
rate. It appears that both units are able to achieve a low degree
of primary-type suspended solids removal of stormwa*er related
contain I nants.
13.6 Pya Studies
Several dye studies were performed on the Swirl unit
during selected storm events. Times for the dye to complete one
revolution for varying flow rates are as follows:
336
-------
TABLE 25
Summary: West Roxbury Evaluation
Suspended Solids Removal and Efficiency Rates
Date
t iow
cfs
Removal
Rate %
Efficiency Sett 1 eabl 1 1 ty
% Results
Swirl Concentrator
6-29-80
7-29-80
10-3-80
10-25-80
6-9-81
6-22-81
8-4-81
2.0
3.0
6.0
3.0
0.8
3.0
2.2
2.0
6.0
12.0
21.0
35.3
36.0
29.7
34.0
27.0
27.0
33.0
9.5
5.8
8.2 *
27.5
32.0 **.§
21.4 »»•
2.75
19.0 «•» 21.4 - 28.6 %
16.0 **» 4-5 - 22.5$
21 .0
5.5 6
5.7
6-29-80
7-29-80
10-3-80
10-25-80
6-9-81
6-22-81
8-4-80
1 .5
2.5
4.5
2.5
.8
.8
2.0
6.0
12.3
6.0
Hel
23.0
36.5
16.0
36.5
41.2
50.5
25.5
31.0
10.0
12.9
leal Bend
6.0 *
26.5
10.4 «
26.5 ••
10.0 »
19.3
13.5 34.2 - 42.9$
27.0 8
8.0
8.9
* Influent *« Influent & Foul *** Influent & Clear
9 Sampling problem I cfs = 28.2 l/s
337
-------
Date Flow (efsl Tlma (sec)*
4-4-80 .8 150
7-29-91 .9 139
4-4-80 1.0 135
6-22-61 1.5 90
4-4-80 2.2 74
8-4-81 12.3 10
* Time to circulate through unit for 1 complete
revelution
Inspection of these results Indicates that as the flow
Into the unit Increased, the tine for a complete revolution
of the liquid Is reduced. Flo* Into the Swirl was expected to
enter the chamber tangentially at the Inlet ramp and circulate
throughout the unit from the bottom progressively working
Its way to the surface. ObservalIons during dye studies
Indicate that the flow always rises Immediately to the
surface as It exits the Inlet ramp and then circulates
throughout the chamber.
A dye study was performed on the Helical Bend
during the June 22. 1981 event with flow Into the unit
varying between l.Scfs to lOcfs (42.5 l/s to 283 l/s). The purpose
of the study was to visually observe the flov path In the
unit during an event. Flow velocity measurements throughout the
unit were not taken. All observations were of a visual nature.
Nearly perfect plug flow conditions were observed.
338
-------
REFERENCES
I. Scholl, J., at al. Report to Congress on Combined Sewer
Overflow In the United States.U.S.EPA Report No.EPA-430/9-
78-006. June 1978.
2. Sm I sson,B., "Design, Construction and Performance of Vortex
Overflows", Proc., Symp. on Storm Sewage Overflows, Inst..
Civil Eng. (G.B.), 1967.
3. American Public Works Association. The Swirl Concentrator
as a Combined Sewer Overflow Regulator Facility. U.S. EPA
Report No. EPA-R2-72-008, NTIS No. 214 134, September 1972.
4. Sullivan, R.H., et al. Relationship Between Diameter and
Height for the Design of a Swirl Concentrator as a Combined
Sewer Overflow Regulator. U.S. EPA Report No. EPA-670/2-74-
039, NTIS No.
5. Sullivan, R.H., et al, - The Swirl Concentrator es a Grit
Separator Device, EPA Report No. EPA-670/2-74-026, NTIS No.
PB 233 964, June 1974.
6. Heaney, J.P., Huber W.C., et al. Storm Water Management
Model: Refinement^, Testing and Decision-Making. University
of Florida, Department of Environmental Engineering
Sciences, June 1973.
7. Sullivan, R.H., et al - Field Prototype Demonstration of
Swirl Degrltter. U.S. EPA Report No. EPA-600/2-77-185, NTIS
No. PB 272 068, September, 1977.
8. Drehwlng, F.J. et al, "Disinfection/Treatment of Combined
Sewer Overflows" Syracuse, New York, U.S. EPA Report No.
EPA-600/2-79-134.
9. Drehwlng, F.J., et al. Combined Sewer Overflow Abatement
Program - Rochester, New York - Volume II. Pilot Plant
Evaluations. U.S. EPA Report No. EPA-600/2-79-031Ib, August
1979.
10. Sullivan, R.H. et al The Swirl Primary Separator:
Development and Pilot Demonstration. EPA Report No. EPA-600/
2-78-122, August, 1978.
11. Pr us-Chaci nsk I, T.M., and Wlelgorskl, J.W., "Secondary
Motions Applied to Storm Sewage Overflows, "Proceedings,
Symposium on Storm Sewage Overflows, Institution of Civil
Engineers, May 4, 1967.
339
-------
12. Sullivan, R.H., et al. The Helical Bend Combined Sewer
Overflow Regulator Report No. EPA-600/2-75-062, NT IS No. PB
250 519. Dsccsbcr, 1976.
13. Sulllva"h, R.H., et al. Design Manual Secondary Flow
Pollution Control Devices. U.S. EPA Grant No. R803157. (at
press).
14. Meridian Engineering, Inc. City of Lancaster, Pennsylvania,
Silo Demonstration project. U.S. EPA Grant No. 11023 GSC.
1973.
15. Metcalf & Eddy, Inc., Report to City of San Francisco on
Swirl Concentrator, 1978.
•
16. Dalrymple, R.J., et al. Physical and Settling
Characteristics of Participates In Storm and Sanitary
Wastewaters. U.S. EPA Report No. EPA-670/2-73-01 1 , NTIS No.
PB 242 - 01, Apr!I, 1975.
17. Personal Communication, Fred Parkinson, La Salle
Laboratories with Richard Field, U.S. EPA, February, 1978.
18. Metcalf & Eddy, Inc. Report to City of Saglnaw, Michigan
upon recommended Plan for Abating Pollution from Combined
Sewer Overflows. March, 1972.
19. Plsano, W.C. et al Preliminary Engineering Study for the
Control and Treatment of Combined Sewer Overflows to the
Saglnaw River, U.S. EPA Grant No. S005359, June, 1980.
.20. Plsano, W.C. et al. Facility Plan for the Control and
Treatment of Combined Sewer Overflows to the Saglnaw River.
U.S. EPA Grant No. S005359, March, 1981.
21. Plsano, W.C. et al. Design Development Document for the
Webber Street Combined Sewer Overflow Facility. U.S. EPA
Grant No. S005359, June, 1982.
22. Shelley, G.J. et al. Field Evaluation of a Swirl Degrltter,
Tamworth N.S.W. Australia. U.S. EPA Grant No. R 806706. (at
press).
23. Bergstedt, J.M. et al. Laboratory Evaluation of methods ho
Separate Fine Grained Sediment from Stormwater. U.S. EPA
Report No. EPA-600/2-79-076. July 1979.
24. Wilson, G.E., et al, "Design of Teacup Solids Separators for
Treatment of Sewer Overflows," Prog. Wat. Tech. 1978, Vol.
10, Nos. 5/0, pp. 811-820.
340
-------
25. Eutek Systems, Inc. Summary: Stormwater Teaup Performance at
the SacrajR5f! + 9 City M»'" Nastewater Treatment Facility,
Sacramento, California. August, 1978.
26. Maynard, Froud and Stevens Consulting Engineers, Pilot Plant
at Blackwell In the Long Ashton Rural District, 1967.
27. Hydro Research t'a Mexico A.C.S., Report on the Investigation
of the Technical Feasibility of Treating and Recycling Water
at the Tubacero Factory In Monterrey, Mexico, June, 1981.
28. Hydro Research de Mexico A.C.S., Report on the Investigation
of Treatment of Raw Sevage using the Dynamic Separator at
Aqua Industrial, Monterrey, Mexico, June, 1981.
29. R.P.M. Smlsson, Hydro Research & Development (UK) Ltd. with
W.C. Plsano, Environmental Design & Planning, Inc.,
December, 1981.
3U. "Applications of Storm Water Management Models, January 5-9,
1976". A Short Course Sponsored by the U.S. Environmental
Protection Agency otorm & Combined Sewer Section- In
Cooperation with the Department of Civil Engineering,
Environmental Engineering Program, University of
Massachusetts, & Division of Continuing Education.
31. Plsano, W.C., Watson, D.S. and Aronson, 6.L. The Value of
Fl ow CalIbratlon forDeclslon - Making In I nf 11tratI on/InfIow
Stud'es. U.S. EPA Grant No. R804578. January, 1981.
32. Plsano, W.C. et al. Dry Weather Deposition and Flushing for
Combined Sewer Overflow Pollution Control. U.S. EPA Report
No. EPA-600/2-79-133. August, 1979.
33. Aronson, G.L., Plsano, W.C. and Watson D.S. Development and
Testing of Procedures for Determination of Combined Sewer
Overflow SettleeblIIty Characteristics. U.S. EPA Grant No.
R804578. April, 1981.
34. Water Research Centre, Settling Column Testing, Edinburgh,
Scotland. 1978.
35. Standard Methods for the Examination of Water and Wastewater
14th Edit'-::.. 1975 APHA-AWWA-WPCF.
36. Oceanography International Corp. U.S. EPA Approved
Alternative Method, Federal Register Vol. 43, No. 45. March
1978,
341
------- |