September, 2007
Revised April, 2009
07/30/WQPC-SWP
EPA/600/R-07/120
Environmental Technology
Verification Report
In-Drain Treatment Device
The Hydro International Up-Flo™ Filter
Prepared by
Penn State Harrisburg
Middletown, Pennsylvania
Under a Cooperative Agreement with
U.S. Environmental Protection Agency
-------
THE ENVIRONMENTAL TECHNOLOGY VERIFICATION
PROGRAM
U.S. Environmental Protection Agency
NSF International
ETV Joint Verification Statement
TECHNOLOGY TYPE:
APPLICATION:
TECHNOLOGY NAME:
TEST LOCATION:
COMPANY:
ADDRESS:
WEB SITE:
EMAIL:
UPFLOW WATER TREATMENT
IN-DRAIN TREATMENT DEVICE
UP-FLO™ FILTER WITH CPZ MIX™ FILTER MEDIA
PENN STATE HARRISBURG
HYDRO INTERNATIONAL
94 Hutchins Drive
Portland, Maine 04102
http://www.hydrointernational.biz/us/
stormwaterinquiries@hil-tech.com
PHONE: (207)756-6200
FAX: (207)756-6212
NSF International (NSF), in cooperation with the U.S. Environmental Protection Agency (EPA), operates
the Water Quality Protection Center (WQPC), one of six centers under the Environmental Technology
Verification (ETV) Program. The WQPC recently evaluated the performance of the Up-Flo™ Filter,
manufactured by Hydro International. The Up-Flo™ Filter was tested at the Penn State Harrisburg
Environmental Engineering Laboratory in Midletown, Pennsylvania.
EPA created ETV to facilitate the deployment of innovative or improved environmental technologies
through performance verification and dissemination of information. The ETV program's goal is to further
environmental protection by accelerating the acceptance and use of improved and more cost-effective
technologies. ETV seeks to achieve this goal by providing high quality, peer-reviewed data on technology
performance to those involved in the design, distribution, permitting, purchase, and use of environmental
technologies.
ETV works in partnership with recognized standards and testing organizations; stakeholder groups, which
consist of buyers, vendor organizations, and permitters; and with the full participation of individual
technology developers. The program evaluates the performance of innovative technologies by developing
test plans that are responsive to the needs of stakeholders, conducting field or laboratory tests (as
appropriate), collecting and analyzing data, and preparing peer-reviewed reports. All evaluations are
conducted in accordance with rigorous quality assurance protocols to ensure that data of known and
adequate quality are generated and that the results are defensible.
07/30/WQPC-SWP
The accompanying notice is an integral part of this verification statement.
VS-i
September 2007
-------
TECHNOLOGY DESCRIPTION
The following description of the Up-Flo™ Filter was provided by the vendor and does not represent
verified information.
The Up-Flo™ Filter is a passive, modular filtration system that incorporates multiple elements of a
treatment train into a single, small-footprint device. The Up-Flo™ Filter uses a sedimentation sump and
screening system to pretreat runoff before it flows up through the filter media, housed in one to six filter
modules, where final polishing occurs. A high-capacity, siphonic bypass safeguards against upstream
ponding during high-flow events. The siphon also serves as a floatables baffle to prevent the escape of
floatable trash and debris from the Up-Flo™ Filter chamber.
The Up-Flo™ Filter is self-activating and operates by simple hydraulics. Challenge water enters the
chamber from an inlet pipe or an overhead grate and flows into the sump region where gross debris and
coarse grit are removed by settling. Runoff continues to fill the chamber until there is enough driving
head to initiate flow through the filter media. At this point, the water flows up through the angled screen
into the filter module. In the filter module, flow passes up through the filter media and is conveyed to the
outlet module via the flow conveyance channel. Flows in excess of the filtration capacity are discharged
directly to the outlet module by the siphonic bypass. The siphon also serves as a floatables baffle to
prevent the escape of buoyant litter and debris. The Up-Flo™ Filter is equipped with a drain-down
mechanism to ensure that the filter media sits above the standing water level during no-flow conditions, to
prevent anoxic conditions that could promote bacterial growth in the filter media and the release of
harmful leachates. As flows subside, water slowly drains out of the chamber through four small drain-
down ports located at the base of the outlet module. The drain-down ports are covered with a layer of
filter fabric to provide treatment to the drain-down flows.
Performance of a regularly maintained Up-Flo™ Filter should provide removal of over 80% of total
suspended solids (TSS) from challenge water runoff. It will also remove a portion of metals, organics and
other pollutants commonly found sorbed to the surface of suspended sediment particles. Each filter
module filled with the CPZ Mix™ will have a flow rate of 20-25 gpm when the water level in the
chamber provides 20 in. of driving head. Water will continue to be filtered up through the filter media
until the water level in the chamber falls to zero inches of driving head. When the inflows exceed the
filtration capacity, the excess flows will discharge through the bypass siphon directly to the outlet
module.
VERIFICATION TESTING DESCRIPTION
Methods and Procedures
The testing methods and procedures employed during the study were outlined in the Test Plan for Hydro
International, Inc. Up-Flo™ Filter for Stormwater Treatment (February 2006). The Up-Flo™ Filter was
installed in a specially designed testing rig to simulate a catch basin receiving surface runoff. The rig was
designed to provide for controlled dosing and sampling, and to allow for observation of system
performance.
The Up-Flo™ Filter was challenged by a variety of hydraulic flow and contaminant load conditions to
evaluate the system's performance under normal and elevated loadings. The test conditions are
summarized in Table 1. Additional tests were conducted at the vendor's request to determine the media's
sediment removal capabilities with challenge water consisting of only sediments and nutrients (no
hydrocarbons) at continuous flow. The results of these tests will be published in an addendum at a later
time.
07/30/WQPC-SWP The accompanying notice is an integral part of this verification statement. September 2007
VS-ii
-------
Table 1. Test Phase Summary
Phase
I
II
III-l
III-2
III-3
IV
and Flow Condition
Intermittent Flow
Contaminant Capacity
Hydraulic Capacity, Clean Water
Hydraulic Capacity, Synthetic
Wastewater
Hydraulic Capacity, Spiked
Wastewater
Contaminant Capacity at High
Hydraulic Throughput
Flow
1 1 gpm, 15 min on, 15
min off
16 gpm continuous
10 to 45 gpm, increased
in 5 gpm increments
10 to 45 gpm, increased
in 5 gpm increments
10 to 45 gpm, increased
in 5 gpm increments
32 gpm continuous
Loadings
Normal
Normal
None
Normal
Spiked
(4X)
Normal
Test Duration
40 hr
Continue until
exhaustion
15 min at each
flow interval
15 min at each
flow interval
15 min at each
flow interval
Continue until
exhaustion
A synthesized wastewater mixture containing petroleum hydrocarbons (gasoline, diesel fuel, motor oil,
and brake fluid), automotive fluids (antifreeze and windshield washer solvent), surfactants, and sediments
(sand, topsoil and clay), was used to simulate constituents found in surface runoff from a commercial or
industrial setting. Influent and effluent samples were collected and analyzed for several parameters,
including TSS, suspended sediment concentration (SSC), total phosphorus (TP), and chemical oxygen
demand (COD). Complete descriptions of the testing and quality assurance/quality control (QA/QC)
procedures are included in the verification report.
PERFORMANCE VERIFICATION
System Installation and Maintenance
The Up-Flo™ Filter was found to be durable and easy to install, requiring no special tools. Maintenance
on the system during testing consisted of replacing the filter media bags, and removing sediment and
water collected in the sump. Maintenance took approximately 30-45 minutes, with the most difficult
activity being removal of the filter media bags, due to their size and weight.
Hydraulic Capacity
The hydraulic capacity of the Up-Flo™ Filter was determined using clean water (Phase III-l), synthetic
wastewater (Phase III-2), and synthetic wastewater with spiked constituents (Phase III-3). Capacity was
evaluated as a function of influent and effluent flow rates, and water levels in the sump. The testing
determined the effluent flow rates were comparable to the influent for all flow rates tested, up to and past
the point where the bypass was activated. The hydraulic capacity results are expressed graphically in
Figure 1.
An Up-Flo™ with new filter media can accept a hydraulic flow of up to approximately 30 gpm with no
bypass, depending on the concentration of contaminants in the wastewater. At flows greater than 30 gpm
the water elevation in the sump approaches the bypass siphon elevation, and a portion of the influent flow
exits the system as untreated bypass. The maximum treated flow decreases as the filter media trap
contaminants, preventing water from flowing through the filter bags. This was particularly evident with
the Phase III-3 (spiked contaminant loadings), where the effluent flow diminished prior to eventually
reaching bypass conditions.
07/30/WQPC-SWP
The accompanying notice is an integral part of this verification statement.
VS-iii
September 2007
-------
Flow Rate - Phase I
50
40 -
30 -
&
ra
a:
20 -
10 -
influent
III-2
III-3
d
10 15 20 25 30 35
Flow Conditions (GPM)
40 45
Figure 1. Comparison of influent versus effluent flow rates for Phase III hydraulics testing.
Contaminant Removal
Table 2 summarizes the influent and effluent constituent concentrations and the respective removal
efficiencies for the Phase I (intermittent flow) and Phase II (continuous flow tests). During both of these
tests, the flow was held constant at 11 gpm for Phase I and 16 gpm for Phase II, both of which are less
than the Up-Flo™ Filter's 20 gpm rated capacity. These tests were done consecutively, and were
completed when filter media exhaustion or blinding was observed. During testing, the filter media was
blinded off by contaminant loading prior to breakthrough occurring. In general, the effluent constituent
concentrations remained constant throughout testing.
Table 2. Up-Flo™ Filter Treatment Efficiency Summary for Phase I and Phase II Tests
Influent Concentration
Results (mg/L)
TSS
ssc
TP
COD
Mean
136
147
47
157
Median
112
130
44
134
Max.
492
555
183
523
Min.
<5
<5
0.6
60
Effluent Concentration
Results (mg/L)
Mean
36
39
38
63
Median
30
30
38
65
Max.
100
108
81
89
Min.
9
<5
0.6
33
Removal Efficiency ("/o)1
Mean
73
74
19
60
Median
73
77
14
51
Max.
92
99
91
88
Min.
-1,280
-480
-530
-3.3
1. Mean and median removal efficiencies are calculated using the calculated mean and median influent and effluent
concentrations, while maximum and minimum removal efficiencies are evaluated from the paired sample data points.
The median sediment removal efficiency is 73% and 77% for TSS and SSC, respectively, which is
slightly below the vendor's 80% sediment removal efficiency performance claim. The Up-Flo™ Filter
was also shown to be capable of reducing TP and COD, demonstrated by median removal efficiencies of
14% and 51%, respectively.
07/30/WQPC-SWP
The accompanying notice is an integral part of this verification statement.
VS-iv
September 2007
-------
Media Blinding/Bypass
During the Phase II and Phase IV tests, the testing organization observed that when the filter media
reached capacity, it would shift within the filter module. This shift opened a preferential pathway in the
corner of the filter module for water to pass through the system without passing through the filter media.
This failure mechanism was not anticipated by the vendor. The vendor indicated that the Up-Flo™ Filter
would fail as the filter bags clog, forcing a rise of the water level in the tank to an elevation that would
eventually reach the bypass siphon and flow out through the bypass.
Quality Assurance/Quality Control
NSF personnel completed a technical systems audit during testing to ensure that the testing was in
compliance with the test plan. NSF also completed a data quality audit of at least 10% of the test data to
ensure that the reported data represented the data generated during testing. In addition to QA/QC audits
performed by NSF, EPA personnel conducted an audit of NSF's QA Management Program.
Original signed by Original signed by
Sally Gutierrez October 15, 2007 Robert Ferguson October 3, 2007
Sally Gutierrez Date Robert Ferguson Date
Director Vice President
National Risk Management Research Laboratory Water Systems
Office of Research and Development NSF International
United States Environmental Protection Agency
NOTICE: Verifications are based on an evaluation of technology performance under specific,
predetermined criteria and the appropriate quality assurance procedures. EPA and NSF make no expressed
or implied warranties as to the performance of the technology and do not certify that a technology will
always operate as verified. The end user is solely responsible for complying with any and all applicable
federal, state, and local requirements. Mention of corporate names, trade names, or commercial products
does not constitute endorsement or recommendation for use of specific products. This report is not an NSF
Certification of the specific product mentioned herein.
Availability of Supporting Documents
Copies of the Protocol for the Verification ofln-Drain Treatment Technologies, April 2001, the
verification statement, and the verification report (NSF Report Number 07/30/WQPC-SWP) are
available from:
ETV Water Quality Protection Center Program Manager (hard copy)
NSF International
P.O. Box 130140
Ann Arbor, Michigan 48113-0140
NSF website: http://www.nsf.org/etv (electronic copy)
EPA website: https://www.epa.gov/etv (electronic copy)
Appendices are not included in the verification report, but are available from NSF upon request.
07/30/WQPC-SWP The accompanying notice is an integral part of this verification statement. September 2007
VS-v
-------
Environmental Technology Verification Report
In-Drain Treatment Device
The Hydro International Up-Flo™ Treatment
Device
Prepared by:
Penn State Harrisburg
Middletown, Pennsylvania 17057
Under a cooperative agreement with the U.S. Environmental Protection Agency
Raymond Frederick, Project Officer
ETV Water Quality Protection Center
National Risk Management Research Laboratory
Water Supply and Water Resources Division
U.S. Environmental Protection Agency
Edison, New Jersey
September 2007
Revised April 2009 with Supplemental Vendor Testing (Chapter 6)
-------
Notice
This document has been peer reviewed and reviewed by NSF and EPA and recommended for
public release. Mention of trade names or commercial products does not constitute endorsement
or recommendation by the EPA for use or certification by NSF.
in
-------
Foreword
The EPA is charged by Congress with protecting the Nation's land, air, and water resources.
Under a mandate of national environmental laws, the Agency strives to formulate and implement
actions leading to a compatible balance between human activities and the ability of natural
systems to support and nurture life. To meet this mandate, EPA's research program is providing
data and technical support for solving environmental problems today and building a science
knowledge base necessary to manage our ecological resources wisely, understand how pollutants
affect our health, and prevent or reduce environmental risks in the future.
The National Risk Management Research Laboratory (NRMRL) is the Agency's center for
investigation of technological and management approaches for preventing and reducing risks
from pollution that threaten human health and the environment. The focus of the Laboratory's
research program is on methods and their cost-effectiveness for prevention and control of
pollution to air, land, water, and subsurface resources; protection of water quality in public water
systems; remediation of contaminated sites, sediments and ground water; prevention and control
of indoor air pollution; and restoration of ecosystems. NRMRL collaborates with both public
and private sector partners to foster technologies that reduce the cost of compliance and to
anticipate emerging problems. NRMRL's research provides solutions to environmental
problems by: developing and promoting technologies that protect and improve the environment;
advancing scientific and engineering information to support regulatory and policy decisions; and
providing the technical support and information transfer to ensure implementation of
environmental regulations and strategies at the national, state, and community levels.
This publication has been produced as part of the Laboratory's strategic long-term research plan.
It is published and made available by EPA's Office of Research and Development to assist the
user community and to link researchers with their clients.
Sally Gutierrez, Director
National Risk Management Research Laboratory
IV
-------
Contents
Verification Statement VS-i
Notice iii
Foreword iv
Contents v
Figures vi
Tables viii
Abbreviations and Acronyms ix
Chapter 1 Introduction 1
1.1 ETV Purpose and Program Operation 1
1.2 Testing Participants and Responsibilities 1
1.2.1 U.S. Environmental Protection Agency 1
1.2.2 NSF International - Verification Organization (VO) 2
1.2.3 Testing Organization - Penn State Harrisburg 3
1.2.4 Vendor - Hydro International 3
1.3 Verification Testing Site 4
Chapter 2 Up-Flo™ Filter Equipment Description and Operating Processes 5
2.1 Equipment Description 5
2.1.1 Up-Flo™ Filter Components 5
2.2 Hydraulic Flow Path 6
2.3 Flow Conditions 6
2.3.1 Operating Flow Conditions 6
2.3.2 Bypass Flow Conditions 7
2.3.3 Drain Down 8
2.4 Sizing and Hydraulic Capacity 8
2.5 Test Unit Specifications and Test Setup Description 9
2.6 Up-Flo™ Filter Capabilities and Claims 12
2.6.1 System Capability 12
2.6.2 Vendor Claims 12
2.7 Performance Measures for the Verification Test 13
2.7.1 Contaminant Selection and Monitoring for Performance 13
2.7.2 System Component Operation and Maintenance Performance 14
2.7.3 Quantification of Residuals 14
Chapters Verification Testing Procedures 15
3.1 Testing Objectives 15
3.2 Test Equipment 16
3.3 Test Phases -Hydraulic Loading 19
3.3.1 Phase I — Performance under Intermittent Flow Conditions 19
3.3.2 Phase II - Determination of the Capacity of the Unit 20
3.3.3 Phase III - Performance Under Varied Hydraulic and Concentration Conditions21
3.3.4 Phase IV- Contaminant Capacities at High Hydraulic Throughput 22
3.4 Influent Characterization 23
3.4.1 Synthetic Challenge Water 23
3.4.2 Stock Solutions 24
3.4.3 Influent Characterization during the Verification Testing 25
-------
3.4.4 Solids Characterization during the Verification Testing 25
3.5 Effluent Characterization 26
3.6 Residue Management 26
3.7 Operation and Maintenance Observations 26
Chapter 4 Verification Testing Results and Discussion 28
4.1 Synthetic Wastewater Composition 28
4.2 Synthetic Wastewater Laboratory Analytical Results 31
4.3 Test Phases in the Test Plan 32
4.3.1 Phase I - Performance under Intermittent Flow Conditions 32
4.3.2 Phase II - Determination of the Capacity of the Unit 36
4.3.3 Phase III - Performance under Varied Hydraulic and Concentration Conditions 42
4.3.4 Phase IV- Contaminant Capacities at High Hydraulic Throughput 56
4.4 Phases I-IVData Summary and Discussion 60
4.4.1 Installation and Operation & Maintenance Findings 61
4.5 Summary of Findings 63
Chapter 5 Quality Assurance/Quality Control 65
5.1 Audits 65
5.2 Precision 65
5.2.1 Field and Laboratory Precision Measurements 65
5.3 Accuracy 67
5.4 Representativeness 67
5.5 Completeness 68
Chapter 6 Vendor Supplemental Testing 69
6.1 Up-Flo® Filter Modifications 69
6.2 Test Procedure Modifications 70
6.2.1 Synthetic Challenge Water 70
6.2.2 Analytical Methods 71
6.3 Synthetic Challenge Water Laboratory Analytical Results 71
6.4 Test Results 72
6.4.1 Phase I - Performance under Intermittent Flow Conditions 72
6.4.2 Phase II - Determination of the Capacity of the Unit 77
6.5 Sediment Retained in Sump 81
6.6 Test Summary and Discussion 83
Appendices 87
A Test Plan 87
B UpFlo™ Filter O&M Manual 87
C Analytical Data 87
Figures
Figure 2-1. Up-Flo™ Filter components 5
Figure 2-2. Filter module components 6
Figure 2-3. Flow path during normal operating conditions 7
Figure 2-4. Flow path during bypass flow conditions 7
Figure 2-5. Flow path during drain down conditions 8
VI
-------
Figure 2-6. Bypass water levels for standard Up-Flo™ Filter (left) and shallow Up-Flo™ Filter.
9
Figure 2-7. Up-Flo™ Filter test unit isometric view 10
Figure 2-8. Up-Flo™ Filter test unit plan view 10
Figure 2-9. Section view of Up-Flo™ Filter test unit 11
Figure 3-1. Test rig process flow diagram 16
Figure 3-2. Sediment particle size distribution graph 18
Figure 4-1. Phase I TSS influent and effluent results 33
Figure 4-2. Phase I SSC influent and effluent results 34
Figure 4-3. Phase I total phosphorus influent and effluent results 34
Figure 4-4. Phase I COD influent and effluent results 35
Figure 4-5. Phase I particle size distribution summary 35
Figure 4-6. Phase II TSS influent and effluent results 37
Figure 4-7. Phase II SSC influent and effluent results 37
Figure 4-8. Phase II total phosphorus influent and effluent results 38
Figure 4-9. Phase II COD influent and effluent results 38
Figure 4-10. Phase II PSD summary 39
Figure 4-11. Phase I and II TSS cumulative loading results 40
Figure 4-12. Phase I and II SSC cumulative loading results 40
Figure 4-13. Phase I and II total phosphorus cumulative loading results 41
Figure 4-14. Phase I and II COD cumulative loading results 41
Figure 4-15. Phase I and II PSD summary 42
Figure 4-16. Phase III Part 1 relationship between influent and effluent flow rates using clean
water 43
Figure 4-17. Phase III Part 1 tank water depth as a function of influent flow rate 44
Figure 4-18. Phase III Part 1 drawdown flow rates 44
Figure 4-19. Phase III Part 2 relationship between influent and effluent flow rates 45
Figure 4-20. Phase III Part 2 tank water depth as a function of influent flow rate 46
Figure 4-21. Phase III Part 3 relationship between influent and effluent flow rate 46
Figure 4-22. Phase III Part 3 tank water depth as a function of influent flow rate 47
Figure 4-23. Comparison of influent versus effluent flow rates for Phase III hydraulics testing.48
Figure 4-24. Phase III Part 2 TSS influent and effluent results 51
Figure 4-25. Phase III Part 2 SSC influent and effluent results 51
Figure 4-26. Phase III Part 2 total phosphorus influent and effluent results 52
Figure 4-27. Phase III Part 2 COD influent and effluent results 52
Figure 4-28. Phase III Part 3 TSS influent and effluent results 53
Figure 4-29. Phase III Part 3 SSC influent and effluent results 53
Figure 4-30. Phase III Part 3 total phosphorus influent and effluent results 54
Figure 4-31. Phase III Part 3 COD influent and effluent results 54
Figure 4-32. Phase III Part 2 PSD summary 55
Figure 4-33. Phase III Part 3 PSD summary 56
Figure 4-34. Phase IV TSS influent and effluent cumulative loading results 58
Figure 4-35. Phase IV SSC influent and effluent cumulative loading results 58
Figure 4-36. Phase IV total phosphorus influent and effluent cumulative loading results 59
Figure 4-37. Phase IV COD influent and effluent cumulative loading results 59
Figure 4-38. Phase IV particle size distribution analysis 60
vn
-------
Figure 6-1. Modifications to Up-Flo® Filter module 69
Figure 6-2. Modifications to Up-Flo® Filter module showing improved support details 70
Figure 6-3. Phase I TSS influent and effluent results 74
Figure 6-4. Phase I SSC influent and effluent results 74
Figure 6-5. Phase I TP influent and effluent results 75
Figure 6-6. Phase IRP influent and effluent results 75
Figure 6-7. Phase I tank water level 76
Figure 6-8. Phase I influent and effluent PSD summary 77
Figure 6-9. Phase II TSS influent and effluent results 79
Figure 6-10. Phase II SSC influent and effluent results 79
Figure 6-11. Phase II TP influent and effluent results 80
Figure 6-12. Phase IIRP influent and effluent results 80
Figure 6-13. Phase II tank water level 80
Figure 6-14. Phase II influent and effluent particle size distribution summary 81
Figure 6-15. Depth of sedimentation in sump 82
Figure 6-16. Sump particle size distribution analysis results 82
Tables
Table 3-1. Particle Size Distribution 17
Table 3-2. Revised Synthetic Challenge Water Concentrations 23
Table 3-3. Synthetic Challenge Water Mix Concentrations 24
Table 4-1. Desired Feed Rates at "Normal" Settings (matching the concentrations in the original
challenge solution) 29
Table 4-2. Desired Feed Rates at "4X Concentration" Settings 30
Table 4-3. Synthetic Wastewater Analytical Data Comparison Test Plan Concentration Mean
Testing 31
Table 4-4. Phase I Analytical Data Summary 32
Table 4-5. Phase II Analytical Summary 36
Table 4-6. Phase III Influent and Effluent Flow Summary 47
Table 4-7. Phase III Part 2 Analytical Data 49
Table 4-8. Phase III Part 3 Analytical Data 50
Table 4-9. Phase IV Analytical Summary 57
Table 4-10. Characterization of Material Captured in Up-Flo™ Filter Sump 63
Table 5-1. Replicate Laboratory Sample RPD Summary 66
Table 5-2. Laboratory Control Sample Data Summary 67
Table 6-1. Modified Synthetic Challenge Water Concentrations 70
Table 6-2. Synthetic Challenge Water Analytical Data Comparison to Desired Feed
Concentration 72
Table 6-3. Phase I Analytical Data Summary 73
Table 6-4. Phase II Analytical Data Summary 78
Vlll
-------
Abbreviations and Acronyms
cfs
COD
dso
EPA
ETV
ft2
ft3
g
gal
gpm
hr
in.
L
LAS
Ib
NRMRL
mg/L
min
mL
um
NIST
NSF
OBC
O&M
P
PI
PSH
QA
QAPP
QC
RPD
SAP
ssc
STPP
TCLP
TO
TP
TSS
USGS
VO
WQPC
WSC
Cubic feet per second
Chemical oxygen demand
Diameter of 50th percentile particle
U.S. Environmental Protection Agency
Environmental Technology Verification
Square feet
Cubic feet
Gram
Gallon
Gallon per minute
Hour
Inch
Liter
Linear alkylbenzene sulfonate (represented by Dodecylbenzenesulfonic
acid)
Pound
National Risk Management Research Laboratory
Milligram per liter
Minute
Milliliter(s)
Micron
National Institute of Standards and Technology
NSF International
Oil-based constituents
Operations and maintenance
Phosphorus
Principal Investigator
Penn State Harrisburg
Quality assurance
Quality Assurance Project Plan
Quality control
Relative percent difference
Sampling and Analysis Plan
Suspended sediment concentration
Sodium tripolyphosphate
Toxicity Characteristic Leachate Procedure
Testing Organization (Penn State Harrisburg)
Total phosphorus
Total suspended solids
United States Geological Survey
Verification Organization (NSF)
Water Quality Protection Center
Water-soluble constituents
IX
-------
Acknowledgements
Penn State Harrisburg was responsible for all elements in the testing sequence, including test
setup, calibration and verification of instruments, data collection and analysis, data management,
data interpretation, and the preparation of this report.
Penn State Harrisburg
777 W. Harrisburg Pike
Middletown, Pennsylvania 17057
Contact Person: Shirley Clark
The vendor of the equipment is:
Hydro International
94 Hutchins Dr.
Portland, Maine 04102-1930
Contact Person: Kwabena Osei
The Testing Organization thanks J. Bradley Mikula for his planning of the testing apparatus,
James Elligson, Julia Hafera and David Spyker for their assistance in assembling the test
apparatus and Christopher Roenning, Kelly Franklin, Christine Siu and Brett Long for their
many hours testing the Up-Flo™ Filter and analyzing the resulting samples. The TO also thanks
the Environmental Engineering Program for its support and patience during the testing period as
we occupied a large portion of the wastewater laboratory.
-------
Chapter 1
Introduction
1.1 ETV Purpose and Program Operation
The U.S. Environmental Protection Agency (EPA) has created the Environmental Technology
Verification (ETV) Program to facilitate the deployment of innovative or improved
environmental technologies through performance verification and dissemination of information.
The ETV Program's goal is to further environmental protection by substantially accelerating the
acceptance and use of improved and more cost-effective technologies. ETV seeks to achieve this
goal by providing high-quality, peer-reviewed data on technology performance to those involved
in the design, distribution, permitting, purchase, and use of environmental technologies.
ETV works in partnership with recognized standards and testing organizations (TOs);
stakeholders groups that consist of buyers, vendor organizations, and permitters; and the full
participation of individual technology developers. The program evaluates the performance of
innovative technologies by developing test plans that are responsive to the needs of stakeholders,
conducting field or laboratory tests (as appropriate), collecting and analyzing data, and preparing
peer-reviewed reports. All evaluations are conducted in accordance with rigorous quality
assurance/quality control (QA/QC) protocols to ensure that data of known and adequate quality
are generated and that the results are defensible.
NSF International (NSF) operates the Water Quality Protection Center (WQPC) in cooperation
with EPA. The Source Water Protection Area of the WQPC evaluated the performance of the
Hydro International's Up-Flo™ Filter, which is an in-drain device designed to remove
hydrocarbons, organically bound metals, sediments, and other organic chemical compounds from
commercial or industrial runoff and wet weather flow. This document provides the verification
test results for the Hydro International Up-Flo™ Filter.
It is important to note that verification of the equipment does not mean that the equipment is
"certified" by NSF or "accepted" by EPA. Rather, it recognizes that the performance of the
equipment has been determined and verified by these organizations for those conditions tested by
the TO.
1.2 Testing Participants and Responsibilities
The ETV testing of the Up-Flo™ Filter was a cooperative effort between the following
participants: EPA, NSF, PSH, and Hydro International.
The following is a brief description of each ETV participant and their roles and responsibilities.
1.2.1 U.S. Environmental Protection Agency
The EPA Office of Research and Development, through the Urban Watershed Branch, Water
Supply and Water Resources Division, NRMRL, provides administrative, technical, and QA
-------
guidance and oversight on all ETV WQPC activities. This peer-reviewed document has been
reviewed by NSF and EPA and recommended for public release.
The key EPA contact for this program is:
Mr. Ray Frederick, Project Officer, ETV Source Water Protection Program
(732) 321-6627 e-mail: Frederick.Ray@epamail.epa.gov
USEPA, NRMRL
Urban Watershed Management Research Laboratory
2890 Woodbridge Ave. (MS-104)
Edison, NJ 08837-3679
1.2.2 NSF International - Verification Organization (VO)
NSF is EPA's verification partner organization for administering the WQPC. NSF is a not-for-
profit testing and certification organization dedicated to public health safety and the protection of
the environment. Founded in 1946 and located in Ann Arbor, Michigan, NSF has been
instrumental in the development of consensus standards for the protection of public health and
the environment. NSF also provides testing and certification services to ensure that products
bearing the NSF Name, Logo, or Mark meet those standards.
NSF's responsibilities as the VO include:
• Review and comment on the test plan;
• Review the quality systems of all parties involved with the TO and subsequently, qualify the
TO;
• Oversee the TO activities related to the technology evaluation and associated laboratory
testing;
• Carry out an on-site audit of test procedures;
• Oversee the development of a Verification Report and Verification Statement;
• Coordinate with EPA to approve the Verification Report and Verification Statement;
• Provide QA/QC review and support for the TO.
Key contacts at NSF for the test plan and program are:
Mr. Thomas Stevens, Program Manager Mr. Patrick Davison, Project Coordinator
(734) 769-5347 (734) 913-5719
e-mail: Stevenst@NSF.org davison@nsf.org
NSF International
789 Dixboro Road
Ann Arbor, Michigan 48105
-------
1.2.3 Testing Organization - Penn State Harrisburg
Penn State Harrisburg (PSH) acted as the TO for the verification testing. The PSH
Environmental Engineering Wastewater Laboratory had the space and large-scale equipment
(tanks, pumps, etc.) to perform the testing on the Up-Flo™ unit, and the PSH Stormwater
Management Research Group Laboratory has the equipment and experience to perform the
analytical work for this test plan.
The TO provided all needed logistical support, established a communications network, and
scheduled and coordinated activities of all participants. The TO was responsible for ensuring that
the testing location and feed water conditions were such that the verification testing could meet
its stated objectives. The TO prepared the test plan; oversaw the testing; managed, evaluated,
interpreted, and reported on the data generated by the testing; and reported on the performance of
the technology.
TO employees manufactured and prepared the testing rig, assured the required test conditions
were met, and measured and recorded data during the testing. The TO's Project Manager
provided oversight of the daily tests.
The key personnel and contacts for the TO are:
Shirley E. Clark, Ph.D., P.E.
Assistant Professor of Environmental Engineering
Penn State Harrisburg Environmental Engineering Program
777 W. Harrisburg Pike TL-105
Middletown, PA 17057
1.2.4 Vendor — Hydro International
Hydro International is the vendor of the Up-Flo™ Filter. The vendor was responsible for
supplying a field-ready Up-Flo™ unit and filter media, and was available during all tests to
provide technical assistance as needed.
The primary contact for the vendor is:
Kwabena Osei, Research & Development Manager
(207) 756-6200
e-mail: kosei@hil-tech.com
Hydro International
94 Hutchins Dr.
Portland, Maine 04102-1930
-------
1.3 Verification Testing Site
The verification testing was performed at PSH's campus in Middletown, Pennsylvania. The
testing rig was set up in the PSH Environmental Engineering Wastewater High-Bay Laboratory,
which is capable of performing a wide array of research programs. The laboratory was equipped
with the necessary storage tanks and equipment to provide flows up to 50 gpm with storage of
1,700 gal in the clean-water tank.
Samples of the synthetic wastewater mixture used for testing were created and analyzed in the
Environmental Engineering Program's Stormwater Research Laboratory, which is located in the
same building as the Wastewater High-Bay Laboratory.
-------
Chapter 2
Up-Flo™ Filter Equipment Description and Operating Processes
2.1 Equipment Description
The Up-Flo™ Filter is a passive, modular filtration system that incorporates multiple elements of
a treatment train into a single, small-footprint device. The Up-Flo™ Filter uses a sedimentation
sump and screening system to pretreat runoff before it flows up through the filter media where
final polishing occurs. A high-capacity, siphonic bypass safeguards against upstream ponding
during high-flow events. The siphon also serves as a floatables baffle to prevent the escape of
floatable trash and debris from the Up-Flo™ Filter chamber.
2.1.1 Up-Flo™ Filter Components
The Up-Flo™ Filter has no moving parts and no external power requirements. It consists of a
cylindrical concrete vessel with plastic internal components and a stainless steel support frame.
The concrete vessel is a standard cylindrical manhole with an inlet pipe or a grate opening. An
inspection port at ground level provides access to the sump for sediment removal. The internal
components consist of angled stainless steel screens, wedge-shaped filter modules, a bypass
siphon with a floatables baffle, and an outlet module. The base of the outlet module is equipped
with a drain-down port design that enables standing water to drain out of the filter media
between storm events, preventing the re-release of captured pollutants. The Up-Flo™ Filter
components are shown in Figure 2-1.
Bypass Siphon/
Floatables Baffle
Inlet Grate/
Maintenance Access
Outlet Module
Filter Module
Angled Screen
•Drain-down port
(not seen from this
view)
Figure 2-1. Up-Flo™ Filter components.
-------
The filter module houses the media pack. The media pack consists of two filter media bags and
two layers of flow distributing media. The internal components of the filter module are shown in
Figure 2-2.
Filter Module Lid
P - L. LXf^^^i "Media Restraint
Conveyance channel-
^^^ Media Pack
Filter Module
Support Brackets •— Angled Screen
Figure 2-2. Filter module components.
2.2 Hydraulic Flow Path
The Up-Flo™ Filter is self-activating and operates on simple fluid hydraulics. The configuration
of the internal components directs the flow in a pre-determined path through the vessel as
described below.
2.3 Flow Conditions
2.3.1 Operating Flow Conditions
Challenge water enters the chamber from an inlet pipe or an overhead grate and flows into the
sump region where gross debris and coarse grit are removed by settling. Runoff continues to fill
the chamber until there is enough driving head to initiate flow through the filter media. At this
point, the water flows up through the angled screen into the filter module. In the filter module,
water passes up through the filter media and is conveyed to the outlet module via the flow
conveyance channel. The flow path through the Up-Flo™ Filter during normal operating
conditions is illustrated in Figure 2-3.
-------
Figure 2-3. Flow path during normal operating conditions.
2.3.2 Bypass Flow Conditions
Flows in excess of the filtration capacity are discharged directly to the outlet module by the
siphonic bypass. The siphon also serves as a floatables baffle to prevent the escape of buoyant
litter and debris. The flow path through the Up-Flo™ Filter during bypass flow conditions is
shown below in Figure 2-4.
Figure 2-4. Flow path during bypass flow conditions.
-------
2.3.3 Drain Down
Filter media continuously submerged in water can become anoxic, producing an environment
that promotes bacterial growth and the release of other harmful leachates. The Up-Flo™ Filter is
equipped with a drain-down mechanism to ensure that the filter media sits above the standing
water level during no-flow conditions. As flows subside, water slowly drains out of the chamber
through drain-down ports located at the base of the outlet module. The drain-down ports are
covered with filter fabric to provide treatment to the drain-down flows. The flow path for the
drain down mechanism is shown in Figure 2-5.
Figure 2-5. Flow path during drain down conditions.
2.4 Sizing and Hydraulic Capacity
The Up-Flo™ Filter is sized to treat the peak treatment flow of a water quality design storm. The
peak flow is determined from calculations based on the contributing watershed hydrology and
from a design storm magnitude set by the local challenge water management agency. The
number of filter modules included in an Up-Flo™ Filter is determined by the peak treatment
flow.
The flow rate through each filter module depends on the nature and type of media within the
module and the water level in the Up-Flo™ Filter chamber. By adjusting media blends and the
height of the water column in the chamber, each filter module can be engineered to have a
treatment flow rate of 10 to 25 gpm. The flow rate through each filter module will determine the
number of modules needed to treat the peak treatment flow of the storm event.
The Up-Flo™ Filter is equipped with a bypass siphon designed to discharge flows in excess of
the treatment flow. When influent flows exceed the filtration capacity, the water level in the
-------
Up-Flo™ Filter chamber rises until it reaches the height of the internal weir of the bypass. Once
water starts to flow over the weir, the bypass siphon begins drawing water out of the chamber
discharging the excess flows through the outlet module to the outlet pipe.
The height of the bypass can be adjusted to accommodate shallow retrofits or restrictive
hydraulic profiles. The standard Up-Flo™ Filter bypasses up to 7 cfs with 2.5 feet of hydraulic
drop. A shallow unit, depicted in Figure 2-6, has a bypass capacity of 4 cfs with 1.7 feet of
hydraulic drop.
Figure 2-6. Bypass water levels for standard Up-Flo™ Filter (left) and shallow Up-Flo™
Filter.
2.5 Test Unit Specifications and Test Setup Description
The unit to be tested is a full scale, commercially available catch basin system. For the standard
catch basin configuration, the Up-Flo™ Filter is comprised of one to six filter modules. In
normal business practice, the number of filter modules included in an Up-Flo™ Filter is
dependent upon the required peak treatment flow rate. Because the Up-Flo™ Filter is sized on a
per-module basis, it is important for modular systems to be characterized on a per-module basis.
TSS, phosphorous and hydraulic capacity performance claims will be verified on a one-module
Up-Flo™ Filter setup. The two-module Catch Basin Up-Flo™ Filter set up is shown in Figure 2-
7, Figure 2-8, and Figure 2-9.
-------
Up-Flo™ Filter Test Tank
Viewing/Clean Out Port •
Sump
_-...
Bypass Siphon/
Floatables Baffle
Outlet Module
Outlet Pipe
Filter Module
Support Brackets
with Legs
Figure 2-7. Up-Flo™ Filter test unit isometric view.
The test unit has a 24-in. sump depth, a 12-in. outlet, and an 18-in. acrylic viewing port. The
height of the bypass is set so that there can be 21 in. of driving head acting on the Up-Flo™
Filter before bypass levels are reached. The test tank will be set up such that inflows pour into
the chamber through the open top, replicating a grated-inlet field installation.
Up-Flo™
Filter test tank
Viewing/Clean
Out port
Sump
Filter Module
Bypass Siphon/
Floatables Baffle
Outlet Pipe
Support
Bracket Legs
Figure 2-8. Up-Flo™ Filter test unit plan view.
10
-------
Up-Flo™ Filter Test
Tank
Filter Module
Viewing/Clean Out Port
Sump
Bypass Siphon/
Floatables Baffle
Outlet Module
Outlet Pipe
Support/Legs
Angled Screen
Figure 2-9. Section view of Up-Flo™ Filter test unit.
The tests will be performed at PSH's Environmental Engineering Program's Wastewater High-
Bay Laboratory. The PSH laboratory is set up to handle testing of this type with physical
facilities that includes a water supply up to 50 gpm, tanks, mixers, and pumps to store and feed
the synthetic water, and all other associated piping, controls and related equipment. The
Up Flo™ Filter is a passive unit that does not require any utility connections to operate.
Therefore, there will be no electrical requirements needed for operation of the unit. The
laboratory is equipped with water and electrical needs to supply the synthetic challenge water to
the unit, operate pumps, mixers, and sampling equipment, etc. However, none of these
requirements would be needed in a field application.
The synthetic challenge water described later in this test plan contained simulated challenge
water solids and a source of particulate phosphorus. The contaminant concentrations in the
synthetic water were similar to those found in challenge water runoff, based on data generated
both during the Nationwide Urban Runoff Program (NURP) and the more-recent analysis of
outfall data. The solids that accumulated as part of testing were solid waste that required disposal
after testing. The solids were tested prior to disposal to ensure they are not regulated materials
that require special disposal. In addition, a carbon filter was used to treat the discharge water
after the effluent settling tank to ensure that the organic pollutants were removed to acceptable
levels prior to discharge.
11
-------
2.6 Up-Flo™ Filter Capabilities and Claims
2.6.1 System Capability
The Up-Flo™ Filter is a compact treatment-train device that targets the wide range of
contaminants typically found in water runoff. Each Up-Flo™ Filter includes a sedimentation
sump, coarse screens and polishing filter media. Coarse grit and gross debris is removed by
settling in the sump, neutrally buoyant debris is removed by screening, and fine suspended
sediment is removed by filtration. The filter media may be customized to target other site-
specific pollutants such as metals and organics.
A single filter module was used in this verification program. The filter media installed in the
module was Hydro International's CPZ Mix™, which is made up of activated carbon,
manganese-coated zeolite and peat. Granular activated carbon is a traditional filter media for
targeting organic chemicals, pesticides and herbicides. The manganese-coated zeolite targets
TSS, iron, manganese and ammonium in challenge water runoff. The small fraction of peat is
highly efficient at removing organics and metals.
Each filter module filled with the CPZ Mix™ has a design flow rate of 20-25 gpm when the
water level in the chamber provides 20 in. of driving head. Water is filtered through the filter
media until the water level in the chamber falls to zero inches of driving head. When the inflows
exceed the filtration capacity, the excess flows discharge through the bypass siphon directly to
the outlet module. The bypass is designed to accommodate 7 cfs of excess flows. This high-
capacity bypass siphon ensures that head-loss and flow-restrictions due to the filter media will
not cause collection system backups and ponding on the surface during events with high flow
rates.
Maintenance of the sump and replacement of the filter bags is important for the successful long-
term operation of the Up-Flo™ Filter. The flow capacity of the Up-Flo™ Filter will decrease as
it accumulates sediments. The filter bags should be replaced once a year (or as needed) to ensure
that fine sediment build-up is not allowed to accumulate such that the flow rate of the filter will
be significantly reduced. Sediment and gross debris must also be periodically removed from the
sump to ensure that accumulated sediment does not block the intake of the filter module.
This test plan was designed to meet the basic protocol requirements and focused on the treatment
capability of the unit to remove sediment and particulate phosphorous from synthetic challenge
water. The experimental design and sampling and analysis plan presented in the following
sections provide details on the test protocol and the constituents targeted for this verification.
2.6.2 Vendor Claims
The Up-Flo™ Filter is designed to incorporate multiple elements of a treatment train into a
single, small-footprint device. The Up-Flo™ Filter utilizes settling, screening and filtration to
remove gross debris and suspended sediment from challenge water runoff. Specifically, the
Up-Flo™ Filter will remove over 80% of fine total suspended solids (TSS) from challenge water
runoff, and it will also remove a portion of metals, organics and other pollutants commonly
12
-------
found sorbed to the surface of suspended sediment particles. Verification of the removals of
metals, organics (except that measured as chemical oxygen demand [COD]), and other pollutants
was not included as part of the test plan.
Regular maintenance events are necessary to ensure optimal performance of the Up-Flo™ Filter
over time. In-field maintenance includes removing floatables, sediment and other pollutants from
the sump and changing out the media packs. In-field inspection should occur regularly. In-field
media pack replacement should occur once a year or as needed. The in-field maintenance of each
Catch Basin Up-Flo™ Filter unit should take a half-hour or less. Maintenance on the Up-Flo™
Filter test unit will occur after each phase of performance testing. The side of the Up-Flo™ Filter
test tank is equipped with an 18-in. access port to facilitate sump cleanout (see Figure 3-7). To
replace media packs, entry into the test tank is necessary. The tank is spacious enough to provide
comfortable access for one maintenance person. Confined space issues did not need to be
addressed during this testing since the test tank was open to the atmosphere.
To properly maintain the Up-Flo™ Filter, the steps detailed in the Up-Flo™ Filter Operation &
Maintenance (O&M) Manual were followed.
2.7 Performance Measures for the Verification Test
The performance capabilities of the Up-Flo™ Filter were assessed both quantitatively and
qualitatively. Sampling and analysis of the influent, effluent, and residues provided data to
determine the treatment efficiency of the unit with quantitative data. Recording of visual
observations, operational issues and maintenance requirements provided a basis for qualitatively
assessing the unit's performance. The test plan, including the Experimental Design, Sampling
and Analysis Plan (SAP), and Quality Assurance Project Plan (QAPP), focused on obtaining
performance-based data that served as the foundation of the verification report and the
verification statement.
2.7.1 Contaminant Selection and Monitoring for Performance
The Up-Flo™ Filter unit is designed to remove solids and solids-associated pollutants, such as
particulate-bound phosphorus in runoff. Based on the unit's capabilities a list of targeted
contaminants that will be monitored for removal by the unit has been selected. The targeted list
is as follows:
Targeted Contaminant List
• Suspended sediment concentration (SSC)
• Total suspended solids (TSS)
• Total phosphorus (TP)
These constituents, in addition COD [as a surrogate for the added organics], were measured in
influent and effluent samples in accordance with the experimental design and the SAP. The
results provided data for determining the performance capability of the unit to remove targeted
contaminants and provide data on the additional and secondary contaminants as well. All of
13
-------
these data are reported in the verification report as part of the quantitative performance
measurements.
2.7.2 System Component Operation and Maintenance Performance
The overall system performance was measured both quantitatively and qualitatively. Quantitative
measurements included determination of the range of hydraulic flow conditions that can be
handled by the unit. The hydraulic capacity of the unit was determined by measuring the
hydraulic flow rate in volume of water treated and flow rate handled. The experimental design
included both hydraulic loading tests and loading of contaminants to the unit. The filter media
and containment bag combination was stressed to exhaustion and spike loads were charged to the
unit at high flow rates. The mass removal of contaminants was determined.
Qualitative measures were assessed by observations of and experience with the unit during the
setup and testing phases. Records were maintained on the ease and time of installation, the time
and ease of maintenance for cleanout and absorption medium replacement, and other operating
observations. The unit is a simple design with no controls, instrumentation, alarms, or other
mechanical or electrical devices that will require operation. The unit was monitored for solids or
debris buildup, clogging of entry paths, and other related operational issues. The O&M Manual
provided by Hydro International was reviewed for its specificity and completeness. These
observations, experiences, records and review will be the basis for evaluating the system
performance in terms of operation and maintenance.
2.7.3 Quantification of Residuals
Testing the Up-Flo™ Filter created residual material, such as removed contaminants, sediments,
and spent filter media. The quantity of residual materials requiring disposal was a factor in
performance measurements.
14
-------
Chapter 3
Verification Testing Procedures
3.1 Testing Objectives
The objective of in-drain treatment system verification testing under the ETV Source Water
Protection Protocol for In-Drain Treatment Technologies is to evaluate the contaminant removal
performance and operational and maintenance performance of commercially available systems.
The objective of this testing was to determine the performance attained by the Hydro
International's Up-Flo™ Filter when used to treat synthetic challenge water containing a variety
of contaminants, including sediments, hydrocarbons, water-soluble organics and fertilizer. In
order to estimate the "life" of the device before maintenance in a rapid period of time, the
concentrations of all contaminants except sediment tested were higher than those typically seen
in urban stormwater, but lower than those anticipated to be seen in mixes of stormwater and
washwater.
The objective was achieved by implementing testing procedures presented in the protocol and
test plan (Appendix A). A synthesized challenge water containing sediments, petroleum
hydrocarbons, and surfactants was prepared to simulate contaminants at concentrations typically
found in a mixture of surface water runoff and other wet-weather flows at a commercial or
industrial setting. The treatment system was challenged under a variety of hydraulic loading
conditions using the synthetic wastewater. Influent and effluent samples collected from the unit
were measured for various contaminants as determined by indicator tests (e.g., COD, TSS, SSC,
Particle Size Distribution, and TP). The results were used to calculate removal efficiencies and
system capacities, and to determine the system treatment effectiveness. The treatment system
was also monitored for operation and maintenance characteristics, including the performance and
reliability of the equipment and the level of operator maintenance required.
The experimental design followed the methods and procedures defined in the protocol. The
design incorporated all of the elements described in the protocol and included all of the phases of
testing prescribed. There were two anticipated deviations or exceptions from the protocol as
understood by the TO. These deviations were as follows:
1. The measurement of head loss was not directly applicable due to the design of the
Up-Flo™ Filter; and
2. The synthetic challenge water concentrations set to reflect the requested challenge water
concentrations, and, since no description of the sediment was provided, the particle size
distribution of the sediment was selected based on those required by New Jersey
Department of Environmental Protection challenge water device evaluation protocols.
The verification test was a controlled test. The testing was performed on a full-scale unit
(containing one filter cartridge) and was set up in the PSH Wastewater Research laboratory. The
PSH Wastewater Research laboratory is a physical testing laboratory with space, tanks, piping,
utilities, etc., to perform medium scale (10 - 50 gpm) testing of this type. The synthetic
challenge used for the testing was made as described later in this section and dosed to the unit as
15
-------
prescribed in the protocol, with the exceptions noted in this report due to the low concentrations
required.
3.2 Test Equipment
The Up-Flo™ Filter unit was placed in a specially designed (vendor-supplied) testing tank that
simulated a typical catch basin used in stormwater runoff conveyance systems. The testing rig
designed and constructed by PSH personnel controlled influent and effluent flow and constituent
feed rates. The rig also provided for collection of influent and effluent liquid samples for
laboratory analysis, and observation of performance conditions, such as bypass, in a simple and
effective manner.
Figure 3-1 shows the process flow diagram and equipment configuration for the test setup. City
water stored in a 1,700 gal holding tank served as the main water feed. Oil-based constituents
(OBC) (gasoline, diesel fuel, motor oil, and brake fluid) and water-soluble constituents (WSC)
(windshield washer fluid, antifreeze, and surfactants) were stored in two-liter bottles and fed by
variable-speed peristaltic pumps into the inlet pipe containing the water. The inlet pipe was a
12-in. PVC plastic pipe that received water from the feed tank and dispensed the water mixture
into the Up-Flo™ device. A dry feeder above the channel dispensed the solids mixture into the
water stream at controlled rates.
Synthesized
Contaminants
Two (2) Tanks
Soil Dry Feeder
Open Channel
<
L
Influent ^L
7 1
i T
Influent
Sampling Point
Clean
Water
Up-Flo™
Filter and
test catch
basin
( FE J Effluent
Effluent
X
Effluent Sampling
Point
Figure 3-1. Test rig process flow diagram.
16
-------
The test site was the PSH Environmental Engineering Wastewater Research laboratory in
Middletown, Pennsylvania. The physical laboratory was set up to handle medium-range flow
testing and full-scale unit testing. The facility had space to set up several large tanks and piping
to convey the challenge wastewater to the full-scale test unit. The laboratory setup designed for
this verification activity could supply up to 50 gpm of city water as a main feed during the
testing. Ample electrical service was available to run all pumps, controllers, samplers, and
associated equipment.
The Up-Flo™ Filter unit used for the verification test was that of a full scale commercially
available one-module catch basin configuration which would be used in catch basin applications.
Influent to the unit was pumped into the same elevation as the grate inlet (relative to the unit) so
that flow could move through the system by gravity and the driving head in a manner similar to a
field application. Effluent from the unit flowed by siphon out of the side of the test unit in the
same manner that the flow would exit the unit in the field.
Figure 3-1 shows the process flow diagram and equipment configuration for the test setup. City
water served as the main water feed with a maximum flow rate of up to 50 gpm. A flow control
valve controlled the flow. The flow rate of the water was measured using a standard "paddle
wheel" style flow meter that showed flow rate (gpm) and totalized the volume processed (gal).
Synthetic challenge water was made by adding pre-mixed and sized solids, with a specific
amount of a particulate phosphorus source (ground slow-release fertilizer) to the city water. The
solids were mixed in the appropriate ratio using a cement mixer and stored in a bucket near the
test device. Periodic samples were collected from the solids dry-feeder (hopper) of the device to
ensure that the mixture had not separated during storage. The original intention was to add the
solids by slurry, but initial tests in the lab found that the sand could not be kept in suspension
even in a stirred sample bottle. Therefore, the hopper was used to feed the solids into the pipe
with sufficient mixing area available in the pipe between the solids-addition point and the entry
to the device. The sieve size analysis of the selected solids mix, as specified by a New Jersey
testing plan, is provided in Table 3-1, and is displayed graphically in Figure 3-1.
Table 3-1. Particle Size Distribution
Description
Coarse sand
Medium sand
Fine sand
Very fine sand
Silt
Fine Silt
Clay
Particle Size
Gun)
500-1,000
250-500
100-250
50-100
8-50
2-8
1-2
Sandy loam
(% by mass)
5
5
30
15
25
15
5
This distribution can be approximated by mixing a pre-sieved concrete plant sand with the
Sil-Co-Sil 250 to be purchased from U.S. Silica, Inc. (or one of its distributors). The mix is
42.5% sand and 57.5% Sil-Co-Sil 250.
17
-------
100
100
Particle Size (jam)
1000
10000
Figure 3-2. Sediment particle size distribution graph.
The original intention was to use Miracle-Gro™ topsoil as the source of particulate phosphorus
to generate the required additional phosphorus (above that supplied by the STPP in the organic
mixture). Testing at PSH on this topsoil showed that the phosphorus content measured as TP is
approximately 0.04 mg TP/g Miracle-Gro™. However, the calculation of the solids mixture
requirement was that the topsoil did not have sufficient total phosphorus to be used (the sand
would have been eliminated from the mixture entirely). A search for a source of particulate
phosphorus resulted in the use of Scott's Starter Slow-Release Lawn Fertilizer. This fertilizer is
granular with a coating designed for slow release. The intact particles were too large to use in the
solids feed so the fertilizer was ground using a coffee grinder and the phosphorus concentration
measured. The testing showed the approximate TP concentration of the Scott's Starter to be
0.3 mg TP/g fertilizer.
All sampling was performed manually for all test sequences. This eliminated the concern
regarding the collection of representative solids when using automatic sampling equipment.
The synthetic challenge water entered the treatment unit through the open top of the device
grating, flowed through the sump/sediment collection section, and passed over/through the
adsorbent materials. The treated water exited through the outlet pipe along the side of the unit.
Flow rates were measured both at the beginning and outlet of the system. A sampling port was
located in the effluent pipe for collection of manual grab samples. All sampling was performed
manually for all test sequences.
18
-------
Attempts were made to use an automatic flow measuring device to confirm the readings from the
flow meter. However, the low flow levels in the effluent pipe that accompanied the unit made the
use of the meter impossible. In addition, the TO was concerned about the trapping of solids
around the flow meter installation. To compensate for the lack of automatic flow readings in the
effluent pipe, manual flow measurements (bucket and stopwatch) were made periodically
throughout the testing - typically with every sample collection or every two hours, whichever
came first. Head measurements in the tank were taken at every sample collection and flow
measurement time.
3.3 Test Phases - Hydraulic Loading
The unit was tested under varying hydraulic load conditions to simulate typical conditions found
in wash water applications (i.e., catch basins and drain inlets in streets, parking lots, etc., that
contain substantial dry-weather flows or truck-washing facilities) and during challenge water
flows. The primary operational characteristics tested included the following:
• Performance under intermittent flow conditions;
• Performance under different hydraulic loadings, including peak flow;
• Performance at different contaminant loadings; and
• Capacity of the unit to contain contaminants.
The testing was done in four phases that included conditions designed to test all of these
operating scenarios. The phases described below followed the same phases that are discussed in
the protocol.
3.3.1 Phase I — Performance under Intermittent Flow Conditions
In Phase I the system was operated intermittently to simulate actual in-drain treatment
applications during intermittent loadings at flow rates that are typical average flow rates over a
period of time. The Up-Flo™ Filter catch basin unit, with one to six filter modules, is designed
to treat flow rates of up to approximately 20 gpm per filter module before any water is bypassed
through the overflows. A more typical average flow rate at a catch basin or drain inlet is
expected to be in the 10-15 gpm range. A flow rate of approximately 11 gpm was used for the
Phase I four-to-five day test. The intermittent tests were run for a 40-hr period. During the ON
cycle, the unit received flow for 15 min, followed by a 15-min period with no flow. The result
was 16 flow periods during eight-hour ON cycle (two 15 min flow periods per hour for 8 hr).
The flow was constant during the dosing periods at a flow rate of approximately 11 gpm. Flow
rates were recorded throughout the ON-cycle period and the effluent flow rate was recorded
periodically during the OFF cycle to determine drain down flows.
Samples of both the influent and the effluent were collected by manual grabs. Samples were
only collected when flow is being sent to the unit. Samples for both the solids and phosphorus
analysis were collected manually with 500 mL of sample collected every 500 gal (approximately
once every 2 hr). Table 5-1 in the Sampling and Analysis Plan Section of the test plan provides a
summary of all sampling and analysis schedules for verification test.
19
-------
Observations included a physical description of the effluent water with respect to color, oil
sheen, etc. The unit was observed for any evidence of clogging, change in operating head or
head loss, flow patterns, or any evidence of bypass or short-circuiting. These observations are
described in Chapter 4. The protocol called for the measurement of head loss as part of the
monitoring of flow conditions in the unit. The Up-Flo™ Filter unit, however, is designed to
bypass any flow that does not pass through the absorption media. Given that the unit is fed by
gravity, is open at the top, and has an overflow capacity greater than the inlet, it is not possible to
measure head loss on the influent stream to the unit. An approximation of the depth of water
over the filter media in the treatment chamber was monitored by noting whether water was
bypassing the treatment media, and reported as an estimate of the head loss through the media.
This head loss, however, would only impact the capacity of the unit to treat water and would not
impact the concern regarding flooded conditions. Water depth measurements, therefore, were
recorded whenever samples and/or flow rates were measured.
The unit has a relatively large capacity (approximately 40 ft3) for holding sediment (settled
solids). The challenge water had a target sediment concentration of approximately 300 mg/L.
Assuming 100% removal from a flow of 12,000 gal and a 90 lb/ft3 bulk density, the retained
sediment would occupy 0.2 ft3. Therefore, sediment cleanout was not anticipated until the
Phase I test was completed. At the end of the Phase I period, the unit was inspected to determine
the condition of the sediment chamber and the absorbents. Observation of an increase in water
depth in the test tank during the test run would indicate whether the media was beginning to
blind or plug. If the media and sediment chamber were in good operating condition, the media
would be used for the capacity study. If the sediment chamber appeared to be filling quicker than
expected or the media was beginning to plug as indicated by water draining through the bypass
holes during the low-flow testing, the unit would be cleaned and the media pack will be replaced
as described in the O&M Manual.
3.3.2 Phase II - Determination of the Capacity of the Unit
The objective of the Phase II testing was to run the unit to "exhaustion" with respect to the
capacity of the absorbent material to remove suspended solids and/or phosphorus. During this
phase of testing, the unit was operated under continuous flow conditions for 12 hr/day until the
unit plugged with solids or the absorption capacity was exceeded. This was not a continuous test
sequence since it would be highly unusual for an in-drain unit to flow at near maximum flow
continuously until exhaustion occurred. Therefore, operating on a 12-hr basis was selected since
it would most resemble real-world conditions.
The test plan allowed for using the loadings from the Phase I test to contribute to the loadings in
Phase II. The total loading from Phase I would then be added to Phase II data to calculate total
capacity. The flow rate for this test was set at approximately 16 gpm, which is approximately
80% of the maximum rated flow capacity of 20 gpm.
If the unit capacity had not been exceeded in the first 12 hr run (about 11,500 gal of water), the
test plan called for the unit to be operated for a second 12 hr period or until the solids capacity
was reached. If after the second 12-hr period indicates exhaustion has not been achieved, then
20
-------
the unit would be started again and would continue to be dosed on a 12 hr run schedule until the
maximum absorption capacity was reached.
Samples were collected on a grab sample basis. Samples from the influent and effluent were
collected at the start of the test and after approximately each 10,000 gal of influent flow, and
analyzed for the primary constituents (TSS, SSC, TP). Samples were collected on the same
schedule until the capacity was achieved.
Flow rates were monitored throughout the test period on a minimum of a once per hour basis.
The water depth over the filter media was monitored and recorded. Increasing water elevation in
the test tank was an indication that plugging was occurring. At the end of the Phase II test, the
unit was cleaned and the media pack was replaced as described in the vendor's O&M Manual.
3.3.3 Phase III - Performance Under Varied Hydraulic and Concentration Conditions
This phase of testing focused on determining the unit's hydraulic capacity and how well it
handled spike loads of constituents. Phase III had three distinct parts.
3.3.3.1 Part 1: Hydraulic Capacity with Clean Water
The vendor stated that one filter module has a rated capacity of 20 gpm for treating water. Flows
above 20 gpm would be bypassed through the bypass openings in the top of the unit. In order to
confirm the rated treatment capacity the unit was challenged with increasing flow rates using
clean water in the Part I test.
The test started with a clean unit containing fresh media. Only the clean water line was used for
this test. The drain-down ports on the base of the outlet module were plugged prior to testing.
Flow started at 10 gpm of fresh water for a period of 15 min. After 15 min, the flow was
increased to 15 gpm for a period of 15 min. Flow continued to be increased by an additional
5 gpm (20 gpm, 25 gpm, 30 gpm, etc.) in 15-min increments until flow began through the
bypass. The maximum flow rate achieved, before bypass and after bypass occurs, was recorded.
Flow increases continued until the maximum available fresh water rate was reached. All flow
rates and operating observations were recorded.
Observations of the water elevation at various flows were made so that estimates of media head
loss can be made. The overflow was monitored and water height at various bypass flow rates was
recorded.
3.3.3.2 Part 2: Hydraulic Throughput with Synthetic Challenge Water
The Part 2 testing followed the same approach as the Part 1 testing except that the synthetic
challenge water was used as the influent water. In this part, the chemical feed pumps and hopper
were used to add the stock solutions to the fresh water. At each increase in flow rate, the pumps
and feeder were increased in rate in ratio to the fresh water feed to maintain a constant
concentration of constituents in the synthetic challenge water.
21
-------
The test was conducted after the Part 1 test and used the same filter media that was used for the
Part 1 test. Flow started at 10 gpm for a period of 15 min. After 15 min, the flow was increased
to 15 gpm for a period of 15 min. Flow continued to be increased by an additional 5 gpm
(20 gpm, 25 gpm, 30 gpm, etc.) in 15-min increments until flow began through the bypass holes.
The maximum flow rate achieved before bypass and after bypass begins was recorded in the
logbook. After achieving the maximum treated rate, the flow continued to be increased to
challenge the bypass system. All flow rates and operating observations were recorded in the
logbook along with any physical observations regarding the unit response during the test.
Grab samples of the influent and effluent were collected at each flow rate condition. All samples
will be analyzed for the complete list of constituents (solids and phosphorus).
3.3.3.3 Part 3: Impacts of Spike Concentration Loadings
Part 3 was a test series designed to evaluate the impact that spike loadings may have on the unit's
ability to remove key constituents. The key constituents for the Up-Flo™ Filter are TSS, SSC,
PSD, and TP. The hydraulic loadings were increased following the same protocol as for Part 2.
Using the same unit (no cleanout or media pack) as for Part 2, the test procedure started at a flow
rate of approximately 10 gpm. The chemical feed pump rates of the stock solutions and dry
feeder were set at a factor of four times higher than used in the previous tests. This increased the
concentration of constituents approximately by a factor of four. Grab samples of the influent and
effluent were collected at each flow rate condition until the unit flooded or the maximum
available feed water capacity was reached. All samples will be analyzed for all constituents of
interest. At the end of the Phase III tests, the unit was cleaned and the media pack replaced as
described in the vendor's O&M Manual.
3.3.4 Phase IV — Contaminant Capacities at High Hydraulic Throughput
The influence on treatment efficiency of high hydraulic loads on the unit were tested in
Phase IV. The Phase IV test was a capacity or "exhaustion test" similar to Phase II, except the
unit will be under higher hydraulic loads typical of a very large flow event. The Up-Flo™ Filter
unit was somewhat unique in that it treats all of the water that can pass through the treatment
chambers and then bypass the remaining water. Thus, at higher flows (above treatment capacity)
it will not backup and flood an area around the inlet, but rather will treat a set flow, about
20 gpm/ft2 of filter media, and the additional flow will be bypassed to the catch basin outlet and
enter the collection system. Under this high flow rate test, the unit was operated above the rated
treatment capacity with the bypass flowing and removing the extra flow. The flow rate was set at
approximately 32 gpm, which is above the treatment capacity. The test was designed to
demonstrate the system's treatment capability when it is operating in bypass mode. The test time
period was 12 hr continuous flow per 24-hr day. However, the unit did not end up in bypass. The
results are described later in the verification report.
Observation of the flow rates through the treatment unit and the bypass were to be used as the
primary indicator that solids capacity has been reached. When flow rates in the treatment section
decreased by 25% or more for 30 minutes, capacity was considered to have been reached.
22
-------
Samples were collected on a grab sample basis. Samples from the influent and effluent were
collected at the start of the test. It was anticipated that the flows would be sampled every
10,000 gal of water treated and analyzed for the primary constituents. However, given the nature
of the breakthrough pattern and timing seen in Phase II testing, it was determined that higher
resolution sampling was required. Samples were collected every 30 min for the first 2 hr of
testing and then once per hour after that. Flow rates were monitored throughout the test period on
a minimum of a once per hour basis.
3.4 Influent Characterization
3.4.1 Synthetic Challenge Water
The verification test will be performed using synthetic water (Table 3-2) made from a mixture of
solids - one of which will provide the particulate phosphorus required by the test plan. The
following products will be used to make the synthetic challenge water:
• Regular unleaded gasoline;
• Diesel fuel;
• 10W-30 motor oil;
• Brake fluid;
• Antifreeze (glycol based);
• Vehicle washing detergent (specific chemical addition - see below);
• Windshield washer fluid;
• Sil-Co-Sil 250;
• Slow release phosphorus-supplying fertilizer; and
• Concrete plant sand sieved to a size of all passing through 5,000 jim.
Table 3-2. Revised Synthetic Challenge Water Concentrations
Concentration
Parameter (mg/L)
SSC 300
TSS 300
Total phosphorous (as P) 5-10
COD 100-200
A formula using a mix of the above named products/materials has been made and tested in the
laboratory to determine the conformance to these specifications. The synthetic mix that was
prepared and tested is shown in Table 3-3. The results of testing the ground fertilizer for
phosphorus content is 0.3 mg TP/g Scott's Lawn Starter Fertilizer. The addition of fertilizer
replaced approximately 10% of the sand in the mixture.
23
-------
Table 3-3. Synthetic Challenge Water Mix Concentrations
Product or Material
Concentration in Water (mg/L)
Regular unleaded gasoline
Truck diesel fuel
10W-30 motor oil
Brake fluid
Antifreeze (glycol based)
Dodecylbenzenesulfonic acid (LAS)
Sodium tripolyphosphate (STPP)
Windshield washer fluid
Solids Mixture
0.08
3.9
19
0.97
10
10
2
10
300
The product concentrations in Table 3-3 represent a deviation in the constituent concentrations
identified in the original protocol. The hydrocarbon concentrations specified in the protocol
were not achievable in prior testing due to the insolubility of hydrocarbons with water. For this
test plan, the VO agreed that the hydrocarbon concentrations could be decreased further (to a
targeted concentration range of 10 to 20 mg/L) since the vendor makes no specific claims for
hydrocarbon treatment. Since the vendor did request an evaluation of particulate phosphorus
removal, a slow-release fertilizer was used to increase phosphorus concentrations to
approximately 5-10 mg/L (when combined with the STPP required by the VO). The VO, TO,
and vendor agreed that the materials that comprise the synthetic challenge water should provide a
condition suitable to adequately verify the performance of the Up-Flo™ Filter against the
protocol requirements. This mixture was designed to represent a mixture of stormwater runoff
and a dry-weather flow/washwater that contains a substantially higher load of detergents and un-
emulsified hydrocarbons than is typically seen in urban runoff. The use of this mixture at the
higher loadings shortened the testing time required for the Up-Flo™ compared to using a
simulated solids mixture and increased the blinding of the media by the OBC and WSC
constituents. The concerns raised by this mixture would be likely to be seen in applications with
heavy influences of detergents and/or locations with visible free-floating hydrocarbon products.
3.4.2 Stock Solutions
The standard mix determined above (Table 3-3) was used for all of the verification tests. The
Sil-Co-Sil, fertilizer, and sand was supplied by the hopper and set to meet the concentration
targets in the established mix. The solids were premixed prior to filling the hopper to
homogenize the solids feed. The hopper had to be refilled frequently to ensure that the solids did
not separate during the test. In addition, the humidity in the laboratory testing required regular
maintenance on the hopper to prevent solids "cementing" in the influent line.
The remaining products were mixed into two separate solutions. One solution included the
hydrocarbon-based products (gasoline, diesel fuel, motor oil, and brake fluid), while the other
solution included the water-soluble products (antifreeze, LAS, STPP, and windshield washer
fluid). The two solutions were prepared using the following specifications:
24
-------
• Hydrocarbon mixture (fed into the water at a rate of 0.03 mL per L of water):
o 10 grams (g) motor oil
o 2 g diesel fuel
o 0.05 g gasoline
o 0.5 g brake fluid
• Water-soluble mixture (fed into the water at a rate of 0.1 mL per L of water):
o 10 g windshield washer fluid
o 10 g antifreeze
o 10 g LAS
o 2 g STPP
o Mixture diluted to 100 mL with tap water
3.4.3 Influent Characterization during the Verification Testing
The influent synthetic challenge water was sampled and analyzed during all of the various test
conditions described in Phases I - IV. While the protocol allowed for single daily samples of the
influent in several test cases, the approach used in the test plan was to match influent and effluent
samples as often as possible for all sampling periods. This was to ensure that the actual influent
concentrations would be known for all test conditions.
Because of the large water volumes needed for these tests, it was not practical to make a single
large daily batch of synthetic water to supply the entire day's flow. Instead, the system used
more concentrated stock solutions that were injected into the fresh water flow in the open
channel section.
3.4.4 Solids Characterization during the Verification Testing
Influent and effluent solids were characterized using the Coulter Counter Particle Size Analyzer
for particles in the range of 0.6 |j,m up to 240 |j,m. Particles above 250 |j,m were characterized by
sieving the samples through a stainless steel sieve with a mesh size of 250 |j,m. The combination
of the Coulter Counter results and the sieve analysis for the large particles allowed for a
complete characterization of the influent and effluent particle distribution between 0.6 |j,m and
5,000 |j,m. The results for the solids analysis were subdivided into removal for the following
particle size ranges:
• 0.6 - 3 |j,m
• 3 - 12 urn
• 12-30 urn
• 30-60 urn
• 60 - 120
• 120 - 240
• > 240
25
-------
3.5 Effluent Characterization
The effluent quality was monitored during all phases of testing except during the fresh water
hydraulic test in Phase III, Part 1. The sampling and analysis approach focused frequent sample
collection and analysis on the key parameters for evaluating the UpFlo™ Filter unit as described
previously. Specific details on the sampling and analysis frequency and parameter list are
provided in the SAP section of the test plan and in the previous sections describing the test
phases.
3.6 Residue Management
Residues, including sediment in the settling chamber and the absorbent media, were removed
from the unit at the end of some Phases of testing as described in Section 4.3. This task included
recording the volume of residues/media collected and the wet weight of residues/media
collected. These data were used to provide information on typical cleanout volume and weights
that can be expected from normal operation.
Solid residues were collected from the sedimentation chamber in the unit after the majority of the
water in the unit had been removed using a sump pump. The sediment was removed using a
vacuum system (wet/dry shop vacuum) to simulate the typical removal system used in the field
(vacuum truck). The content of the shop vacuum reservoir was removed using scoops, spatulas,
scrapers, etc. to remove as much material as possible. These solids were measured for wet weight
and volume in order to evaluate the amount of solids that can be expected to be generated and
cleaned out of the unit on a volume throughput/loading basis. Samples of the solids were also
measured for solids content so that a dry weight of solids produced could also be calculated.
Three sub samples of the sediment were collected and percent solids measured. The weight of
solids collected was used to relate the accumulation rate of solids to total water treated.
One representative sample of the spent filter media and retained sediments was analyzed for
COD and TP and for leachate testing following the TCLP procedure. Attempts were made to
weigh the filters and obtain masses of residue gathered on the media. However, because of the
differences in weights due to moisture content between the new bags (which were not completely
dry) and the used bags, this measurement could not be taken accurately.
3.7 Operation and Maintenance Observations
The Up-Flo™ Filter unit was operated by PSH during the test period. The vendor-supplied O&M
Manual is presented in Appendix B. Hydro International will also provide consultation on
installation and operation of the unit.
Installation of the unit was straight forward as the unit arrived at the PSH lab pre-assembled.
Support brackets with legs sit on the base of the test tank. The filter modules were secured onto
the support brackets. The outlet module had a pipe stub that fits up to the tank outlet via
standard Fernco® coupling. The test tank had an open top.
26
-------
Laboratory personnel maintained a detailed logbook describing all observations made during the
tests. Any unit cleaning, clearing of debris, unclogging of the screens or media, etc. were
recorded. Observations were also recorded on the ease of installation, operation, and
maintenance. These observations included a qualitative assessment of the degree of difficulty
encountered during the cleaning of the unit at the end of each phase and on the ease of replacing
the media pack.
Flow rates, volume of water processed, amount of stock solutions pumped from the stock feed
tanks, and related operational data for each test run were recorded in the operational log. Any
deviations or changes from the prescribed test plan were thoroughly documented. The
measurements of residue volumes and weights were recorded after cleaning periods.
Any other observations on the operating condition of the unit or the test system as a whole were
recorded for future reference. Observations of changes in effluent quality based visual
observations, such as color change, oil sheen, obvious sediment load, etc., were recorded for use
during the verification report preparation.
27
-------
Chapter 4
Verification Testing Results and Discussion
4.1 Synthetic Wastewater Composition
The protocol and test plan set forth a requirement that the TO maintain constituent feed rates in
the synthetic wastewater of ±50 percent of the target feed during the course of testing so that the
system would be properly challenged. Prior to beginning the testing, the TO created calibration
curves for each pump (water, OBC, WSC) using the appropriate feed mixture. Then the flow
rates were set based on the calibration curves. The flow meter calibrations are shown in Figures
4-1, 4-2 and 4-3 for the feed water, WSC and OBC, respectively.
M -
2D •
Water Flow Meter Calibration
Mnlrr Rising {GPM} = 0 fiSlAcjiial Flow Bain) - 4 9H?
Ft' - 0JW
,-*
n 10 711 in 40
Actual Flow Rate I (3PM;
Figure 4-1. Calibration of the flow meter for the feed water.
i- 6.0
Hydrocarbon Peristallsc Pump CaliSMIion
y Rail i;inL''s; = o.i2:Puirp Srt-rg: * 0 OT
0.6 -
Pump
Figure 4-2. Calibration of the hydrocarbon feed peristaltic pump.
28
-------
Water-Soluble Constituents Peristaltic Pump Calibration
0,2 •
Pump Setting'
Figure 4-3. Calibration of the WSC peristaltic pump.
Based on the calibration equations, the desired flow readings for the water and the peristaltic
pump settings were selected for each flow rate-concentration combination. The following tables
for settings were established (Tables 4-1 and 4-2). These were then corresponded to settings on
the hopper and the peristaltic pumps.
Table 4-1. Desired Feed Rates at "Normal" Settings (matching the concentrations in the
original challenge solution)
Water Flow Rate
(gpm)
10
15
20
25
30
35
40
45
50
Solids Feed Rate
(mg/sec)
189
284
379
473
570
662
757
852
947
OBC Feed Rate
(mL/sec)
0.019
0.028
0.038
0.047
0.057
0.066
0.076
0.085
0.095
WSC Feed Rate
(mL/sec)
0.063
0.095
0.13
0.16
0.19
0.22
0.25
0.28
0.32
29
-------
Table 4-2. Desired Feed Rates at "4X Concentration" Settings.
Water Flow Rate Solids Feed Rate OBC Feed Rate WSC Feed Rate
(gpm) (mg/sec) (mL/sec) (mL/sec)
10
15
20
25
30
35
40
45
757
1,140
1,510
1,890
2,270
2,650
3,030
3,410
0.076
0.114
0.151
0.189
0.227
0.265
0.303
0.341
0.252
0.379
0.505
0.631
0.757
0.883
1.01
1.14
Generally, the OBC and WSC feed rate was higher than targeted because of the inherent
difficulties posed by the low flows required of the testing. Two different sizes of pump tubing
were tried for each measure to better accurately target the desired flow rate range. However, the
desired range, especially for the OBC mixture, fell between the two pump tube sizes. The WSC
settings were in range but all at the lowest end of the range. For both OBC and WSC, it was
decided to proceed with the larger tubing, based on the belief that over-challenging the filter was
better than under-challenging it. Part of the desire of the protocol is to evaluate the impact that
hydrocarbons and other constituents of washwater (similar to that found in service station and
maintenance yard drains) have on blinding of the filter media. Therefore, the use of the larger
tubing was warranted in order to not undercut the concentration of the two "fouling" agents.
Two general problems were encountered with the dosing of the solids. The humidity generated in
the laboratory due to the water flow created a clogging problem in the solids hopper and
removed the option of using the lowest motor settings. The calibration of the hopper therefore
was inconsistent and had to be maintained regularly. Physical measurements of the hopper solids
being dispensed into the water stream indicated that when the hopper was fully functional, the
dispensing was in the desired range. However, the partial clogging was an issue throughout the
tests. The second general problem was encountered in all sample collection procedures where the
solids mixture contains comparatively large particles and is the question of where to sample in
the influent flow stream. For this device, based on initial observations of the flow, the influent
was sampled in the stream as the stream "united" entering the device. However, the solids results
showed that, although they were not observed un-entrained in the system, the sand particles were
falling out of the inlet pipe and were not evenly dispersed in the influent. At the end of the
Phase III testing, the sampling location was moved up to the edge of the influent dispensing pipe
and near the center bottom of the flow stream. This also has been documented to cause errors of
measurement, with a potential bias toward higher solids measurements than actually occurring.
This is because the sample is collected along the center flow path which is deemed to be where
the larger particles flow. The testing rig, in agreement with prior testing performed under a
different verification protocol with a different device, is not equipped to provide adequate mixing
of large solids into the water column. The addition of baffles or mixers was considered but was
rejected because of the concerns of forming a solids settling location, ensuring that the solids did
not end up in the Up-Flo™ Filter.
30
-------
4.2 Synthetic Wastewater Laboratory Analytical Results
During testing, 60 influent samples were collected during the normal constituent feed conditions
(Phase I, Phase II, Phase III Part 2, Phase IV) and analyzed for the various constituents specified
in the test plan. Table 4-3 provides a comparison of the mean analytical results for these influent
samples versus the analytical results for the synthetic wastewater mix specified in the test plan.
Table 4-3. Synthetic Wastewater Analytical Data Comparison Test Plan Concentration
Mean Testing
Measured Mean Desired Feed
Constituent Concentration (mg/L) Concentration (mg/L)
TSS 132 300
SSC 130 300
TP 44 5 -10
COD 121 70-100
The mean synthetic wastewater data for the primary constituents were measured to be greater
than desired for TP, and less than desired for TSS and SSC. They were within the ±50% of the
desired concentration for COD. This supports the observations that the test mixture was difficult
to dose correctly in a humid environment and when very little of the OBC and WSC were
required. The decision was made by the TO to supply a concentration of the OBC and WSC that
could be regulated correctly by the pumps. This meant that the flow rate was set for
approximately 20 drops per minute at the lower flow rates of testing - not a steady stream, but
sufficient to provide a measurable concentration. A review of the data shows that the COVs for
all parameters ranged between 0.5 and 1.0. In addition, as discussed earlier in the report,
collecting of samples in the "correct location" in the influent stream caused difficulties. A review
of the data by phases showed that the influent concentration of TSS and SSC for Phase II (the
last phase run) was over 230 mg/L and was within the ±50 percent guideline of the test plan.
The hopper dosage measurements were within the guideline for the test plan. Therefore, although
the mean analytical TSS and SSC concentrations were lower than the 300 mg/L concentration
specified in the test plan, the hopper dose measurements suggest that the theoretical test plan
concentration was close to the 300 mg/L goal. This also suggests that the analytical samples
were biased toward underreporting the actual solids concentration. On the basis of the hopper
dosage measurements, the overall objective for sediment loading was met.
The variances between the test plan and mean testing concentrations for the secondary
parameters exceeded the ±50% guideline for most parameters, but the vendor makes no claims
for the secondary parameters. Therefore, the variation from the targeted concentrations is
deemed to have no impact on meeting the testing objectives. However, the potential effects of
the increased secondary parameter concentrations on the sediment removal performance is not
known.
-------
4.3 Test Phases in the Test Plan
This section summarizes the analytical and flow data for the test phases specified in the test plan
(Phases I through IV). The efficiency values reported in this section are a function of the total
influent and total effluent concentrations and do not take into account the effects of water
bypassing the filter media.
4.3.1 Phase I - Performance under Intermittent Flow Conditions
As described in Section 3.5.1, the Phase I test took place over 40 hr, 8 hr per test day on days 1
and 4 and 12 hr per test day on days 2 and 3, with the flow alternating on and off for 15-min time
periods. The influent flow rate was set at 12 gpm throughout the test.
4.3.1.1 Analytical Data
The effect of blinding or clogging of filter media should be evident in the results comparing flow
rate through the media to the effluent concentrations. The TSS, SSC, TP and COD analytical
data as related to cumulative volumetric loading on the media are summarized in Table 4-4. The
test plan required that a minimum of one set of samples be collected each test day. The
verification organization collected a total of 26 sets of samples. The increase was to verify
whether filter media breakthough was occurring. Three sets of these samples encountered very
low or non-detectable sediment concentrations due to the issues outlined in Section 4.1. For the
purposes of verification, these three sample sets were considered testing anomalies and were
removed from statistical evaluations. These data are reported in the data set enclosed in
Appendix C.
Removal efficiencies for TSS and SSC ranged from 73% to 77%, depending on the analyte and
the statistical evaluation, which is slightly below the vendor's 80% performance claim. The VO
observed dark particles in the effluent at the beginning of the test phase. These dark particles
were likely the result of washing of fine sediments in the media bags. Large negative removal
efficiencies were observed, primarily at the beginning of the test phase, which is likely the result
of media bags washing out fine particles or bridging in the test rig's sediment dispenser, resulting
in low or non-detectable influent sediment concentrations. The TP data showed a mean and
median removal efficiency of 13%. The COD analytical data showed a mean and median
removal efficiency of 62% and 53%, respectively.
Table 4-4. Phase I Analytical Data Summary
Influent Concentration Effluent Concentration
(mg/L) (mg/L) Removal Efficiency (%)*
Analyte Mean Median Max Min Mean Median Max Min Mean Median Max Min
TSS
SSC
TP
COD
123
132
40
168
107
125
40
139
435
480
126
523
5
5
0.6
77
32
34
35
63
29
29
35
65
83
106
64
89
9
5
0.6
33
74
74
13
62
73
77
13
53
92 -1280
99 -480
91 -533
88 5.1
32
-------
1. Mean and median removal efficiency is a function of mean and median influent and effluent concentrations, and
maximum and minimum removal efficiencies are a function of individual paired data points.
A graphical examination of the data was also conducted to illustrate the results discussed above.
Figures 4-1, 4-2, 4-3, and 4-4 compare the influent and effluent concentrations for TSS, SSC,
TP, and COD, respectively.
Total Suspended Solids - Phase I
500
400 -
300 -
CO
CO
200 -
100 -
—•— Influent
—-O— Effluent
0 2000 4000 6000 8000 10000 12000
Cumulative Volumetric Loading (gal)
Figure 4-1. Phase I TSS influent and effluent results.
14000
The Up-Flo™ Filter did not exhibit signs of clogging or blinding during the test run. A review
of the water depth measurements at each sample time showed that the tank water level remained
consistent between 40 and 42 in. No buildup of head was noted in the unit, further indicating that
the media capacity had not been exhausted in the Phase I testing.
Particle size distribution analysis was also performed on representative influent samples and on
all the effluent samples. Since Phase I was not the chronologically first phase performed, many
influent samples had been analyzed prior to this and it was determined that the influent
distribution was relatively consistent. Figure 4-5 shows the results of the PSD analysis for
Phase I.
Figure 4-5 shows that the influent sample had the largest d50, indicating a reduction in the media
particle size in the solution as it passed through the Up-Flo™ Filter. This confirms the
predictions of the manufacturer that the Up-Flo™ Filter would be capable of removing the larger
particles in the solution. The data show the mean influent dso was 57 um and the mean effluent
d50 was 24 um.
33
-------
500
Suspended Sediment Concentration - Phase I
400 -
300 -
E,
o
200
100 -
—•— Influent
— O— Effluent
0 2000 4000 6000 8000 10000 12000 14000
Cumulative Volumetric Loading (gal)
Figure 4-2. Phase I SSC influent and effluent results.
Total Phosphorus - Phase I
140
120 -
100 -
en
E
80 -
60 -
40 -
20 -
Influent
—-O— Effluent
0 2000 4000 6000 8000 10000 12000
Cumulative Volumetric Loading (gal)
Figure 4-3. Phase I total phosphorus influent and effluent results.
14000
34
-------
600
Chemical Oxygen Demand - Phase I
500 -
400 -
O)
Q
O
O
300 -
200 -
100 -
—•— Influent
—-O— Effluent
_O.-O--OO
0 2000 4000 6000 8000 10000 12000 14000
Cumulative Volumetric Loading (gal)
Figure 4-4. Phase I COD influent and effluent results.
Phase I Influent & Effluent
100
—•—— Influent
—-o— Effluent
Dry solids mix
80 -
10 100
Diameter (Mm)
Figure 4-5. Phase I particle size distribution summary.
1000
35
-------
4.3.2 Phase II - Determination of the Capacity of the Unit
As described in Section 3.5.2, in Phase II the system was run to "exhaustion" with respect to the
capacity of the filter media to remove suspended solids or petroleum hydrocarbons. The unit was
operated under continuous flow conditions at a constant flow rate of 15 gpm until the unit
plugged with solids or the contaminant absorption capacity was exceeded. The test plan specified
a flow rate of 16 gpm for this test, based on the vendor's claims that the system could treat water
at a maximum flow rate of approximately 20 gpm.
4.3.2.1 Analytical Data
As specified in the test plan, samples were collected approximately every 10,000 gal and the
system was run until breakthrough was noted. Breakthrough was noted by the failure of the
media bags to remain in place in the system. The data are summarized in Table 4-5 and are
expressed graphically in Figures 4-6 through 4-9.
Table 4-5. Phase II Analytical Summary
Influent Concentration
Results (mg/L)
Mean Median Max. Min.
Effluent Concentration
Results (mg/L)
Mean Median Max. Min.
Removal Efficiency ("/o)1
Mean Median Max. Min.
TSS
ssc
TP
COD
215
237
89
82
164
171
75
67
492
555
183
134
84
96
47
60
60
71
56
60
50
64
50
62
100
108
81
80
30
<5
30
43
72
70
36
27
70
63
33
7.5
82
81
59
53
40
47
7.1
-3.3
1. Mean and median removal efficiency is a function of mean and median influent and effluent concentrations, and
maximum and minimum removal efficiencies are a function of individual paired data points.
36
-------
Phase II - Total Suspended Solids
600
500
400 -
300 -
CO
05
200 -
100 -
—•— Influent
--O- Effluent
0 10000 20000 30000 40000 50000
Cumulative Volumetric Loading in Phase II (gal)
Figure 4-6. Phase II TSS influent and effluent results.
Phase II - Suspended Sediment Concentration
60000
600
500
400 -
-*— Influent
• -O- Effluent
300 -
o
GO
oj
200 J
100 -
10000 20000 30000 40000 50000
Cumulative Volumetric Loading in Phase II (gal)
60000
Figure 4-7. Phase II SSC influent and effluent results.
37
-------
Phase II - Total Phosphorus
200
180 -
160 -
140
O. 100
I-
80 H
60
40 -\
20
— -o—
Influent
Effluent
0 10000 20000 30000 40000 50000
Cumulative Volumetric Loading in Phase II (gal)
Figure 4-8. Phase II total phosphorus influent and effluent results.
Phase II - Chemical Oxygen Demand
60000
140
120 -
100 -
Q
O
O
60
40 -
20
—•— Influent
- O— Effluent
0 10000 20000 30000 40000 50000
Cumulative Volumetric Loading in Phase II (gal)
Figure 4-9. Phase II COD influent and effluent results.
60000
38
-------
In general, the Up-Flo™ Filter was 40% to 82% effective in removing TSS and 47% to 81%
efficient in removing SSC from the influent. The TSS and SSC removal efficiencies actually
increased over the life of the test. The TP removal efficiencies ranged from 7% to 60% and the
COD removals ranged from <0% to 53%, with COD removal efficiencies increasing across the
test.
PSD analysis was also performed on the Phase II samples, as shown in Figure 4-10. The mean
d50 for the influent was 60 um and the mean effluent d50 was 26 um.
100
Phase II Influent & Effluent
Influent
Effluent
Dry solids mi>:
10 100
Diameter
1000
Figure 4-10. Phase II PSD summary.
The Phase II test was stopped when the TO noticed that the Up-Flo™ filter bags had been moved
out of place and that substantial solids were appearing the effluent samples compared to previous
samples. The mesh retaining the filter bags below the cartridge lid was displaced, allowing water
to bypass the filter bags. This breakthrough was noted prior to the water level reaching the
designed bypass level.
Because the media bags were not changed between Phases I and II, a full evaluation of the
Up-Flo™ Filter requires an evaluation of performance across the entire testing sequence on these
bags. Figures 4-11 through 4-14 summarize the media bag behavior across the entire range of
testing for TSS, SSC, TP and COD, respectively. Figure 4-15 summarizes the PSD analysis for
Phase I and Phase II combined. The mean dso for the influent was 59 um and the mean effluent
dso was 24 um.
39
-------
600
500 -
400 -
Phase l+ll - Total Suspended Solids
O)
E,
to
CO
300 -
Influent
— O— Effluent
200 -
100 -
10000 20000 30000 40000 50000 60000 70000
Cumulative Volumetric Loading (gal)
Figure 4-11. Phase I and II TSS cumulative loading results.
Phase l+ll - Suspended Sediment Concentration
600
500 -
400 -
O)
O
CO
co
300 -
200 -
100 -
0
—•— Influent
—-O— Effluent
Q = 15gpm
.-O
0 10000 20000 30000 40000 50000 60000 70000
Cumulative Volumetric Loading (gal)
Figure 4-12. Phase I and II SSC cumulative loading results.
40
-------
200
Phase l+ll -Total Phosphorus
Q = 10gpm
150 -
0.
—•— Influent
—-O— Effluent
Q = 15gpm
0 10000 20000 30000 40000 50000 60000 70000
Cumulative Volumetric Loading (gal)
Figure 4-13. Phase I and II total phosphorus cumulative loading results.
Phase l+ll - Chemical Oxygen Demand
600
500 -
400 -
O)
E,
Q
O
O
300 -
200 -
100
0 10000 20000 30000 40000 50000 60000 70000
Cumulative Volumetric Loading (gal)
Figure 4-14. Phase I and II COD cumulative loading results.
41
-------
Phase l+ll Influent & Effluent
100
—• Influent
•-o— Effluent
^—^— Dry solids mix
80 -
10 100 1000
Diameter (pm)
Figure 4-15. Phase I and II PSD summary.
4.3.3 Phase III - Performance under Varied Hydraulic and Concentration Conditions
As described in Section 3.5.3, Phase III testing focused on determining the unit's hydraulic flow
capacity and how well it handles spike loads of constituents. Phase III had three distinct parts:
• Part 1: Hydraulic capacity with clean water;
• Part 2: Hydraulic capacity with synthetic wastewater (regular constituent feed
concentrations);
• Part 3: Hydraulic capacity with spiked constituents (four times constituent feed
concentrations).
The Phase III tests were performed first because the information gathered in Phase III would
help set the flow rates in Phases II and IV.
4.3.3.1 Flow Data
In Phase III Part 1, clean water was used to determine the maximum hydraulic capacity of the
system before water bypassed the unit and whether drain backup would occur, resulting in
potential flooding of the catch basin. The test started at 10 gpm and ran for a minimum of
15 minutes. The flow rate was then increased at 5 gpm increments, and the test was rerun until
bypass occurred. Test Phases III Part 2 and III Part 3 were identical to Phase III Part 1, with the
exception that constituents were added to the clean water.
42
-------
The Phase III-l data are shown graphically in Figures 4-16, 4-17 and 4-18, representing the
relationship between influent and effluent flow rates, between influent flow rate and tank water
depth and the drawdown flow rate as a function of water depth. The elevation of the bypass
siphon was 60 in. above the tank floor, and served to prevent water depths greater than 60 in. for
these flow conditions.
50
40
CL
CD
£
ffi)
QC
30 -
; 20-
E
LLJ
10-
Effluent Flow Rate as a Function of Influent Flow Rate
• Effluent 1
O Effluent 2
T Effluent 3
10 20 30
Influent Flow Rate (GPM)
50
Figure 4-16. Phase III Part 1 relationship between influent and effluent flow rates using
clean water.
43
-------
Tank Water Depth as a Function of Influent Flow Rate
60-
40-
20
• Effluent 1
O Effluent 2
T Effluent 3
* *
10
40
50
20 30
Influent Flow Rate (GPM)
Figure 4-17. Phase III Part 1 tank water depth as a function of influent flow rate.
Drawdown: Flow Rate vs Water Level
20
15
E
Q.
CD
I 10
DC
5 -
50 48 46 44 42 40
Tank Water Level (in)
38
33
Figure 4-18. Phase III Part 1 drawdown flow rates.
44
-------
The data show that the influent and effluent flow rates through the system are nearly identical
once the flow is greater than approximately 15 gpm. The data also show the Up-Flo™ Filter can
operate up to approximately 35 gpm before the bypass level is triggered. The drawdown data
snowed that the drawdown time for the device was less than one hour and was linearly related
(visual assessment only) to the water level in the tank.
The same information was graphed for Phase III Part 2 and Phase III Part 3 (Figures 4-19, 4-20,
4-21, and 4-22, respectively for the relationship between flow rates and between influent flow
rate and water depth in the tank. Phase III Part 2 and Phase III Part 3 used the same media bags
as in Phase III Part 1.
Effluent Flow Rate as a Function of Influent Flow Rate
40
B 30 -
ra
20 -
u
E
LLJ
10 -
10 20 30
Influent Flow Rate (GPM)
40
SO
Figure 4-19. Phase III Part 2 relationship between influent and effluent flow rates.
45
-------
Tank Water Depth as a Function of Influent Flow Rate
70
SO-
'
50 i
jf.
c
ro
,E 40 -
-------
Tank Water Depth as a Function of Influent Flow Rate
70
60 -
50 -
40
30 -
Q
10 -
10 20 30
Influent Flow Rate (GPM)
50
Figure 4-22. Phase III Part 3 tank water depth as a function of influent flow rate.
The results of the flow rates through the media as a function of influent flow rate were compared
in Table 4-7 and graphically in Figure 4-23. The results show that, in general, the effluent flow
rates were comparable to the influent for all flow rates tested (up to and past the point where the
bypass was activated). Hydraulic performance appears to decrease during the 40 gpm and
particularly the 45 gpm testing in Phase III Part 3. When Figure 4-22 is evaluated with Table 4-7
and Figure 4-23, it appears that the bypass siphon (with an elevation of 60 in.) was preventing
the tank water level from exceeding 60 in., and at influent flows greater than 30 gpm, a portion
of the effluent was likely untreated bypass water.
Table 4-6. Phase III Influent and Effluent Flow Summary
Influent Flow
Rate (gpm)
10
15
20
25
30
35
40
45
Phase III Part 1
8.43
15.5
20.0
24.2
31.1
36.3
40.3
47.1
Effluent Flow Rate (gpm)
Phase III Part 2
7.39
15.0
20.0
22.1
34.9
36.4
40.7
48.1
Phase III Part 3
10.3
14.6
16.9
20.5
32.0
33.6
38.2
41.9
47
-------
Flow Rate - Phase
10 15 20 25 30 35
Flow Conditions (GPM)
40
45
Figure 4-23. Comparison of influent versus effluent flow rates for Phase III hydraulics
testing.
4.3.3.2 Analytical Data
Samples were collected during Phase III-2 and Phase III-3 testing at each flow rate condition (10,
15, 20, 25, 30, 35, 40, and 45 gpm). The analytical data are summarized in Tables 4-8 and 4-9.
For Phase III-2, the TSS and SSC analytical data showed a reduction starting above 90% and
decreasing to 0% at the two highest flow rates (40 and 45 gpm settings, 45 and 50 gpm
measured). TP removals ranged from <0% to 65%, while COD removals ranged from <0% to
85%. At the higher challenge concentrations of Phase III-3, performance degradation was noted
much sooner for all parameters compared to Phase III-2. The results are shown graphically in
Figures 4-24 through 4-27 for Phase III-2 and 4-28 through 4-31 for Phase III-3. The graphics
illustrate the much more rapid loss of performance in Phase III-3. This would be expected, since
the device would be challenged beyond its design flow capabilities, and a portion of the flows
would pass through the bypass mechanism without treatment.
48
-------
Table 4-7. Phase III Part 2 Analytical Data
Influent Flow Rate
Analyte (gpm)
TSS 10.7
6.43
21.6
24.6
39.2
44.7
48.9
50.8
SSC 10.7
6.43
21.6
24.6
39.2
44.7
48.9
50.8
TP 10.7
6.43
21.6
24.6
39.2
44.7
48.9
50.8
COD 10.7
6.43
21.6
24.6
39.2
44.7
48.9
50.8
Influent Cone.
(mg/L)
140
110
209
125
49
283
45
45
120
215
242
184
71
391
71
49
14
32
19
20
13
46
36
45
35
30
41
43
286
105
168
130
Effluent Cone.
(mg/L)
5
6
<5
8
3.3
17
26
45
<5
8.3
3.6
13
7.1
20
32
49
79
54
61
13
11
16
33
32
71
60
34
43
44
164
367
48
Removal
Efficiency (%)
96
95
>99
94
93
94
42
0
>99
96
99
93
90
95
55
0
-464
-69
-221
35
15
65
8.3
29
-103
-101
18
1.4
85
-56
-118
63
49
-------
Table 4-8. Phase III Part 3 Analytical Data
Influent Flow Rate
Analyte (gpm)
TSS 11.4
15.7
19.5
25.4
31.5
31.9
41.1
45.6
SSC 11.4
15.7
19.5
25.4
31.5
31.9
41.1
45.6
TP 11.4
15.7
19.5
25.4
31.5
31.9
41.1
45.6
COD 114
15.7
19.5
25.4
31.5
31.9
41.1
45.6
Influent Cone.
(mg/L)
331
253
430
624
314
370
511
575
NA
311
396
671
292
416
415
603
123
150
168
216
79
132
197
229
275
363
463
264
151
377
181
207
Effluent Cone.
(mg/L)
13
75
138
269
219
255
511
409
9.8
NA
142
273
340
320
824
399
60
59
107
162
138
187
237
242
27
89
152
190
186
188
222
198
Percent
Efficiency (%)
96
70
68
57
30
31
0
29
NA
NA
64
59
-16
23
-99
34
51
61
36
25
-75
-42
-20
-5.7
90
75
67
28
-23
50
-23
4.3
50
-------
Total Suspended Solids: Phase 111-2
ouu
250 -
,§ 200 -
c
| 150-
<-> 100 -
50 -
n .
i i Influent
(=> Effluent
—
— I
P
— i
. — .
i i
LL D_
6,4 10.7 21.6 24.6 39.2 44.7
Measured Flow Rate (gpm)
Figure 4-24. Phase III Part 2 TSS influent and effluent results.
48.9
50.8
Suspended Sediment Concentration: Phase III-2
500
400 -
§ 300
'•+-»
1
8
g 200
o
o
03
(f)
100
Influent
Effluent
6.4 10.7 21.6 24.6 39.2 44.7
Measured Flow Rate (gpm)
Figure 4-25. Phase III Part 2 SSC influent and effluent results.
48.9 50.8
51
-------
Total Phosphorus: Phase 111-2
80 -
§> 60
40 -
o
O
£L
6.4 10.7 21.6 24.6 39.2 44.7 489 50.8
Measured Flow Rate (gpm)
Figure 4-26. Phase III Part 2 total phosphorus influent and effluent results.
Chemical Oxygen Demand: Phase III-2
'-tuv
5* 300 -
O)
E.
c
0
£ 200 -
8
c
0
O
Q
g 100-
n .
i
i
3 Influent
3 Effluent
I — i
f .|
1
6.4 10.7 21.6 24.6 39.2 447
Measured Flow Rate {gpm)
Figure 4-27. Phase III Part 2 COD influent and effluent results.
48.9 50.8
52
-------
Total Suspended Solids: Phase 111-3
/uu -
600 -
0) 500 -
c
•2 400 -
P
Concenti
CO
0
o
$ 200-
100 -
n .
i i Influent
i= Effluent
]
—
1
1 — 1
— 1
— 1
1 —
—
1 — 1
11,4 15.7 19.5 25.4 31.5 31.9
Measured Flow Rate (gpm)
Figure 4-28. Phase III Part 3 TSS influent and effluent results.
41.1
45.6
Suspended Sediment Concentration: Phase III-3
1000
800 -
§ 600 -
03
g 400-
O
o
(D
CO
200 J
Influent
Effluent
11.4 15.7 19.5 25.4 31.5 31.9
Measured Flow Rate (gpm)
Figure 4-29. Phase III Part 3 SSC influent and effluent results.
41.1
45.6
53
-------
Total Phosphorus: Phase 111-3
300
250 -
£? 200 -
£ 150 -
c
O 100 -
Q.
50 -
Influent
Effluent
11.4 15.7 19.5 25.4 31.5 31.9 41.1 45.6
Measured Flow Rate (gpm)
Figure 4-30. Phase III Part 3 total phosphorus influent and effluent results.
Chemical Oxygen Demand: Phase III-3
500
400 -
o 300
Sfl
00
I
§
O
Q
O
200 -
100 -
Influent
Effluent
11.4 15.7 19.5 25.4 31.5 31.9
Measured Flow Rate (gpm)
Figure 4-31. Phase III Part 3 COD influent and effluent results.
41.1
45.6
54
-------
PSD analysis was also performed for all samples in Phase III. The results are shown graphically
in Figures 4-32 and 4-33. For Phase III Part 2, the influent mean dso was 60 um, while for Phase
III Part 3, the mean influent dso was 43 um. The mean effluent dso for Phase III Part 2 was 11 um
and for Phase III Part 3 was 39 um. This poorer performance in reducing the d50 of the influent
was not unexpected given the higher loading entering the filter during Phase III Part 3. This
indicates that the Up-Flo™ Filter was capable of removing particulates from the influent during
normal operations, and removals are reduced when the filter is challenged, and part of the flow is
bypassed, as would be expected.
Phase 111-2 Influent & Effluent
100
Influent
o— Effluent
Dry solids mix
o
0
10 100
Diameter (|jm)
Figure 4-32. Phase III Part 2 PSD summary.
1000
55
-------
Phase 111-3 Influent & Effluent
100
Influent
Effluent
Diy solids mix
10 100
Diameter (IJITI)
1000
Figure 4-33. Phase III Part 3 PSD summary.
4.3.4 Phase IV— Contaminant Capacities at High Hydraulic Throughput
As described in Section 3.5.4, in Phase IV the system was run to exhaustion (similar to Phase II),
except that the unit was under higher hydraulic loads and proportional contaminant loads.
The unit was operated under continuous flow conditions at a constant flow rate of 32 gpm until
the unit plugged with solids, or the contaminant absorption capacity was exceeded. The test plan
specified a flow rate of 30 gpm, based on the vendor's claims that the system could treat water at
a maximum flow exceeding 20 gpm.
During the first day (approximately two hours into the testing), the TO observed the media bags
"broke through" their mesh retainer, causing visible solids in the effluent. New bags were
installed and the test rerun the following two days. No samples were analyzed from the first day.
The testing under sustained contaminant and flow loading conditions until failure highlighted a
failure mode that had not been anticipated by the vendor. Under these conditions, failure
occurred through what appeared to be inadequate support of the top flow distribution media,
allowing bypassing to occur within the filter module, as opposed to bypassing through the bypass
mechanism. Because of the nature of this failure mode, the protocol was modified and additional
samples were taken during the first two hours of the rerun.
56
-------
4.3.4.1 Analytical Data
As described above, samples were collected in accordance with the test plan, plus additional
supplemental samples were collected to confirm test observations. A total of 15 sets of samples
were collected and analyzed. The results of the testing are summarized in Table 4-10. Visual
evidence of breakthrough was noted at 19,800 gallons. Two sets of confirmatory samples were
collected and analyzed once failure was observed. As anticipated, the Up-Flo™ Filter
performance was more variable than during the earlier runs, even for TSS and SSC. The median
removal efficiency for TSS and SSC during Phase IV was 62%, the median TP removal
efficiency was less than -8%, and the median COD removal was -42%. Comparison of this data
to the Phase I and Phase II data shows that the ability of the Up-Flo™ Filter to remove dissolved
and fine particulate pollutants may be compromised by this failure mode, particularly at flow
rates 150% above the design flow rate. The data is displayed graphically in Figures 4-34 through
4-37.
Table 4-9. Phase IV Analytical Summary
Influent Concentration (mg/L) Effluent Concentration (mg/L) Removal Efficiency ("/o)1
Analyte Mean
TSS 131
SSC 121
TP 42
COD 66
Median
95
102
36
59
Max
480
389
163
180
Min
<5
<5
0.9
18
Mean
45
46
39
107
Median
36
39
39
84
Max
113
129
80
370
Min
6.5
7.3
3
42
Mean
65
62
7
-63
Median
62
62
-8
-42
Max
95
93
81
41
Min
-1,640
-1,420
-4,680
-1,960
1. Mean and median removal efficiency is a function of mean and median influent and effluent concentrations, and
maximum and minimum removal efficiencies are a function of individual paired data points.
57
-------
500
400 -
Phase IV - Total Suspended Solids
300 -
O)
CO
CO
200 -
100 -
5000 10000 15000 20000
Cumulative Volumetric Loading in Phase IV (gal)
Figure 4-34. Phase IV TSS influent and effluent cumulative loading results.
Phase IV - Suspended Sediment Concentration
500
—••— Influent
—-O— Effluent
400 -
5000 10000 15000 20000
Cumulative Volumetric Loading in Phase IV (gal)
Figure 4-35. Phase IV SSC influent and effluent cumulative loading results.
58
-------
180
Phase IV - Total Phosphorus
o>
£
160 -
140 -
120 -
100 -
—•— Influent
—-O— Effluent
0 5000 10000 15000 20000
Cumulative Volumetric Loading in Phase IV (gal)
Figure 4-36. Phase IV total phosphorus influent and effluent cumulative loading results
Phase IV - Chemical Oxygen Demand
400
350 -
300 -
250 -
CO
200 -
Q
O
O
150 -
100 -
50 -
0
—•— Influent
— -O— Effluent
0 5000 10000 15000 20000
Cumulative Volumetric Loading in Phase IV (gal)
Figure 4-37. Phase IV COD influent and effluent cumulative loading results.
59
-------
The PSD analysis for Phase IV (Figure 4-38) shows that the Up-Flo™ Filter was not as effective
at reducing the influent dso during its operation at high hydraulic loadings. This is likely due to
the breakthrough of the filter bags and holder seen during this phase. The operation of the filter
was halted when the bag breakthrough was noticed visually, but it likely occurred to a slight
extent prior to its being visible in the operation.
Phase IV Influent & Effluent
100
Influent
1>— Effluent
Dry solids mix
10 100 1000
Diameter (|jm)
Figure 4-38. Phase IV particle size distribution analysis.
4.4 Phases I-IV Data Summary and Discussion
The flow and analytical data in the four test phases provided the following general observations:
• The Up-Flo™ Filter was capable of removing sediments from the influent water. TSS
and SSC removals were variable, resulting primarily from variable influent
concentrations with effluent concentrations remaining fairly consistent. The mean
removal efficiencies during typical operating conditions were slightly less than 80%, and
in some circumstances were as high as 90% to 95%. Performance was poorer when the
hydraulic flows or pollutant loadings were higher.
• Particle size distribution analysis confirmed these results, with poorer removals (or no
removal) occurring during the test phases with sustained high hydraulic flows or pollutant
loadings.
60
-------
• Total phosphorus removals ranged from negative to approximately 60%. COD removals
also ranged from negative to greater than 85%. In general, filter performance improved as
the filter aged, up to the point where failure began.
• The failure mechanism noted by the TO was not the one anticipated by the vendor. The
vendor indicated that the Up-Flo™ Filter would fail by having the filter bags clog,
forcing a build up of water in the tank, which would eventually reach the bypass and flow
out through the bypass. Instead, the TO noticed (at the ends of Phases II and IV) that the
water built up in the tank to a specific level, indicating that the filter was clogging.
However, prior to activation of the bypass, the pressure on the bags apparently built up to
a level sufficient to move the bags and mesh supports in the filter module. This uplifting
of the bags provided an opening large enough (at one corner) to allow water to flow
freely past the bags. This failure was noted in two ways: (1) the edges of the bags were
noted above the effluent opening of the cartridge container, and (2) the water level
suddenly dropped noticeably after the slow buildup to the bypass level.
In general, the Up-Flo™ Filter was capable of removing solids consistent with claims made by
the vendor, at the concentrations used during testing. It is anticipated that the results of this
series of tests could be adjusted to calculate loadings throughout the filter's life and used to
develop design curves that can be used to predict behavior when challenged by lower TP and
hydrocarbon concentrations. What is unknown (from a performance prediction standpoint) is
what effect the higher hydrocarbon loadings had on blinding the filter bags. This data is not
easily translated without additional work at different hydrocarbon loadings, where the effect of
hydrocarbon loading on filter life can be evaluated.
4.4.1 Installation and Operation & Maintenance Findings
The TO performed O&M on the system as outlined in the vendor's written O&M procedures
between test phases and as necessary during testing. O&M procedures and observations focused
on:
• Ease of installation;
• Weight of filter media bags, before and after testing;
• Clarity of written O&M procedures;
• Ease and time needed to clean unit and replace filter media; and
• Characteristics of waste materials.
4.4.1.1 Installation
To evaluate the ease of installation of the Up-Flo™ Filter, the TO installed the system in the test
rig supplied by the vendor in accordance with the vendor's instructions for use in a catch basin.
In general, the TO found the installation instructions were clear and the procedures were simple
to follow.
61
-------
4.4.1.2 Filter Media Bags
The TO was unable to observe differences in the sizes and dry weights of the filter media bags
from phase to phase since the bags were shipped slightly damp. Therefore, obtaining a point of
comparison was impossible. According to the vendor, the net weight of the carbon-based filter
bags is approximately 50 Ib. Two bags of the CPZ media were installed in the unit between
phases (chronologically, new filter bags were installed after Phases III, II and IV). Because of the
height of the test unit compared to the location of the filter cartridge, it was difficult to lift the
heavy filter bags in and out, especially one-to-two days after use when they were still nearly
soaked (although not dripping to a measurable extent). The TO was concerned that the straps
used to lift the bags in and out of the cartridge and the device would not hold up during
installation and removal. These concerns were unfounded. The bags stayed intact until they were
sliced open to observe the depth of penetration of the pollutants into the filter media.
4.4.1.3 General O&M/Svstem Cleanout
System cleanout consisted of pumping down the water to the sediment level, taking care not to
disturb the sediment. One person entered the device and one person remained outside to hand in
materials as needed. The replacement of the bags consisted of opening the module lid, removing
the top layer of mesh, the two bags and the bottom layer of mesh. The interior of the filter
module was wiped down with a paper towel and tap water to remove grit trapped along the
edges. The sediment was cleaned out between phases using a wet/dry shop vacuum. Once the
filter module was visually clean, new mesh and filter bags were installed as outlined in the O&M
manual with care taken to fit the bags to the edges of the filter module and to fit the top mesh
below the mesh sill of the filter module. The typical O&M session took between 30 and 45 min
with approximately half of the time devoted to pumping down the water in the tank.
4.4.1.4 Waste Material Characterization
Waste material characterization focused on two primary areas: physical and chemical. Physical
characterization determined the mass and volume of waste material generated during a cleanout
session, while chemical characterization determined hazardous characteristics important in waste
disposal considerations.
As waste materials were generated, representative composite samples of the recovered sediments
were submitted for analysis for sediment COD and sediment phosphorus. Three samples from
different sections of the sump were collected and analyzed. The results were as listed in
Table 4-11. The hopper solids were also tested and the solids' concentration was >0.0875 g TP/g
solid. The statistical analysis showed less than 20% deviation among the hopper solids samples
indicating that the contaminants of concern were well-distributed across the hopper solids. While
there are no waste disposal regulations that specifically address these pollutants, the results
indicate that the sump is capable of trapping particulate pollutants and that further testing may be
required if the influent water contains one or particular contaminants of concern, such as metals
(which were not included in the scope of this project).
62
-------
Table 4-10. Characterization of Material Captured in Up-Flo™ Filter Sump
Sample Number COD (g COD/g waste) TP (g TP/g waste)
1 0.070 0.016
2 0.10 0.022
3 0.087 0.017
Average 0.086 0.018
Standard Deviation 0.015 0.0034
COV 0.18 0.19
4.5 Summary of Findings
A newly maintained Up-Flo™ System, operating in the design range, is capable of reducing
sediment concentrations in this test wastewater in a range of 50% to 90%, as measured by TSS
and SSC. Hydrocarbon removals, as measured by COD analyses, were highly variable and
ranged from negative to >85%. TP removals were in the same range as the COD removals. The
TO observed the following regarding the removals of TP and COD. The filter performance was
optimum during the middle of the filter run (after the filter had 'aged' and before breakthrough
began). Filtration performance was best when the filter was operating on an intermittent schedule
and at the design flow rates or below. This is in agreement with filter treatment theory.
An Up-Flo™ with new filter media can accept a hydraulic flow of up to approximately 35 to
40 gpm, without bypassing, depending on the concentration of contaminants in the wastewater.
The maximum treated flow decreases as the filter media trap contaminants, preventing water
from flowing through the filter bags. The activation of the bypass was only observed during the
testing across the operational flow rates (Phase III) testing. A different failure mechanism (where
the pressure on the bags was sufficient to dislodge the bags and open a flow path through the
cartridge) was observed in Phases II and IV. This failure mechanism was new to the vendor, who
indicated that this failure mechanism had not been noted before in the vendor's laboratory. The
TO supposes that this may be due to the different test mixtures used in the vendor's laboratory
compared to the TO, who was following the test plan. The test plan had a mixture that was much
closer to washwater than stormwater.
In addition to hydrocarbon and phosphorus treatment, the Up-Flo™ system was also capable of
reducing suspended solids concentrations in the treated effluent. Sediment removal efficiency
was measured three ways:
1. the TSS and SSC analytical methods;
2. theoretical methods (measuring the mass of solids fed into the synthetic wastewater by
the test rig); and
3. particle size distribution comparison of influent and effluent.
An important consideration in determining overall system efficiency is the propensity of
contaminants to plug the filter media, resulting in untreated wastewater bypassing the filter
media. When the Up-Flo™ Filter failed in the method described by the vendor, it was simple to
63
-------
verify visually. The water in the tank built up to the bypass level. This would be easy to observe
in the field also if the storm inlet is covered by a grate. The failure due to the shifting of the bags
and mesh will not be visible in normal applications. It was visible here because the effluent
flowed directly into the treatment basin prior to discharge to the TO's sewer system.
Filter media blinding, which is a function of the influent flow rate and pollutant loading, did not
occur immediately, even at the high flow rate and high influent concentration conditions.
However, as can be seen in the data for Phases III (high concentration) and Phase IV (high
sustained flow rate) when compared to the Phase I+II results, treatment efficiency is decreased
across the run and the run is shortened. Because of the elevated concentrations of detergents and
because of the behavior of the sediments as floes, the TO can only predict that performance in
the field would be extended compared to that in the laboratory. The length of that extended
performance is unknown because of the manner in which the blinding occurred, with an oily
slime appearing on the media face and oil particles plus detergent creating rings around the test
tank. The results of these tests are more directly applicable to the performance of an Up-Flo™
Filter at hotspots where substantial vehicular maintenance and washing could be expected.
O&M procedures are relatively simple and can be completed in approximately 30-45 min.
64
-------
Chapter 5
Quality Assurance/Quality Control
The test plan included a QAPP with critical measurements identified and several QA/QC
objectives established. The verification test procedures and data collection followed the QAPP,
and summary results are reported in this section. The full laboratory QA/QC results and
supporting documentation are presented in Appendix C.
5.1 Audits
The VO conducted one audit of the PSH Environmental Engineering Laboratory at the start of
the verification test. The audit found that the field and laboratory procedures were generally
being followed, and that the overall approaches being used were in accordance with the
established QAPP. Recommendations for changes or improvements were made, and the
responsible parties responded quickly to these recommendations.
5.2 Precision
Throughout the verification test, the laboratory performed laboratory duplicates or matrix
spike/matrix spike duplicates to monitor laboratory precision. Field duplicates were collected to
monitor the overall precision of the sample collection and laboratory analyses. The test plan data
quality objectives for precision were based on laboratory precision for the analyses. The test plan
did not set field precision targets, as it was recognized that precision impacted by sampling and
constituent mixtures would be highly constituent- and equipment-dependent.
The relative percent difference (RPD) recorded from the sample analyses was calculated to
evaluate precision. RPD is calculated using the following formula:
%RPD = ^=^ x 100% (5-1)
x
where:
x\ = Concentration of compound in sample
X2= Concentration of compound in duplicate
x = Mean value of xi and X2
5.2.1 Field and Laboratory Precision Measurements
The laboratory performed precision analyses in two methods: laboratory standard measurements
and analysis of field replicates. Triplicate analyses for all samples collected during Phase III (the
first phase chronologically) for TSS, SSC, COD, and TP were performed. These field samples
were individual bottles collected after the system had sufficient time to stabilize.
For the laboratory, the required analytical tolerance limits are 10% for all analytes used in this
test plan. The samples all fell within this tolerance, with the exception of TSS, which was within
the 30% tolerance seen for prior TSS sampling. Several papers have been written addressing the
65
-------
limitations and relationship between TSS and SSC, including one under review by Dr. Clark, the
principal investigator (PI) on this project. The data from the PSH laboratory where the TSS
results are 70% to 80% of the SSC values for the same samples is in agreement with that seen by
other researchers working with stormwater samples. The statistical analysis of the data contained
in that paper (based on 215 sample pairs) showed that there was no statistical difference between
the TSS and SSC results, indicating that the variability seen between samples is sufficiently large
to drown out the differences between the analytical methods. This is particularly true when
influent samples were analyzed (and not as true for effluent samples). These results are due to
the larger particles that are in the influent samples (and not in the effluent). The TSS sampling
methods are not easily able to sample particles larger than 100 to 200 um.
For COD and TP, the field replicates are not in the 0% to 25% COV range deemed tolerable by
the test plan. The reason for this difference is the non-continuous distribution of TP and COD in
the influent. In order to obtain the dosing required by the test plan, only periodic dosing (adding
periodic drops of solution, rather than a continuous stream) was required of the OBC and WBC
solutions. The solids dosing was more consistent, although problems were noted with dosing due
to clogging of the solids hopper and distribution system occasionally. This resulted in the
installation of a technician at the solids' hopper to monitor the dosing of the system.
The field precision results are summarized in Tables 5-1. All of the data are presented in the
Appendices to this report. These samples are based on triplicate influent samples collected
during Phase III Part 2.
Table 5-1. Replicate Laboratory Sample RPD Summary
Analyte
TSS
SSC
TP
COD
Number of
Samples
24
24
8
8
Mean
(mg/L)
109
168
56
71
Standard
Deviation
67
115
46
50
COV
0.61
0.69
0.82
0.71
All of the TOC laboratory data was within the established precision limits, although this analysis
may not have provided a true result for the samples, as discussed in this Section 5.5.
While the results were not always within the limits established by the test plan, the procedures
were reviewed regularly and standards analyzed. These standards' results showed that laboratory
procedures, calibrations, and data were found to be in accordance with the published methods
and good laboratory practice.
The design of the sampling program anticipated that precision might be low for some of the
constituents due to the nature of the water being tested. The sampling plan included collection of
several aliquots over time to make composite samples. The data evaluation also was based on
mean data collected over a large volume of flow and long time periods. This approach was used
to help mitigate minute-by-minute changes that might occur in the water, particularly in the
influent water. Also, the careful monitoring of the total volume of water used and the total mass
66
-------
of constituents fed to the system provided a basis for calculating influent concentration. The
sampling techniques and laboratory procedures were carefully reviewed before and during the
test. The procedures used were in accordance with best sampling practice, and the laboratory
methods and procedures were found to be performed in accordance with the published methods.
5.3 Accuracy
Method accuracy was determined and monitored using a combination of matrix spikes and
laboratory control samples (known concentration in blank water) depending on the method.
Recovery of the spiked analytes was calculated and monitored during the verification test.
Accuracy was in control throughout the verification test. Table 5-2 shows a summary of the
laboratory control sample recovery data.
Table 5-2. Laboratory Control Sample Data Summary
Analyte
TSS
ssc
TP (as P)
COD
Actual
(mg/L)
150
350
2.00
300
Measured
(mg/L)
159
344
2.25
294
cov
0.08
0.002
0.05
Deviation from
Standard
Concentration
6%
2%
13%
5%
All the samples were within the quality control limits, with the exception of one COD sample
(151 mg/L) which was much lower than the allowed limits. This does not raise a concern,
because all other COD standard samples were well within their limits. Samples associated with
the COD standard were spot-checked the next day to ensure that the problem was in the standard
only. This was confirmed when the sample analytical results were similar from Day 1 to Day 2
and the standard was measured at the desired level.
The balance used for TSS and SSC analysis was calibrated routinely with weights that were
National Institute of Standards and Technology (NIST) traceable. Calibration records were
maintained by the laboratory and inspected during the on site audits. The temperature of the
drying oven was also monitored using a thermometer that was calibrated with a NIST-traceable
thermometer. Pipettes and graduated cylinders had their calibrations confirmed using the
analytical balance and deionized water.
5.4 Representativeness
The testing procedures were designed to ensure that representative samples were collected of
both influent and effluent wastewater. Supervisor oversight and audits provided assurance that
procedures were being followed. As discussed earlier, the challenge in sampling wastewater is
obtaining representative samples. The data indicated that while individual sample variability
might occur, the long-term trend in the data was representative of the concentrations in the
wastewater, and redundant methods of evaluating key constituent loadings in the wastewater
67
-------
were used to compensate for the variability of the laboratory data. In addition, the results and
shape of the effluent curves were compared to known filter theory to evaluate abnormalities. For
example, while the models were not fitted to this data, it is well known that filter flow rate can be
modeled by a power equation with suspended solids loading or time as the independent variable.
This occurred in this case to the extent seen in prior laboratory work by the TO with up-flow
filters. In addition, the graphs of pollutant behavior over filter life showed the traditional
breakthrough curves, where filter performance was variable at the start of the run, optimal
performance was obtained after the filter aged slightly and the pollutant removals decreased as
the filter neared breakthrough.
The laboratories used standard analytical methods and written standard operating procedures for
each method to provide a consistent approach to all analyses. Sample handling, storage, and
analytical methodology were reviewed during the on-site and internal audits to verify that
standard procedures were being followed. The use of standard methodology, supported by proper
quality control information and audits, ensured that the analytical data were representative of the
actual wastewater conditions.
5.5 Completeness
The test plan set a series of goals for completeness. During the startup and verification testing,
flow data were collected for each day at a minimum of once per two hours for Phases II, IV, and
V, and once per active flow setting for Phases I and III. The flow records are 100% complete.
No scheduled analyses had to be omitted from the testing program. Less than seven TSS or SSC
samples were not sieved prior to analysis. In all cases but two, either the TSS or SSC sieved was
performed and while the protocol called for using the TSS only data to adjust the particle size
distribution for the mass above 250 um, in those cases where the TSS sieve data was missing, the
SSC unsieved and sieved comparison was used. For those two instances where no sieve data was
available, the samples were not included in the particle size distribution analysis presented here.
Given the number of samples collected (which exceeded the requirements of the test plan for all
phases but Phase II), these missing samples were not considered sufficiently important to rerun
the testing phase. Sufficient data was available to document the performance of the device. This
results in less than five omitted data points from a more than 200 data points per analytical
parameter, resulting in greater than 99% completeness, which exceeds the 80% completeness
goal for this program.
While COD was used as a surrogate organic measurement in the protocol to measure the capture
of hydrocarbons, the free product (un-emulsified hydrocarbons) in the device and in the flow
stream affected the repeatability of the tests even from aliquots drawn from the same sample
bottle. All samples collected for COD were analyzed, resulting in 100 percent completeness,
giving a reasonable indication of the bounds of performance of the Up-Flo™ Filter. Similar
variability was seen with the TP measurements because of the small additions of the WBC which
contained the bulk of the dissolved phosphorus. All samples were analyzed, resulting in 100%
completeness and allowing for the bounds of performance to be evaluated.
68
-------
Chapter 6
Vendor Supplemental Testing
The vendor requested that supplemental testing be conducted on the Up-Flo® Filter after they
reviewed the test data and results derived from the ETV testing. They expressed concerns that
the filter media breakthrough in the filter module was something they had not seen in testing at
their facility or in field applications. This test was conducted to examine whether the filter
module design should be modified to reduce the ability of the filter media to move within the
module coupled with whether the synthetic challenge water created a challenge that was beyond
the design considerations of the device and not indicative of real-world situations.
6.1 Up-Flo® Filter Modifications
The filter modules were redesigned to improve support and restraint and prevent the media bags
from shifting and potentially displacing vertically, observed during the original phase of testing.
The number of latches attaching the filter module lid was increased from one to three.
Additionally, the media restraint was redesigned by increasing the width of each structural side.
Figure 6-1 shows the Up-Flo® Filter module and Figure 6-2 shows the modifications made to the
module to improve the support on the filter media.
Latches
Media
Restrain
Conveyance
Channel
Media Pack
Filter Module
Angled Screen
Figure 6-1. Modifications to Up-Flo Filter module.
69
-------
Improved support across top of flow distribution media
Old Module New Module
Figure 6-2. Modifications to Up-Flo® Filter module showing improved support details.
6.2 Test Procedure Modifications
The testing procedures were modified, including modifications to the synthetic challenge water,
elimination of COD as an analyte based on the changes to the synthetic challenge water, and
omission of the Phase III and Phase IV tests. These modifications are outlined in greater detail in
this section.
6.2.1 Synthetic Challenge Water
The verification test was performed using synthetic water (Table 6-1) made from a mixture of
solids - one of which provided the particulate phosphorus required by the test plan. The
following products were used to make the synthetic challenge water:
Sil-Co-Sil® 250;
• Slow release phosphorus-supplying fertilizer; and
• Concrete plant sand sieved to a size of all passing through 5,000 jim.
Table 6-1. Modified Synthetic Challenge Water Concentrations
Concentration
Parameter (mg/L)
SSC 300
TSS 300
Total phosphorous (as P) 3
Reactive phosphorus (as P) 1
70
-------
A formula using a mix of the above named products/materials was made and tested in the
laboratory to determine the conformance to these specifications. The result of testing the ground
fertilizer for phosphorus content is 0.3 mg TP/g Scott's Lawn Starter Fertilizer. The amount of
fertilizer used was decreased from the amount used during the initial testing because phosphorus
recovery during initial testing was found to be greater than the target concentration. This higher
concentration may have been from the combination of the slow-release property of the fertilizer
and the grinding of the pellets into smaller particles, thus releasing more phosphorus because of
the increased surface area that comes into contact with water. For the supplemental testing, the
fertilizer replaced approximately 1% of the sand in the mixture to decrease the phosphorus
concentration to the target concentration.
The other constituents added to the synthetic challenge water in the initial testing (gasoline,
diesel fuel, motor oil, brake fluid, antifreeze, detergents, and windshield washer fluid) were
removed. Observations during the initial testing indicated that the synthetic challenge water
including the hydrocarbon constituents mixed with the solids to form a viscous substance that
was atypical of stormwater and could prematurely blind the filter media.
6.2.2 Analytical Methods
Constituent analysis for this testing included reactive and total phosphorus (RP and TP,
respectively), and solids (PSD, TSS, and SSC). COD was not analyzed because the hydrocarbon
mixture was removed from the synthetic challenge water.
Influent and effluent solids were characterized using wet sieve analysis on samples for particles
less than 20 |j,m to above 250 |j,m. Samples were sieved through stainless steel sieves with mesh
sizes of 20 |j,m, 38 |j,m, 63 |j,m, 106 |j,m, and 250 |j,m. This wet sieve analysis allowed a complete
characterization of the influent and effluent particle distribution from less than 20 |j,m to
5,000 |j,m. The results for the solids analysis were subdivided into removal for the following
particle size ranges:
• <20 urn
• 20-38 urn
• 38-63
• 63-106
• 106-250 urn
• >250
6.3 Synthetic Challenge Water Laboratory Analytical Results
During testing, 46 influent samples were collected during the normal constituent feed conditions
(Phase I, Phase II) and analyzed for the various constituents specified in the test plan. Table 6-2
provides a comparison of the mean analytical results for these influent samples versus the
analytical results for the synthetic challenge water mix specified in the test plan.
71
-------
Table 6-2. Synthetic Challenge Water Analytical Data Comparison to Desired Feed
Concentration
Measured Mean Desired Feed
Constituent Concentration (mg/L) Concentration (mg/L)
TSS 101 300
SSC 299 300
TP 1.26 3
RP 0.73 1
The mean synthetic challenge water data for the primary constituents were measured to be
approximately half of the desired target concentration for TP, approximately 75% of the targeted
RP concentration, approximately one-third the concentration for TSS, and 99% for SSC. A
review of the data shows that the COVs for all parameters ranged between 0 and 1.0. To confirm
reliability of the sampling and to assess the repeatability of the testing with new personnel,
testing was performed again to ensure that the sampling met the required criteria for efficient
solids capture. The differences in solids analysis procedure resulted in capturing almost all solids
by the SSC method but only approximately one-third by the TSS methodology1.
The hopper dosage measurements are consistent with the biases reported for TSS concentrations,
which typically underreport the total sediment concentration in the sample, especially for
sediment with a specific gravity greater than 1 and a dso greater than approximately 75 jim.2
Although the mean analytical TSS concentrations were lower than the 300 mg/L target
concentration goal, the hopper dose measurements suggest that the theoretical test plan
concentration was close to the 300 mg/L goal.
6.4 Test Results
This section summarizes the analytical data, flow data, and observations for the test phases
conducted during the supplemental testing. The efficiency values reported in this section are a
function of the total influent and total effluent concentrations.
6.4.1 Phase I - Performance under Intermittent Flow Conditions
The TSS, SSC, TP, and RP analytical data as related to cumulative volumetric loading on the
media are summarized in Table 6-3. The test plan required that a minimum of one set of samples
be collected each test day, however, the TO collected samples twice per day. The testing
organization collected a total of 20 sets of samples. The increase was to verify whether filter
media breakthough was occurring.
1 An in-depth discussion of solids recovery using the TSS and SSC analytical methods can be found in: Clark, S.E.
and Siu, C.Y.S. "Measuring Solids Concentration in Stormwater Runoff: Comparison of Analytical Methods."
Environmental Science & Technology. 2008, Vol. 42, No. 2, pp. 511-516.
2 Ibid.
72
-------
Table 6-3. Phase I Analytical Data Summary
Influent Concentration Effluent Concentration
(mg/L) (mg/L) Removal Efficiency (%)*
Analyte
TSS
ssc
TP (as P)
RP (as P)
Mean
85
288
1.32
0.73
Median
75
247
1.19
0.66
Max.
243
775
2.55
1.61
Min.
15
85
0.58
0.38
Mean
41
42
1.50
0.92
Median
37
37
1.42
0.86
Max.
68
79
2.27
1.37
Min.
24
20
1.12
0.60
Mean
52
86
-14
-25
Median
51
85
-19
-30
Max. Min.
81 -133
95 32
40 -291
37 -234
1. Mean and median removal efficiency is a function of mean and median influent and effluent concentrations, and maximum
and minimum removal efficiencies are a function of individual paired data points.
The median removal efficiency for TSS was 51%, while the median removal efficiency for SSC
was 85%. The mean and median influent SSC concentration was approximately four times
higher than the mean and median TSS concentrations; and the median TSS and SSC effluent
concentrations were nearly identical. The difference in sediment removal efficiencies can be
explained by the particle size distribution of the synthetic challenge water and the differences in
the analytical methods. The TSS analytical method requires the analyst to shake the sample and
collect an aliquot using a pipette, while the SSC analytical method utilizes the entire sample.
Therefore, the SSC analytical method is perceived as a more effective method to quantify the full
spectrum of solids including the coarser fractions of particles which may fall out of suspension
and finer fractions of particles which will tend to stay in suspension, and as a result, generally
yields higher removal efficiencies than results based on TSS. The Up-Flo® Filter was generally
not effective in treating total phosphorus or reactive phosphorus as presented in the form utilized
in the synthetic challenge water.
A graphical examination of the data also was conducted to illustrate the results discussed above.
Figures 6-3, 6-4, 6-5, and 6-6 compare the influent and effluent concentrations for TSS, SSC,
TP, and RP, respectively. Figure 6-7 shows the tank water levels for each test day.
The Phase I testing was conducted with new filter media bags installed in the Up-Flo® Filter. The
Up-Flo® Filter did not exhibit signs of clogging or blinding during the test run. A review of the
water depth measurements at each sample time showed that the tank water level remained
consistent between 38 and 42 in. No buildup of head was noted in the unit, further indicating that
the media capacity had not been exhausted in the Phase I testing.
73
-------
250
200 -
150 -
C/D
C/3
100 -
50 -
0
0--
—*— Influent
—-O— Effluent
O
0 2000 4000 6000 8000 10000 12000
Cumulative Volumetric Loading (gal)
Figure 6-3. Phase I TSS influent and effluent results.
14000
800
600 -
o»
•§
O
CO
to
400 -
200 -
—•— Influent
— O— Effluent
.-O—O-Q-O--O—QS
0 2000 4000 6000 8000 10000
Cumulative Volumetric Loading (gal)
Figure 6-4. Phase I SSC influent and effluent results.
12000 14000
74
-------
3.0
2.5 -
2.0 -
o>
E 1.5 4
1.0 -
0.5 -
0.0
—•— Influent
—-O— Effluent
0--0
0 2000 4000 6000 8000 10000
Cumulative Volumetric Loading (gal)
Figure 6-5. Phase I TP influent and effluent results.
12000 14000
1.8
1.6 -
1.4 -
1.2 -
1.0 -
Q_ 0.8 -
o:
0.6 -
0.4 -
0.2 -
0.0
—•— Influent
— O— Effluent
0 2000 4000 6000 8000 10000
Cumulative Volumetric Loading (gal)
Figure 6-6. Phase I RP influent and effluent results.
12000
75
-------
50
40 -
30 -
5
> 20 -
c
CO
10 -
Weir Height = 60 inches
i i
CM CO
Day 1 Day 2
Figure 6-7. Phase I tank water level.
CM CO
Day3
I I I I I I
CM CO -
-------
100
o
0
100
1000
Diameter
Figure 6-8. Phase I influent and effluent PSD summary.
6.4.2 Phase II —Determination of the Capacity of the Unit
Upon inspection of the filter media bags after Phase I testing, the bags were found to be covered
with sediment. The TO shared this information with the vendor, who requested that the Up-Flo®
Filter be equipped with new filter media bags prior to the start of the next test phase. Since each
phase began with new filter media bags, with the exception noted previously, Phase I and II data
were not combined during the supplemental testing.
The data are summarized in Table 6-4 and are expressed graphically in Figures 6-9 through 6-12.
The median SSC removal efficiency was 77%, while the median TSS removal efficiency was
41%. Similar to Phase I, the median influent SSC concentration was approximately three times
higher than the median influent TSS concentration, yet the median effluent TSS and SSC
concentrations were nearly identical. The Phase II data also show that the Up-Flo® Filter was
not effective at treating total or reactive phosphorus as presented in the form utilized in the
synthetic challenge water.
Figure 6-13 shows the water levels during each day of testing. At the beginning of the test, the
water level in the sump would rise to around the elevation of the bypass weir (60 in.). As the
testing progressed, the TO observed that the water level in the sump would take progressively
longer to reach the bypass weir elevation. On Day 14, after three consecutive days of the water
level in the tank failing to reach the bypass weir elevation, the TO concluded that the Up-Flo®
77
-------
Filter had reached a point where maintenance would be required to restore original operating
conditions, so the Phase II test was considered finished.
Analytical data results for the final three days of testing did not demonstrate an increase in
contaminant concentrations in the effluent, which could be anticipated if the filter mechanism
was breached and flows were exiting the Up-Flo® Filter without filtration. Figures 6-9 through
6-12 do not show a dramatic change in the effluent contaminant concentrations at the end of the
test. The TO concluded that, there was a change in conditions in the filter modules sufficient to
relieve the pressure in the filter modules and to decrease the head in the tank, but this change did
not result in contaminant concentration increase in the effluent.
Table 6-4. Phase II Analytical Data Summary
Influent Concentration
Results (mg/L)
Effluent Concentration
Results (mg/L)
Removal Efficiency (%}1
Analyte
TSS
ssc
TP (as P)
RP (as P)
Mean
110
307
1.26
0.73
Median
99
289
1.28
0.69
Max.
309
845
2.24
1.61
Min.
43
109
0.40
0.23
Mean
59
64
1.30
0.75
Median
58
65
1.25
0.71
Max.
87
101
2.53
1.38
Min.
30
33
0.51
0.27
Mean
47
79
-3.6
-3.5
Median
41
77
2.3
-2.2
Max.
83
93
46
67
Min.
-14
41
-78
-100
1. Mean and median removal efficiency is a function of mean and median influent
and effluent concentrations, and maximum and minimum removal efficiencies are a
function of individual paired data points.
400
300 -
en
E,
CO
co
200 -
100 -
—•— Influent
—-O— Effluent
20000 40000 60000 80000
Cumulative Volumetric Loading in Phase II (gal)
100000
78
-------
Figure 6-9. Phase II TSS influent and effluent results.
1000
OJ
E,
O
CO
800 -
600
400 -
200
—•— Influent
—-O— Effluent
20000 40000 60000 80000 100000
Cumulative Volumetric Loading In Phase II (gal)
Figure 6-10. Phase II SSC influent and effluent results.
3.0
2.5 -
2.0 -
O)
1.5 -
D_
I-
1.0
0.5 -
0.0
—•— Influent
— O— Effluent
20000 40000 60000 80000
Cumulative Volumetric Loading in Phase II (gal)
100000
79
-------
Figure 6-11. Phase II TP influent and effluent results.
2,0
1.5 -
O)
1.0 -
Q_
ec
0.5 -
0.0
—•— Influent
—-O— Effluent
0 20000 40000 60000 80000
Cumulative Volumetric Loading in Phase II (gal)
Figure 6-12. Phase II RP influent and effluent results.
100000
70
60 -
•> 50 -
c
"S
40 -
-------
PSD analysis also was performed on the Phase II samples, as shown in Figure 6-13. The mean
dso for the influent was 156 um and the mean effluent dso was 16 um. This confirmed the
manufacturer's claims that the Up-Flo® Filter would be capable of removing a high proportion of
the particulates in the solution.
100
Influent
—o — Effluent
Dry solids mix
1 10 100
Diameter (|jm)
Figure 6-14. Phase II influent and effluent particle size distribution summary.
1000
6.5 Sediment Retained in Sump
Figure 6-15 shows the depth of sedimentation in different areas in the sump after running both
Phases I and II. The letters on the figure correlate to grab sample locations. The greatest
sediment depth occurred near the filter modules. The water stream exited the influent pipe in this
general area. As a result, the larger particles most likely settled out of solution beneath the filter
modules.
Figure 6-16 presents the sieve analysis of the three sampled locations within the sump. The
distribution of the hopper solids is given also for comparison. The heavier solids in the mixture
tend to settle out near the inlet outflow area. The grade of the solids appears to become finer
(based on these samples) the further away from initial point of entry into the tank (around
Location E).
81
-------
I
I
N
F
G
Figure 6-15. Depth of sedimentation in sump.
Sample
Location
A
B
C
D
E
F
G
H
I
J
K
L
M
N
Depth
(in.)
4.9
4.0
2.5
4.4
2.8
2.0
1.1
1.3
0.8
1.0
1.4
1.3
1.5
1.5
Sample A
Sample G
Sample K
Hopper Solids
100
Particle Size ((.irn)
Figure 6-16. Sump particle size distribution analysis results.
1000
10000
82
-------
Based on the data shown in Figure 6-14, the mass of solids in the sump after the running of
Phases I and II was estimated to be 85.3 kg. The total sum of loads for both phases was
approximately 104 kg. This indicates that the device retained approximately 82% of the total
solids loading in the sump. The concentration of phosphorus retained in the sump was estimated
to be 31.3 mg PC>43" per gram of solids, or a total phosphorus mass of 3.5 kg PC>43", which
represents approximately 3% of the total solids loading. The increase of the phosphorus loading
from approximately 1% in the influent to 3% in the sump is likely attributable to the affinity of
phosphorus to be retained in sediments.
6.6 Test Summary and Discussion
The flow and analytical data result in the following general observations:
• The Up-Flo® Filter was capable of removing sediments from the influent. Removal
efficiencies for SSC were near 80% and were near 50% for TSS. The difference between
the TSS and SSC removal efficiencies are attributable to the TSS analytical procedure
quantifying only the finer fraction of sediment as opposed to the SSC analytical
procedure which quantifies a full spectrum of coarse and fine sediment. Most of the
sediments removed from the flows were retained within the sump.
• Particle size distribution analysis showed that the Up-Flo® Filter removed a high
proportion of the particulate sediments. The influent dso ranged from approximately
100 um to 300 um, and the effluent d50 was approximately 15 um.
• The Up-Flo® Filter was generally not effective at removing total or reactive phosphorus
as presented in the form utilized in the synthetic challenge water during this phase of
testing.
• The Up-Flo® Filter is designed so that flows exceeding the filtration capacity discharge to
the bypass weir. It is anticipated that clogging of the filter bags over time would decrease
the filtration capacity, which would result in the water elevation and head increasing in
the tank. Flows reaching the bypass module elevation would pass through the weir in the
bypass module without undergoing filtration. Based on this supplemental testing and the
original ETV study, the TO observed that as the filter media ripens, conditions within the
filter modules change, resulting in an increase in the capacity of the flow through the
filter modules and a decrease in the driving head, instead of filter clogging decreasing the
flow through the filter module. This observation is demonstrated graphically in
Figure 6-12.
• The vendor's redesign of the media restraint and the latching mechanisms of the lid of the
filter module prior to the supplemental testing aimed to decrease the ability of the filter
media bags to shift within the filter module and let flows pass between the filter media
and the filter module walls. The latches were able to keep the filter bags encased within
the filter module. As the filter media ripens, it appears that conditions within the filter
modules change, allowing for an increase in the flow capacity through the filter module.
83
-------
• The decrease in the driving head due to the apparent increase in the flow capacity through
the filter modules as the filter media ripened did not coincide with an increase in effluent
analytical concentrations, as might be expected if the flows were bypassing the filter
media. Effluent concentrations toward the end of the Phase II test, when the tank water
level did not reach the weir elevation, were consistent with the effluent concentrations
observed at the beginning of testing.
84
-------
Glossary
Accuracy - a measure of the closeness of an individual measurement or the mean of a number of
measurements to the true value and includes random error and systematic error.
Bias - the systematic or persistent distortion of a measurement process that causes errors in one
direction.
Commissioning - the installation of the in-drain removal technology and start-up of the
technology using test site wastewater.
Comparability - a qualitative term that expresses confidence that two data sets can contribute to
a common analysis and interpolation.
Completeness - a qualitative term that expresses confidence that all necessary data have been
included.
Precision - a measure of the agreement between replicate measurements of the same property
made under similar conditions.
Protocol - a written document that clearly states the objectives, goals, scope, and procedures for
the study. A protocol shall be used for reference during vendor participation in the verification
testing program.
Quality Assurance Project Plan - a written document that describes the implementation of
quality assurance and quality control activities during the life cycle of the project.
Residuals - the waste streams, excluding final effluent, that are retained by or discharged from
the technology.
Representativeness - a measure of the degree to which data accurately and precisely represent a
characteristic of a population parameter at a sampling point, a process condition, or
environmental condition.
Source Water Protection Stakeholder Advisory Group -a group of individuals consisting of
any or all of the following: buyers and users of in-drain removal and other technologies,
developers and vendors, consulting engineers, the finance and export communities, and permit
writers and regulators.
Standard Operating Procedure - a written document containing specific procedures and
protocols to ensure that quality assurance requirements are maintained.
Technology Panel - a group of individuals with expertise and knowledge of in-drain treatment
technologies.
85
-------
Testing Organization - an independent organization qualified by the Verification Organization
to conduct studies and testing of mercury amalgam removal technologies in accordance with
protocols and Test Plans.
Vendor - a business that assembles or sells in-drain treatment equipment.
Verification - to establish evidence on the performance of in-drain treatment technologies under
specific conditions, following a predetermined study protocol(s) and test plan(s).
Verification Organization - an organization qualified by EPA to verify environmental
technologies and to issue verification statements and verification reports.
Verification Report - a written document containing all raw and analyzed data, all quality
assurance/quality control (QA/QC) data sheets, descriptions of all collected data, a detailed
description of all procedures and methods used in the verification testing, and all QA/QC results.
The test plan(s) shall be included as part of this document.
Verification Statement - a document that summarizes the Verification Report reviewed and
approved and signed by EPA and NSF.
Verification Test Plan - a written document prepared to describe the procedures for conducting
a test or study according to the verification protocol requirements for the application of in-drain
treatment technology. At a minimum, the test plan shall include detailed instructions for sample
and data collection, sample handling and preservation, precision, accuracy, goals, and QA/QC
requirements relevant to the technology and application.
86
-------
Appendices
A Test Plan
B UpFlo™ Filter O&M Manual
C Analytical Data
87
------- |