September 2006
                                  Revised July 2008
                                 06/29/WQPC-WWF
                                  EPA/600/R-06/136
Environmental Technology
Verification Report

Stormwater Source Area Treatment
Device

The Terre Hill Concrete Products
Terre Kleen™ 09
               Prepared by

            Penn State Harrisburg
          Middletown, Pennsylvania

        Under a Cooperative Agreement with
       U.S. Environmental Protection Agency

-------
   Environmental Technology Verification Report


     Stormwater Source Area Treatment Device

            The Terre Hill Concrete Products
          Terre Kleen™ 09 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 2006
                            Revised July 2008
   Note: Revised in July 2008 to include a change in the method the drainage area, runoff
volumes and peak runoff intensities were calculated.  The revised drainage area is described in
Section 3.2. The revised runoff volume calculation method is described in Section 5.1.1. The
  revised runoff data is reported in Table 5-2 and is used for the sum of loads calculations in
Table 5-4, Table 5-5, and Table 5-6. Section 5.3.3 was also added to provide additional detail
                      on the Hjulstrom Diagram analysis.

-------
                                       Notice
The  U.S. Environmental  Protection  Agency (EPA) through its  Office  of Research  and
Development has financially supported and collaborated with NSF International (NSF) under a
Cooperative  Agreement.    The  Water Quality  Protection  Center,  operating  under  the
Environmental Technology Verification Program, supported  this verification  effort.   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.

-------
                                      Foreword

The  following is the final report  on an Environmental  Technology Verification (ETV) test
performed for NSF International (NSF) and the United States Environmental Protection Agency
(EPA). The verification test for the Terre Hill Concrete Products'  Terre Kleen™ Stormwater
Treatment Device was conducted at a site in Harrisburg, Pennsylvania, maintained by the City of
Harrisburg Public Works Department.

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.
                                           11

-------
                                       Contents

Verification Statement	VS-i
Notice	i
Foreword	ii
Contents	iii
Figures	iv
Tables	v
Abbreviations and Acronyms	vi
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	2
            1.2.2   Verification Organization	2
            1.2.3    Testing Organization	3
            1.2.4   Vendor	4
            1.2.5    Verification Testing Site	4
Chapter 2 Technology Description	5
  2.1   Technology Overview	5
  2.2   Technology Description	5
  2.3   Applications	9
  2.4   Operation and Maintenance	9
  2.5   Performance Claim	10
Chapters Test Site Description	11
  3.1   Sizing Methodology	11
  3.2   Site Description	11
  3.3   Peak Flow Calculation	16
  3.4   Contaminant Sources and Site Maintenance	19
  3.5   Stormwater Conveyance System and Receiving Water	19
  3.6   Terre Kleen™ Installation	19
Chapter 4 Sampling Procedures and Analytical Methods	21
  4.1   Sampling Locations	21
           4.1.1    Upstream Influent	21
           4.1.2   Downstream Effluent	21
           4.1.3    Rain Gauge	21
  4.2   Monitoring Equipment	21
  4.3   Constituents Analyzed	22
  4.4   Sampling Schedule	22
  4.5   Field Procedures for Sample Handling and Preservation	23
Chapter 5 Monitoring Results and Discussion	24
  5.1   Storm  Event Data	24
           5.1.1    Flow Data Evaluation	24
  5.2   Monitoring Results: Performance Parameters	26
           5.2.1    Concentration Efficiency Ratio	26
           5.2.2   Sum of Loads	28
  5.3   Particle Size Distribution	29
           5.3.1    Particle Size Distribution with Sieve Data	29
                                           in

-------
           5.3.2   Particle Size Distribution with Coulter Counter Data	31
           5.3.3   Particle Size Distribution and HjulStrom Diagram Evaluation	36
  5.4   Retained Solids Analysis	38
           5.4.1   Particle Size Distribution of Retained Solids	38
           5.4.2   TCLP Analysis of Retained Solids	39
Chapter 6 QA/QC Results and Summary	41
  6.1   Laboratory/Analytical Data QA/QC	41
           6.1.1   Bias (Field Blanks)	41
           6.1.2   Replicates (Precision)	41
           6.1.3   Accuracy	44
           6.1.4   Representativeness	44
           6.1.5   Completeness	45
Chapter 7 Operation and Maintenance Activities	46
  7.1   System Operation  and Maintenance	46
Chapter 8 References	48
Appendices	49
A  Design and O&M Guidelines	49
B  Verification Test Plan	49
C  Event Hydrographs and Rain Distribution	49
D  Analytical Data Reports with QC	49


                                        Figures

Figure 2-1. Terre Kleen™ schematic and flow diagram	6
Figure 2-2. Adjusted Hjul Strom diagram (provided by vendor)	8
Figure 3-1.   Site topographic map showing the sampling location and location of the Terre
    Kleen™	13
Figure 3-2.  Aerial  photograph of Harrisburg Public Works Yard showing outlet and drainage
    area delineation	14
Figure 3-3. Photograph of the drainage area	14
Figure 3-4.  View from the sampling location across the paved lot to the edge of the watershed
    (see stop sign in photograph on right)	15
Figure 3-5. Outlet sampling location for 15-inch reinforced concrete storm drain pipe	15
Figure 5-1.  Coulter analysis comparison by count of the Terre Kleen™ influent and effluent
    total particle count for Event 8	32
Figure 5-2.  Coulter analysis comparison by volume of the Terre Kleen™ influent and effluent
    particle volume for Event 8	32
Figure 5-3. Complete particle size distribution for influent and effluent samples from Event 8.33
Figure 5-4.  Particle size distribution for influent and effluent samples from all  sampled storm
    events using mean dio,  d25, dso, dys, and dgo	34
Figure 5-5. Particle size distribution for influent and effluent samples using median dio, d25,  dso,
    dys, andd9o	35
Figure 5-6.  Hjulstrom  diagram plotting the 95th percentile particle size remaining in solution
    versus the horizontal water velocity through the plates	37
Figure 5-7.  Particle size distribution for material captured in the sediment storages areas of the
    Terre Kleen™	39
                                            IV

-------
                                        Tables

Table 2-1. Terre-Kleen™ Sizing Chart	7
Table 3-1. Preliminary Suspended Solids Sampling at the Harrisburg Public Works Yard	11
Table 3-2. Rainfall Depth (in.)	17
Table 3-3. Rainfall Intensities (in./hr)	17
Table 3-4. Peak Flow Calculations (cfs)	17
Table 3-5. Peak Flow Calculations (cfs) Using Time of Concentration	18
Table 4-1. Constituent List for Water Quality Monitoring	22
Table 5-1. Summary of Events Monitored for Verification Testing	25
Table 5-2. Peak Discharge Rate and Runoff Volume Summary	26
Table 5-3. Monitoring Results and Efficiency Ratios for Sediment Parameters	27
Table 5-4. Sediment Sum of Loads Results	29
Table 5-5. Particle Size Distribution Analysis Results (Particle Sizes > 250 |j,m)	30
Table 5-6. Particle Size Distribution Analysis Results (Particle Sizes Smaller than 250 |j,m).... 31
Table 5-7. Results for Cleanout  Solids	40
Table 6-1. Sampler Calibration for TSS using Sil-Co-Sil Mixture	42
Table 6-2. Field Duplicate Sample Relative Percent Difference Data Summary	43
Table 7-1. (a) Initial cleanout of  the sedimentation  chamber, (b) Bottom of primary chamber
    after dewatering and during sediment cleanout	46
Table 7-2. Primary chamber nearing the end of cleanout	47

-------
                          Abbreviations and Acronyms
BMP
cfs
cov
DTU
dxx
EMC
EPA
ETV
ft2
ft3
g
gal
gpm
HOPE
hr
IDF
in.
kg
L
Ib
NRMRL
mg/L
min
mm
NSF
O&M
PSH
QA
QC
SOL
SOP
ssc
TCLP
THCP

TO
TSS
um
VO
WQPC
yd3
yr
Best management practice
Cubic feet per second
Coefficient of variation
Data transfer unit
Particle size at xx percentile of cumulative volume
Event mean concentration
U.S. Environmental Protection Agency
Environmental Technology Verification
Square feet
Cubic feet
Gram
Gallon
Gallon per minute
High density polyethylene
Hour
Intensity-Duration-Frequency (curve)
Inch
Kilogram
Liter
Pound
National Risk Management Research Laboratory
Milligram per liter
Minute
Millimeter
NSF International
Operations and maintenance
Penn State Harrisburg (TO)
Quality assurance
Quality control
Sum of the loads
Standard Operating Procedure
Suspended sediment concentration
Toxicity Characteristic Leachate Procedure
Terre  Hill Silo  Company, Inc.  (T/D/B/A Terre Hill Concrete Products)
(vendor)
Testing Organization (Penn State Harrisburg)
Total suspended solids
Micron
Verification Organization (NSF)
Water Quality Protection Center
Cubic yard
Year
                                         VI

-------
                                      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  goal  of the ETV program 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 testing (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.

NSF International (NSF), in cooperation  with the EPA,  operates the Water Quality  Protection
Center (WQPC). The WQPC evaluated the performance of the Terre Hill Concrete Products'
Terre Kleen™ 09 (Terre Kleen™), a stormwater treatment device designed to remove sediments
from  stormwater runoff.  Faculty and  students from Penn State Harrisburg's Environmental
Engineering Program were  the Testing Organization (TO)  and conducted the  field testing and
laboratory analysis.

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 Terre Kleen™ was a cooperative effort among the following participants:

•  U.S. Environmental Protection Agency
•  NSF International
•  Terre Hill Concrete Products (THCP)
•  Penn State Harrisburg (PSH) Environmental Engineering Program
•  City of Harrisburg

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 Management
Branch, Water Supply and Water Resources Division, National Risk Management Research
Laboratory (NRMRL), provides administrative, technical,  and quality assurance guidance and
oversight  on all ETV Water Quality Protection Center activities.  In addition, EPA provides
financial support for operation of the WQPC and partial support for the cost of testing for this
verification. EPA's responsibilities include:

•  Review and approval of the test plan;
•  Review and approval of the verification report;
•  Review and approval of the verification statement; and
•  Post verification report and statement on the EPA website.

The key EPA contact for this program is:

       Mr. Ray Frederick           ETV WQPC Proj ect Officer
       (732) 321-6627             email: frederick.ray@epa.gov

       U.S. EPA, NRMRL
       Urban Watershed Management Branch
       2890 Woodbridge Avenue (MS-104)
       Edison, New Jersey 08837-3679

1.2.2   Verification Organization

NSF is the verification organization (VO) administering the WQPC in  partnership with EPA.
NSF is a not-for-profit testing and certification  organization dedicated to public health, safety,
and protection of the environment. Founded in 1946 and located in Ann Arbor, Michigan, NSF
has been instrumental  in 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 and/or mark meet those standards.

NSF personnel provided technical oversight of the verification process, which include:

•  Review and comment on the test plan;
•  Review quality systems of all parties involved with the TO, and qualify the TO;
•  Oversee TO activities related to the technology evaluation and associated laboratory testing;
•  Conduct an on-site audit of test procedures;
•  Provide quality assurance/quality control (QA/QC) review and support for the TO;
•  Review the verification report and verification statement; and
•  Coordinate with EPA to approve the verification report and verification statement.

-------
                                        Program Manager
                                        email: stevenst@nsf.org

                                        Project Coordinator
                                        email: davison@nsf.org
Key contacts at NSF are:

       Mr. Thomas Stevens, P.E.
       (734) 769-5347

       Mr. Patrick Davison
       (734)913-5719

       NSF International
       789 North Dixboro Road
       Ann Arbor, Michigan 48105

1.2.3   Testing Organization

Penn  State  Harrisburg's (PSH)  Environmental Engineering Program was  the  TO, and was
responsible for ensuring that the  test location and conditions allowed the verification testing to
meet its  stated objectives; preparing the test plan; overseeing the testing; managing the data
generated by the testing; and preparing the verification  statement and  report.  TO  personnel
measured and  recorded data during the testing. TO employees also analyzed the  samples when
they were returned to the laboratory for analysis.  The TO's Project Manager provided project
oversight.

PSH had primary responsibility for all verification testing, including:

•   Coordinate all testing and observations of the Terre Kleen™ in accordance with the test plan;
•   Supervise the analytical work performed in support of the test plan;
•   Establish a communication network;
•   Schedule and coordinate the activities for the verification testing;
•   Manage data generated during the verification testing; and
•   Prepare the draft verification report and statement for the Terre Kleen™ ETV testing.
The key contact for the TO is:

       Dr. Shirley E. Clark, P.E.
       (717) 948-6127
                                        Assistant Professor of Environmental Engineering
                                        email: seclark@psu.edu
       Penn State Harrisburg
       Environmental Engineering Program
       School of Science, Engineering and Technology
       777 W. Harrisburg Pike
       Middletown, PA 17057
1.2.4

-------
       Vendor

The vendor is Terre Hill Silo Company, Inc. (T/D/B/A Terre Hill Concrete Products). Vendor
responsibilities include:

•  Provide, and possibly  install, the technology  and ancillary  equipment required for the
   verification testing;
•  Provide technical support during the installation and operation of the technology, including
   the designation of a staff person or representative that  will conduct at least one on-site
   inspection during monitoring to ensure the technology is functioning as intended;
•  Provide descriptive details about the capabilities and intended function of the technology;
•  Review and approve the test plan prior to the start of testing; and
•  Review and comment on the draft verification report and verification statement.

The key contact for Terre Hill Concrete Products is:

       Dale Groff                        Project Manager
       (717)445-3110                    e-mail: dgroff@terrehill.com

       Terre Hill Concrete Products
       485 Weaverland Valley Road
       Terre Hill, Pennsylvania  17581

1.2.5   Verification Testing Site

The Terre Kleen™ was installed at the  edge of the primary drainage area for the City of
Harrisburg Public Works Yard on 19th Street in
City of Harrisburg Public Works Department is:
Harrisburg Public Works Yard on 19th Street in Harrisburg, Pennsylvania. The key contact for
       Mr. James Close, Director
       City of Harrisburg Public Works Department
       1690 S.  19th Street
       Harrisburg, PA 17104

-------
                                      Chapter 2
                               Technology Description

The following technology description was supplied by the vendor and does not represent verified
information.

2.1    Technology Overview

The  Terre Kleen™ system manufactured by Terre Hill  Concrete  Products (THCP) that was
verified includes baffles, screens, and lamella plates in a self-contained unit. The design of the
unit  provides for underground installation as an in-line  treatment device at locations where
substantial stormwater solids loadings may be encountered. Appendix A includes design and
operations and maintenance (O&M) guidelines for the Terre Kleen™, and Appendix B includes
photographs of the test site.

Product Name:        Terre Kleen™ (Patent # US 6,676,832 B2)
Company Name:      Terre Hill Concrete Products
Authorized Contact Person and Title: Dale Groff, Project Manager

2.2    Technology Description

The  Terre Kleen™ device combines primary and  secondary chambers, baffles, screen, and
inclined sedimentation,  as well as  oil, litter and debris/sediment storage, into a self-contained
concrete structure. A  schematic diagram of the  Terre Kleen™ is shown in Figure 2-1.  The
product specifications  are included in Appendix B of the test plan, which is found in Appendix B
of this report.

The  principle of operation  is hydrodynamic. The primary benefit of the Terre Kleen™ is  its
ability to efficiently settle solids in the inclined cells (lamella plates) located in the secondary
chamber. The combination of treatment technologies into  a single device has been shown to be
more effective than the  use of a single technology for runoff treatment. The design of the unit
provides for underground installation as an in-line treatment device. It may be  applied at a
critical source area or  a  larger unit  may be installed in a storm sewer main to provide treatment
for larger flows. Installation can be performed using conventional construction techniques. Terre
Kleen™ units can  be designed to provide  specific removal efficiencies based on the size
characteristics of the suspended solids and flow rate of storm water to the device.

The Terre Kleen™ device addresses the concern of being space-effective, providing high particle
removal efficiency given the device's relatively small footprint. The ability to install the device
below grade allows for the use of the aboveground space, and makes it easier for the device to be
retrofitted into a pre-existing storm sewer system. In addition, if the flows exceed the sizing for
the secondary (lamella plate) chamber, all flow  is still treated in the primary settling chamber,
with the secondary chamber seeing the flows that it can effectively treat. The water is then
recombined prior to device discharge. This design allows for some treatment of all water that
enters the device.

-------
                                                                                        3
                                                                                        o
                                                                                        LU
                                                                                        CO
                                                                                      CD
                                                                                      m
                                                                                      O
                                                                                      tu
                                                                                      co
                                   S   Q-
                                                                                      o
                                                                                      LU
                                                                                      CO
Figure 2-1. Terre Kleen™ schematic and flow diagram.

-------
The device is designed to be easily accessible for maintenance purposes and has been designed
to provide storage space for sediment below the inclined cells and in the primary chamber. Re-
suspension of captured material below the inclined cells  is minimized because the stormwater
enters the inclined cells sideways instead of scouring the top of the sediment. The storage space
was designed to provide sufficient storage so that frequent clean out is not required. The device
has been designed with access covers to allow for easy access by vactor truck  hose  to all
chambers when maintenance is required.

The sizing chart for the Terre Kleen™ 09 Unit is shown in Table 2-1. The table values list the
anticipated particle size that will be removed based on the  flow rate entering the unit.  More
information on the sizing of the Terre Kleen™ is provided in Section 3.1.

Table 2-1.  Terre-Kleen™ Sizing Chart
Terre Kleen 09
Performance
Capacity
inCFS
0.3
0.6
1.0
1.3
1.9
3.4
4.3
5.6
6.4
8.9
11.4
12.7
15.0
17.8
21.0
Capacity
inGPM
137
268
428
599
855
1539
1930
2508
2850
3990
5130
5700
6726
7980
9405
Minimum
Particle
Size
Removal
in Micron
10
30
50
70
100
150
200
250
300
400
500
600
700
800
900
Grit
Chamber
Projected
Surface
Area SqFt
57
57
57
57
57
57
57
57
57
57
57
57
57
57
57
Sediment
storage in
CF
74
74
74
74
74
74
74
74
74
74
74
74
74
74
74
Approximate size
Length
6'-0"
6'-0"
6'-0"
6'-0"
6'-0"
6'-0"
6'-0"
6'-0"
6'-0"
6'-0"
6'-0"
6'-0"
6'-0"
6'-0"
6'-0"
Width
4'-6"
4'-6"
4'-6"
4'-6"
4'-6"
4'-6"
4'-6"
4'-6"
4'-6"
4'-6"
4'-6"
4'-6"
4'-6"
4'-6"
4'-6"
Miscellaneous Data
Oil
Storage
Capacity
(Gallons)
140
140
140
140
140
140
140
140
140
140
140
140
140
140
140
Grit
Chamber
Loading
Rate in
GPM per
SqFt
2.4
4.7
7.5
10.5
15.0
27.0
33.9
44.0
50.0
70.0
90.0
100.0
118.0
140.0
165.0
Primary
chamber
loading in
GPM per
SqFt
15
30
48
67
95
171
214
279
317
443
570
633
747
887
1045
               Maximum pipe size 024 inches. For higher flow rates check with our office.

As  with all sedimentation devices, the Terre Kleen™ can be sized for future applications to
remove the desired particle size  based on the Hjulstrom  diagram.  The Hjulstrom diagram
evaluates the ability of particles to be washed away (erosion), travel (transportation), or settle
(sedimentation) in fluid as a function of particle size and fluid velocity. The Hjulstrom diagram
provided by the vendor is shown in Figure 2-2.  This figure provides the empirical loading rate
for  the inclined plate settlers (gpm/ft2), and the particle size for which 100% removal is desired
that will settle at that loading rate. The amount of  settling area required is the flow rate (gpm)
from the inlet pipe divided by the loading rate (gpm/ft2). The number of settling cells to provide
the  required area is then directly proportional to the flow rate to the device.  Future sizing can
then be based on the flow loading calculations similar to those provided in Appendix E of the
Test Plan.  This  claim is  plotted as the  line between Sedimentation  and  Transport on the
Hjulstrom diagram

-------
                                        lEROSION
                                        Terre Kteen adjusted for
                                        Hydraulic Radius and
                                        Inclination
                                                                                     SEDIMENTATION
                                                                   U
                                                           s.  a
                                                           i  I
lit
              i 1
i!in...   _1  §  HiM;Ii.


                                                                                               .
                                                                                                     .
                5  I
1 ! [I tffl

!
                                                                                                                        191
                                                                                                                      •
                                       SH.T
Figure 2-2. Adjusted Hjulstrom diagram (provided by vendor).

-------
Verification of the sizing calculations using the Hjulstrom diagram was performed by the TO
through the analysis of periodic sampler bottles that were not included in the composite sampling
required as part of the Test Plan. The results of these analyses are shown in Figure 5-7 and are
discussed in Section 5.4.1.

Additional  equipment specifications,  test site  descriptions,  testing requirements, sampling
procedures, and analytical methods were detailed in the Environmental Technology Verification
Test Plan  for  Terre Hill Concrete  Products:  The  Terre Kleen™,  City of Harrisburg,
Pennsylvania, November 2004 (test plan). The test plan is included in Appendix B.

2.3    Applications

This Terre Kleen™ is designed to remove solids  from stormwater runoff by improving the
sedimentation performance compared with a traditional detention facility, and without requiring
chemical addition. The potential markets for this device include municipalities and developers
with stormwater runoff not meeting the standards for the receiving water to which it is being
discharged. These potential users may be required  through retrofit (municipalities) or through
installation during construction to treat their stormwater prior to discharge. A more efficient, in-
line treatment device for stormwater runoff would allow these owners  to meet the upcoming
requirements for treating their runoff without incurring the tremendous financial burden that
would result from the purchase of a more complex device or multiple devices.

2.4    Operation and Maintenance

The required O&M for this unit will consist of periodic removal of the sediment from the bottom
of the Terre Kleen™ device. This sediment removal interval will be based on the solids loading
to the device and the sedimentation performance of the  device. The storage areas in the device
have been designed to retain approximately 74 cubic feet (ft3) of sediment. For the test site, it
was  anticipated that sediment removal would not be  required during the testing interval.
However, this did not prove true; a description of the O&M activities is contained later in this
report.

During normal operation  in a post-construction environment, it is anticipated that cleaning would
be required once per year. A vactor truck,  similar to that used in cleaning a  sewer system and
stormwater catch basins,  would be used for the cleaning. The  device has openings built into the
top with removable covers for easy access to  the sediment  storage areas.  When the  device
required cleaning during  the testing period, a City of Harrisburg truck was used to remove the
sediment.

The Terre Kleen™ O&M guidelines are included in Appendix C of the test plan.

-------
2.5    Performance Claim

The vendor claims that the Terre Kleen™ 09 device will remove 100% of the 200-jim particles
and larger in the  runoff when the device is operating at the design storm flow (based on the
25-year storm). THCP predicts that at lower flows, removals of particles smaller than  200 jim
will also be achieved. The device is sized based on the adjusted Hjulstrom Diagram (Figure 2-2).
                                          10

-------
                                      Chapter 3
                                Test Site Description

The  test site for this device was the City of Harrisburg Public  Works  Yard in Harrisburg,
Pennsylvania.  The device was installed in the storm collection system adjacent to the swale
located in the  south corner of the property as shown in Figures 3-1 and 3-2. The drainage area
includes roofs, paving and unpaved areas. The watershed delineation is shown in Figure 3-1.

The  stormwater runoff from the test site was characterized prior to testing by four samples that
were collected at the proposed installation location during the  spring of 2004. The results are
summarized in Table 3-1. The sample location was roughly the same as the effluent sampling
location, as shown on Figure 3-1.

Table 3-1. Preliminary Suspended Solids Sampling at the Harrisburg Public Works Yard
                                 Turbidity    Total Solids       TDS          TSS
                                   (NTU)       (mg/L)       (mg/L)       (mg/L)
4/8/2004
4/1 1/2004
4/1 1/2004
4/12/2004
5:OOPM
8:30 AM
2:30 PM
1:00 PM
151
13
40.3
38.2
1,150
150
280
250
920
140
80
110
230
13
200
140
3.1    Sizing Methodology

The  calculation  of the  peak runoff flow rate using  a  25-year  design  storm is  included in
Appendix E of the Test Plan. In summary, the peak runoff was  calculated using the Rational
Method and was based  on site characteristics and on  the Intensity-Duration-Frequency (IDF)
curve provided by  the Pennsylvania Department of Transportation for the 25-year design storm.
The runoff was calculated to be approximately 3.49 cubic feet per second (cfs). The device was
then sized based  on this flow rate and was determined to be a Terre Kleen™ 09 Unit (Figure B-3
of the Test Plan), based on the removal of 200-jim particles for instantaneous peak flows of 3.2
cfs. It is anticipated that the device will provide overall control for the 150-jim particles based on
the normal flow rates similar to the 5- to 10-year storm, approximately 2.0 - 3.0 cfs.

3.2    Site Description

The drainage area  is part of the city's maintenance yard  occupied by the Bureau of Sanitation,
and includes runoff from buildings and paved and unpaved parking areas. The TO obtained
topographic  maps with 1 ft relief contours from the City of Harrisburg for the area (Figure 3-1).
Based on these maps, the watershed was estimated to be  approximately 1.27 acres and between
90 and 95% impervious. This delineation could not be  confirmed  from the aerial photograph in
Figure 3-2, since relief contours were not clear. In addition, the topographic map available from
the USGS uses  20-ft contours  and the  entire site has an  elevation between 381  and 390 ft.
Therefore, no additional delineation information was available.
                                           11

-------
During field visits to the  site, it was noted that at least the upper crushed dirt and stone area
where the vehicle wash occurs, plus part of the hillside, was all draining to the site. A review of
the site elevations using Google Earth (which was not available when the original plan was
developed) showed that the actual drainage area is much larger than originally estimated. Figure
3-2 shows the delineation  of the revised watershed. The estimated area of the revised watershed
from 20-foot contour maps of the site in  conjunction with 1-ft elevations from Google  Earth is
3.21 acres. Photographs of the drainage area are given in Figures 3-3, 3-4, and 3-5.

Roof drains are installed at the corners and at points along the outside of the building. Based on
the flow pattern from these roof drains,  approximately half of the roof drains  drained to the
device. In addition, Inlets 1-10 and 1-11 (Figure 3-1) drained to the device. The building is partly
office space and the  two wings are garage bays, which have floor drains to collect wash water.
Part of the area is heavily traveled (to and from the garage bays, but with trucks coming from
outside  the delineated watershed -  past  the  "front" of the building,  which  is  not in  the
watershed). In the watershed, there are two gas pumps for refilling city vehicles and a significant
number of parked vehicles in the area. Most of these vehicles are either waiting for maintenance
or are for  sale by the  city.

No road salt or soil  piles  are found in the watershed area. An incinerator, located just  off-site,
underwent significant renovation and was not in operation during most of the verification. The
solids found in the preliminary sampling  are  assumed to originate from atmospheric deposition
or from soils clinging to vehicles passing through or parked in the area. No unusual sources were
found in the watershed.

According to a personal conversation with James Close, Public Works Director, no maintenance
has been done on the storm sewer system  in several years. However, as part of the installation of
the Terre Kleen™, the city performed maintenance on the  piping system and the drainage swale
at the end of the pipe. The walkthrough of the site in June of 2004 indicated stagnant water in the
pipe from 1-11 to the outlet because of sediment buildup in the swale. While no  detailed
investigations  of inappropriate connections or infiltration  and inflow have been  performed, no
dry-weather flows were observed during the verification,  other than occasional  water entering
storm drains through vehicle washing or maintenance.

The city re-graded the swale prior to verification to ensure that the device and the pipe drained
properly. This area is on a  hill, so no flooding  is anticipated and no flooding was observed on site
during  Hurricane Ivan  (which  was the  fifth-worst  flooding  situation recorded  on  the
Susquehanna River in Harrisburg) in September 2004.

Prior  to testing, and in order to verify  the  installation locations and elevations, the vendor
surveyed  part of the site. It  was noted during the survey that the  drainage area was  slightly
smaller than estimated from the topographic information. This was due to the repaving of the site
by adding a layer of  asphalt over the existing  layers. Rutting of the asphalt had resulted in a few
areas where runoff from the pavement was actually directed away from the inlets and toward the
drainage swale downstream of the Terre Kleen™.
                                           12

-------
     V   /
                                   Drainage Area
                                   delineation
                                                                10,000 —
                                                                GASOLINE
                                                               STORA8E TANK
                                         Sampling Location
                                                               i  u , u ,KB_ ff  i   u  a
                                                                FNDWil fr-1 T- "
Figure 3-1.  Site topographic map showing the sampling location and location of the Terre
Kleen™.
                                          13

-------
Figure 3-2.  Aerial photograph  of Harrisburg Public Works  Yard showing outlet and
drainage area delineation.
Figure 3-3. Photograph of the drainage area.
                                        14

-------
Figure 3-4.   View from the sampling location  across the paved lot to the edge of the
watershed (see stop sign in photograph on right).
Figure 3-5. Outlet sampling location for 15-inch reinforced concrete storm drain pipe.
                                         15

-------
During the winter/spring of 2006, the City of Harrisburg Municipal Waste Incinerator came back
on line after a multi-year renovation. While the incinerator itself was not in the drainage area,
trucks carrying ash from the incinerator had to pass through the drainage area on their way to the
on-site ash landfill. Part of the road that they traversed is not paved. It was reinforced with gravel
which quickly sank into the dirt/mud, as illustrated in Figure 3-6. The results in Chapter 5 show
that this change in surface activity changed the character of the influent water flowing to the
Terre Kleen™ unit.
Figure 3-6.  Dirt and gravel road used by incinerator ash trucks.
3.3    Peak Flow Calculation

The rainfall amounts for the one-, two-, ten-, and twenty-five year storms for the drainage area
are presented in Table 3-2. Table 3-3 presents the intensities in inches per hour calculated for the
given  rainfall depths, as given  in  the PA DOT Intensity-Duration-Frequency  Curves for
Pennsylvania (these were read from  the PA DOT charts and were not calculated by the TO.
These data were utilized to generate the peak flows shown in Table  3-4. Table 3-5 presents the
peak flow calculated using the time  of  concentration for  the drainage  basin. The time of
concentration was calculated as described in Appendix E of the test plan and is based on the time
                                           16

-------
of concentration calculation methods described in the United States Department of Agriculture's
Natural Resource Conservation Services' (NRCS) TR-55 method "Urban Hydrology for Small
Watersheds." (NRCS, June 1986). The method of calculation for the peak flow was the Rational
Method (McCuen 2005), where the peak flow rate is equal to the rainfall intensity for the time of
concentration multiplied by the drainage area multiplied by a runoff coefficient (which reflects
the quantity of rainfall that becomes runoff).

Table 3-2.  Rainfall Depth (in.)

     Duration	1-yr	2-yr	10-yr	25-yr
5 min
30 min
Ihr
2hr
12 hr
PA DOT. Field Manual
0.31
0.79
1.0
1.3
2.0
0.35
0.93
1.3
1.5
2.5
of PA DOT Storm Intensity -Duration-Frequency
0.45
1.27
1.7
1.9
3.7
Charts. (1986).
0.51
1.42
1.9
2.3
4.5

Table 3-3. Rainfall Intensities (in./hr)
Duration
30 min
1 hr
2hr
12 hr
24 hr
PA DOT. Field Manual
1-yr
1.5
1.0
0.65
0.18
0.10
2-yr
1.8
1.2
0.77
0.22
0.12
of PA DOT Storm Intensity -Duration-Frequency
10-yr
2.5
1.7
1.1
0.32
0.19
Charts. (1986).
25-yr
2.8
2.0
1.3
0.37
0.23

Table 3-4. Peak Flow Calculations (cfs)
Duration
30 min
Ihr
2hr
12 hr
24 hr
1-yr
4.15
2.76
1.79
0.51
0.28
2-yr
4.97
3.36
2.12
0.61
0.33
10-yr
6.90
4.70
3.03
0.88
0.53
25-yr
7.73
5.31
3.59
1.02
0.63
                                          17

-------
Table 3-5.  Peak Flow Calculations (cfs) Using Time of Concentration

     Duration	1-yr	2-yr	10-yr	25-yr
      25min            4.42             5.13              7.72              8.98
The City of Harrisburg recommends that all storm drain systems for maintenance facilities such
as this one be designed to accommodate the 25-year storm, although a 10-year design storm is
acceptable. A 25-minute time of concentration was determined for the basin, generating a peak
runoff of 7.72 cfs for the 10-year storm event.  The Rational Method was used to  calculate the
peak flows for the device, since the drainage basin is between one and two acres and well within
the guidelines for the limits of the Rational Method (drainage area < 20 to 200 acres, depending
on the reference providing the guidance).  The 15-inch reinforced concrete drainage pipe was
originally sized to pass the 25-year  storm  without bypassing the  piping  system. During
installation of the Terre Kleen, the City of Harrisburg replaced the 15-inch reinforced concrete
pipe with an 18-inch SDR32 PVC pipe from the  Terre-Kleen  test unit to the endwall. The
replacement of the  15-inch  concrete pipe to the smoother  18-inch PVC  was part of the
installation agreement with the  City and was done to ensure that the effluent from the Terre
Kleen™  would drain completely. Backwater in the effluent pipe would  have invalidated  all
effluent sampler measurements.

It  was also noted during  analysis  of the  sampler data that there was  a  dry-weather flow
component entering the Terre Kleen™ on an uneven schedule. To ensure that this was not a
sampler error, this dry weather flow  was noticed  during most  visits to the Terre Kleen for
inspection and maintenance. When the dry-weather influent flow rose to more than a trickle, the
effluent pipe also had  dry-weather flow. Analyzing the sampler data indicates that this dry-
weather component also could be seen in the many storm event samples. The TO decided not to
remove this dry-weather flow from  the calculations and evaluations since  this  flow passed
through the Terre Kleen™.

Because of the nature of the site, trash and debris problems were noted in the influent sampler
side of the Terre Kleen™. These problems and their effects are described in Chapters 4 and 5. In
summary, the influence of debris on readings at the influent sampler caused a concern about the
reliability of the influent sampler for flow readings. Therefore, the effluent sampler was used for
all flow measurements,  with one exception described later.

A review of the hydrologic behavior of the site indicates that the time of concentration calculated
in Table 3-5 actually  may be too low for the site. In the more intense storms, a time-to-peak was
noted at approximately 5 min, indicating a flashy response of the site to the storm. Therefore,
while the numbers in this  section  were calculated assuming a Rational coefficient that reflects
90% to 95% imperviousness, this site behaved like it was 100% impervious.
                                           18

-------
3.4    Contaminant Sources and Site Maintenance

The main pollutant sources within the drainage basin are created by vehicular traffic, as well as
heavy equipment maintenance and the storage of garbage collection trucks on the site. Only part
of the site is paved with asphalt. Much of the rest is gravel embedded in dirt.  Trash and debris
accumulate on the surface and enter the stormwater system through the two inlets on site. These
inlets were sized to accommodate the large storm flows, and the storm sewer catch basins do not
have sumps.  There  are no other  stormwater best  management practices (BMPs) within the
drainage basin.

Minimal site maintenance occurred during the verification test period. The primary maintenance
was the deepening of a 3-in. channel to  funnel water into the last inlet (1-11) in the drainage
system prior to the Terre Kleen™. Additional maintenance consisted of one time delivery of
gravel for the prevention of dust and dirt mobilization on the unpaved areas.

3.5    Stormwater Conveyance System and Receiving Water

As previously discussed, the nearest receiving water is the  Susquehanna River, which is located
approximately one-third  of a mile west of the Public Works Yard and the Terre Kleen™
installation.  All water, either treated or bypass, flows via a drainage swale off the site  in a
southwesterly direction before ultimately flowing into the Susquehanna River.

3.6    Terre Kleen™ Installation

Terre Hill Concrete Products supplied the device for testing. The installation was performed by
THCP  and the City of Harrisburg who provided  the construction  equipment and operators
associated with excavation and  placement of the device.  Installation consisted of placing the
Terre Kleen™ into the existing storm sewer infrastructure.  PSH personnel were at the site
during installation to ensure that the device was installed correctly and to be sure principal
researchers understood the device.  Construction activities were completed in February 2005 and
samplers were installed in February 2005. A malfunctioning effluent sampler was replaced in
March 2005.  The installation and final setup are documented  in Figure 3-7 and Figure  3-8.
Figure 3-7 (a) shows the fifteen-in. reinforced concrete pipe on the upstream side that  was
connected to the devices primary chamber,  and the  18-in.  PVC pipe  that was connected to the
downstream side of the device.  Figure 3-7  (b) shows the Terre Kleen device being installed.
Figure 3-8 (a) shows the sample tubing and flow meter cables installed into the influent pipe, and
Figure 3-8 (b) shows the autosampling equipment.
                                           19

-------
                   (a)                                        (b)
Figure 3-7. Left: Terre Kleen™ installation site looking from upstream to downstream.
                   (a)                                        (b)
Figure 3-8. Sampling equipment installation arrangements.
                                         20

-------
                                     Chapter 4
                  Sampling Procedures and Analytical Methods

Descriptions  of the  sampling locations  and methods used  during  verification testing are
summarized in this section. The test plan presents the details on the approach used to verify the
Terre Kleen™. An overview of the key procedures used for this verification is presented below.

4.1    Sampling Locations

       Two locations in the test site  storm sewer system were selected as  sampling and
monitoring sites to determine the treatment capability of the Terre Kleen™.

4.1.1   Upstream Influent

This monitoring site was selected to monitor the  stormwater flow rates entering the  Terre
Kleen™ and collect samples of the influent stormwater. A flow/velocity/stage meter was located
in the influent pipe, upstream of the Terre Kleen™ and downstream of Inlet I-11, at a distance
where  maintenance could be performed and sufficiently downstream that mixing of  the inlet
water with the in-sewer stormwater would have occurred. Sampler suction tubing to an automatic
sampler and the velocity meter were located in the influent pipe as recommended by American
Sigma, the sampler manufacturer.

4.1.2   Downstream Effluent

This sampling and monitoring site was selected to characterize  the stormwater discharged from
the Terre Kleen™.  A velocity/stage meter and  sampler  suction  tubing,  connected to the
automated sampling equipment, were located in the pipe downstream from the Terre Kleen™.

4.1.3   Rain Gauge

Two rain gauges were located  adjacent to the samplers on  top of the  endwall for the effluent
(downstream) pipe leaving the Terre Kleen™. These gauges were used to monitor the  depth of
precipitation from storm events. They were also used to trigger the automatic samplers since a
small amount of dry  flow was found periodically in the influent  piping.  Triggering by a rain
event ensured that the samplers did not trigger until actual runoff had begun. The data were also
used to characterize the events to determine if they  met the requirements  for a qualified storm
event. Qualified storms were those whose rainfall depth measured at least 0.2 in.

4.2    Monitoring Equipment

The specific equipment used for monitoring flow, sampling water quality, and measuring rainfall
for the upstream and downstream monitoring points included:

•  Samplers: American  Sigma 900MAX automatic sampler with  a data transfer unit (DTU II)
   data logger;
                                          21

-------
•  Sample Containers: Twenty-four 500-mL polyethylene bottles designed to fit in the sampler
   housing;
•  Flow Monitors: American Sigma Area/Velocity Flow Monitors; and
•  Rain Gauge:  American Sigma Tipping Bucket Rain Gauge.

4.3    Constituents Analyzed

The list of constituents for which the stormwater samples were analyzed is shown in Table 4-1.

Table 4-1.  Constituent List for Water Quality Monitoring
              Parameter
               Method
Reporting     Detection
  Units         Limit
Method
Total suspended solids (TSS)
Suspended sediment concentration (SSC)
Particle Size Distribution
mg/L
mg/L
Counts/mL or
H3/mL
5
0.5
1
EPA 160.2
ASTMD3977-97
SM 2560
The ETV Verification Protocol for Stormwater Source Area Treatment Technologies indicates
that SSC ASTM Method D3977-97(C) (wet-sieving filtration) should be used for quantification
of particles larger than and smaller than 62 jim in  size.  For this verification, a more thorough
particle size distribution analyses with SM 2560 was utilized to provide a more thorough analysis
of particle size counts.

4.4    Sampling Schedule

The monitoring equipment was installed in February 2005. From March 2005 through June
2005, several trial events were monitored, and  the equipment tested and calibrated. Verification
testing began in June 2005, and  ended in May 2006. As  defined in the test plan, "qualified"
storm events met the following requirements:

•  The total rainfall depth for the event, measured at the site rain gauge, was 0.2 in. (5 mm) or
   greater.
•  Flow through the treatment device was successfully measured and recorded over the duration
   of the runoff period.
•  A flow-proportional composite sample was successfully collected  for both the influent and
   effluent over the duration of the runoff event.
•  Each composite sample collected was comprised of a minimum of five  aliquots, including at
   least two aliquots on the rising limb of the runoff hydrograph, at least one aliquot near the
   peak, and at least two aliquots on the falling limb.
•  There was a minimum of six hours between qualified sampling events.
                                          22

-------
4.5    Field Procedures for Sample Handling and Preservation

Water samples were collected with Sigma automatic samplers programmed to collect aliquots
during  each sample cycle. A peristaltic pump on the sampler pumped water from the sampling
location through Teflon™-lined  sample tubing to the pump head where water passed through
silicone tubing  and  into the sample collection bottles. After qualified  events, samples were
removed from the sampler, split  and capped by PSH personnel. Samples were analyzed  within
the holding times allowed by the  methods. All samples were analyzed at the PSH Environmental
Engineering Research Laboratory. Custody was maintained according to the laboratory's sample
handling procedures. To establish the necessary documentation to trace sample possession from
the time of collection, field collection was documented for each set of samples and recorded both
in the field book and on the computer system.
                                         23

-------
                                     Chapter 5
                        Monitoring Results and Discussion

Precipitation and stormwater flow records were evaluated to verify that the storm events met the
qualified event requirements. The qualified event  data  is summarized in this  chapter.  The
monitoring results related to contaminant reduction over the events are reported in two formats,
consistent with the protocol:

    1.  Efficiency ratio comparison, which evaluates the effectiveness of the system on an event
       mean concentration (EMC) basis.
    2.  Sum of loads (SOL) comparison, which evaluates the effectiveness of the system on a
       constituent mass (concentration times volume) basis.

The performance of the device will also be discussed in light of the specific maximum particle
size, since the performance of the  device is a function of a specific maximum particle size, and
not the removal of a specific percentage of the total suspended solids or of the suspended
sediment concentration load.

5.1    Storm Event Data

Table 5-1 summarizes the  storm  data for the qualified events. Detailed information on each
storm's runoff hydrograph and the rain depth distribution over the event period are included in
Appendix C.

The sample collection starting times for the influent and effluent samples, as well as the number
of sample aliquots collected, varied from event to event. The samplers were activated when the
rain gauges sensed a minimum of 0.08 in. of rain (each sampler had its own rain gauge). The
value of 0.08  in. of rain was selected based on visual observation of the site during rain events.
At  rain depths  smaller than 0.08  in., the runoff depth in the pipe was too low for adequate
sampling. This also  prevented the  samplers from being turned on when there was a very small
rain event or dry-weather flow entering the storm sewer system.

There was intermittent dry-weather flow (typically due to vehicle washing) entering the Terre
Kleen™. Occasionally, the dry-weather flow volume was sufficient to activate the  effluent
sampler. The effluent sampler was  operated from flow conditions for the first three storm
events, but the trigger for effluent sampling was changed to the rain gauge after the third event to
address the concerns of collecting sample aliquots from dry-weather flows.

5.1.1   Flow Data Evaluation

Table 5-2 summarizes the flow volumes and peak discharge rates for the influent and  effluent
monitoring  locations for each of the qualified events. As  described in the next paragraph, litter
and sediment  build-up was commonly observed on the velocity sensor in the influent pipe. This
appeared to impact the flow readings; however, the  influent auto sampler appeared to function
properly. The effluent sampler flow data were used  in all volume calculations except the storm
of April 3,  2006, where the effluent flow logger recorded no data.  For this  event, the  influent
                                          24

-------
  pipe had been cleaned approximately a day before the storm, so potential  interferences from
  sediment or litter were minimized, and the influent flow meter provided data to determine the
  storm volume.

  When  practical, sensors should be installed at a minimum distance of five times the maximum
  expected level upstream from an obstruction and ten times the expected level downstream from
  an obstruction. Obstacles were not an issue in the downstream pipe, but they  were a concern for
  two reasons in the upstream pipe. The first concern was the accumulation of debris in the pipe
  between inlet I-11  and the Terre Kleen™.  It appeared that the pipe joints were not smooth,
  resulting in a location between pipes where small  rock-based  dams could form in the pipe. In
  addition, the site received substantially  more litter over the course of the testing  than was
  expected. Some of  the litter,  such as plastic grocery bags, tended to snag on the influent flow
  meter.  This litter was entrapped by the Terre Kleen™ and did not impact the  performance of the
  effluent sampler.

  Table  5-1. Summary of Events Monitored for Verification Testing
Event No.
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
Start Date
6/29/05
7/7/05
8/16/05
8/27/05
9/16/05
10/13/05
10/21/05
11/16/05
11/21/05
1 1/29/05
12/25/05
1/2/06
1/11/06
4/3/06
5/13/06
Start Time
12:00
18:40
09:35
19:05
18:55
05:20
22:45
10:30
23:20
04:55
11:50
10:45
12:50
14:40
16:20
End
Date
6/29/05
7/8/05
8/17/05
8/28/05
9/17/05
10/14/05
10/22/05
11/17/05
1 1/22/05
11/30/05
12/25/05
1/3/06
1/11/06
4/3/06
5/15/06
End
Time
14:00
09:40
0:35
09:05
0:35
03:15
23:00
01:10
08:35
0:00
20:30
12:25
23:55
22:30
22:30
Rainfall
Amount
(in.)1
0.31
1.68
0.43
0.68
1.22
0.63
1.17
0.20
0.52
1.04
0.45
0.99
0.42
0.75
0.71
Rainfall
Duration
(hr:min)
2:00
15:00
11:10
14:00
5:40
21:55
24:15
14:40
9:45
19:05
8:40
25:40
11:05
7:50
54:10
1. Rainfall depths recorded by the rain gauge corresponding to the effluent sampler.
                                             25

-------
Table 5-2. Peak Discharge Rate and Runoff Volume Summary
                                              Peak Discharge Runoff Volume
                Event No.	Start Date	Rate (cfs)1	(ft3)1
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
6/29/05
7/7/05
8/16/05
8/27/05
9/16/05
10/13/05
10/21/05
11/16/05
1 1/22/05
1 1/29/05
12/25/05
1/2/06
1/11/06
4/3/06
5/13/06
0.83
0.82
0.029
0.76
2.0
0.50
0.80
0.013
0.37
1.2
0.26
0.14
0.20
0.36
0.089
750
7,900
210
1,800
4,900
960
3,800
110
1,300
6,500
580
940
480
1,500
660
              1.  Peak discharge rate and runoff volume reported from effluent data, with the
                 exception of Event 14, where the effluent sampler functioned properly but
                 did not record flow data..

The flow monitors measured the depth and velocity of water in the pipe and calculated the flow
rate at five-minute intervals using Manning's equation with an  assumption of normal depth.
When the data were reported in the September  2006 version of  this  report, the flow volumes
calculated by the sampler software were substantially higher than what would be expected for
any of the rainfall amounts during the testing.  A review of the flow records showed that the flow
meter was calculating flow with Manning's equation using the  water depth and  pipe slope,
instead of water depth and velocity, which is generally perceived as being more accurate.  The
pipe slope method resulted in a velocity data which were substantially higher than the recorded
velocity measurements,  and resulted in higher calculated flow rates.  Subsequently, the  flow
volumes rates were re-calculated for every qualified event using the recorded depth and velocity
data using the flow meter software.

5.2    Monitoring Results:  Performance Parameters

5. 2. 1   Concentration Efficiency Ratio

The concentration efficiency ratio reflects the treatment capability of the device using the event
mean concentration (EMC)  data obtained for each runoff event.  The concentration efficiency
ratios are calculated by:
                      Efficiency ratio = 100 x (HEMCegiuent/EMCinfiuent])               (5-1)
                                           26

-------
The influent and effluent sample concentrations and calculated efficiency ratios are summarized
by the analytical parameter (sediment) categories of TSS and SSC.

The influent and effluent sample concentrations and calculated efficiency ratios for sediments are
summarized in  Table 5-3. The TSS influent concentrations ranged from 58 to 6,870 mg/L, the
effluent concentrations ranged from 35 to 980 mg/L, and the efficiency ratio ranged from -88%
to 86%. The SSC influent concentrations ranged 110 to 430 mg/L, the effluent concentrations
ranged from 55  to 200 mg/L, and the efficiency ratio ranged from -11% to 87%.

Table 5-3. Monitoring Results and Efficiency Ratios for Sediment Parameters
Event
No.
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
Date
6/29/05
7/07/05
8/16/05
8/27/05
9/16/05
10/13/05
10/21/05
11/16/05
1 1/22/05
1 1/29/05
12/25/05
1/02/06
1/11/06
4/03/06
5/13/06
Influent
(mg/L)
540
190
69
58
870
140
210
250
340
1,100
850
310
890
840
6,900
TSS
Effluent
(mg/L)
380
190
130
35
650
140
150
120
220
590
240
130
840
640
980
Reduction
(%)
30
0
-88
40
25
0
29
52
35
46
72
58
5.6
24
86
Influent
(mg/L)
500
220
140
75
3,800
140
200
220
280
1,600
680
520
810
780
7,000
SSC
Effluent
(mg/L)
360
200
35
38
500
140
170
150
300
1,000
110
130
900
540
1,500
Reduction
(%)
28
9.1
75
49
87
0
15
32
-7.1
38
84
75
-11
31
79
As described in Section 3.2, site conditions changed during the fall of 2005. The non-functioning
incinerator was restarted for test fire burns and construction began to open up the hillside at the
edge of the Terre Kleen's drainage  area  for ash disposal. The  construction had limited, but
periodic, effects on the influent solids  concentration.  The effects of substantially  increasing
influent solids concentration were most notable when the incinerator began full operation of one
burner at the end of November 2005 (from average TSS/SSC less than 350  mg/L influent to
greater than 700 mg/L influent TSS/SSC).

In general, the results  show a similarity between influent TSS and SSC concentrations. Both the
TSS and SSC analytical  parameters  measure sediment concentrations in water. However, the
TSS analytical procedure requires  the analyst to draw an aliquot from the  sample container,
while the SSC procedure  requires use of the entire contents of the sample container. If a sample
contains a high concentration  of settleable (large particle size or high density) solids, acquiring a
                                           27

-------
representative aliquot from the sample container for the TSS analysis may be very difficult. A
disproportionate amount of the settled solids may be left in the container, resulting in a reported
TSS concentration lower than the SSC concentration. Since this phenomenon was not observed
during this study, it appears that the sediment loading consisted primarily of sediments with
small  particle size.  This  observation correlates  with  the particle size distribution data
summarized in Section 5.3.

5.2.2  Sum of Loads

The sum of loads (SOL) is the sum of the percent load reduction efficiencies for all of the events,
and provides a measure of the overall performance efficiency of the Terre Kleen™. The load
reduction efficiency is calculated using the following equation:

                    % Load Reduction Efficiency = 100 x (1 - (A / B))                (5-2)

       where:
       A = Sum of Effluent Load = (Effluent EMCi)(FlowVolumei) +
           (Effluent EMC2)(Flow Volume2) + (Effluent EMCn)(Flow Volumen)

       B = Sum of Influent Load =  (Influent EMCi)(Flow Volumei) +
           (Effluent EMC2)(Flow Volume2) + (Effluent EMCn)( Flow Volumen)

       n = Number of qualified sampling events

As noted in Section 5.1.1, the effluent monitoring location provided the most representative flow
data, so the SOL calculation was made using the effluent volumes for both the influent and
effluent data.

Table 5-4 summarizes results for the SOL calculations for TSS  and SSC. The  SOL  analyses
indicate a 44% reduction for TSS and a 63% reduction for SSC.
                                          28

-------
Table 5-4.  Sediment Sum of Loads Results

Event No.
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
Sum of the

Date
6/29/05
7/7/05
8/16/05
8/27/05
9/16/05
10/13/05
10/21/05
11/16/05
1 1/22/05
1 1/29/05
12/25/05
1/2/06
1/11/06
4/3/06
5/13/06
Loads
Runoff
Volume
(ft3)
750
7,900
210
1,800
4,900
960
3,800
110
1,300
6,500
580
940
480
1,500
660

Removal Efficiency (%)
TSS
Influent
(Ib)
25
93
0.9
6.6
270
8.3
50
1.6
28
450
31
18
27
79
280
1,400
44

Effluent
(Ib)
18
93
1.7
4.0
200
8.3
36
0.8
18
240
8.6
7.7
25
61
40
760


Influent
(Ib)
23
98
1.9
8.5
1,200
8.3
47
1.4
23
650
24
31
24
74
290
2,500

SSC
Effluent
(Ib)
17
98
0.5
4.3
150
8.3
40
1.0
24
410
3.9
7.7
27
51
62
910
63
5.3    Particle Size Distribution

Particle size distribution analysis was conducted by:

•  Sieving the samples to create a second  series  of TSS and  SSC  samples that contained
   particles smaller than 250 |j,m; and,
•  Analyzing the samples using a Coulter Multisizer 3, an instrument (described in the test plan)
   that measures particle concentration as counts according to particle size.

5.3.1   Particle Size Distribution with Sieve Data

With the apertures available in the PSH laboratory,  each  sample could be analyzed over a
particle size range of 0.8 to 240 |j,m. In addition, the fraction of the samples above 250 jam could
be quantified.  The results of the 250-|j,m sieve split are  summarized in  Table  5-5, which
demonstrate that the SSC analysis was a better measure to quantify the larger particles than TSS.
Recalculating the SOL for both TSS and SSC for particles larger than 250 jam shows an 85%
reduction for TSS, and a 98% reduction for SSC. The  lower removal for TSS has two  possible
origins: taking a sample aliquot from a bottle instead of analyzing the whole sample; or the
potential for scour/resuspension of previously captured solids.
                                          29

-------
Table 5-5.  Particle Size Distribution Analysis Results (Particle Sizes > 250
TSS
Event
No
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
Influent
(%)
NA
7
9
11
39
0
15
12
16
28
33
0
0
7
24
Effluent
(%)
NA
1
76
17
16
0
0
3
15
4
0
7
1
4
0
SSC
Influent
(%)
NA
NA
54
15
88
6
11
7
32
20
40
29
0
11
7
Effluent
(%)
NA
NA
0
2
0
4
0
12
42
1
0
15
16
0
11
Sum of the Loads
Removal
Efficiency
(percent)


TSS
Influent
(Ib)
NA
6.5
0.1
0.7
110
0
7.5
0.2
4.4
130
10
0
0
5.6
68
330

Effluent
(Ib)
NA
0.9
1.3
0.7
32
0
0
0
2.7
9.6
0
0.5
0.3
2.4
0
51
85
SSC
Influent
(Ib)
NA
NA
1.0
1.3
1,000
0.5
5.2
0.1
7.3
130
9.8
8.9
0
8.1
20
1,200

Effluent
(Ib)
NA
NA
0
0.1
0
0.3
0
0.1
10
4.1
0
1.1
4.3
0
6.8
27
98
   NA- Not available.

The vendor claimed that the device would remove 100% of the particles greater than 200 jim.
Assuming  that the influent is  well mixed  (in the water column),  sedimentation theory for
flocculating particles indicates that  some, but smaller than total, reduction in particles smaller
than the  cutoff size is  expected. Complete sedimentation  assumes  that the particles do not
interact with each other during sedimentation. These results indicate that complete sedimentation
of particles greater than 250 jim does not occur, likely due to non-ideal behavior of the particles
(such as by the creation of slower-settling floes with  the oils in the runoff) or due to very light
particles in the runoff. Based on the  samples' behavior in the lab and the amount of oil collected
in the primary chamber of the Terre  Kleen, it is believed that the creation of floes and emulsions
created non-ideal settling conditions  in the treatment device.

Using the data from Tables 5-4  and  5-5, it is also possible to calculate the percentage reduction
in loads of particles smaller than 250 jim. Table 5-6  summarizes the results of this calculation.
For particles  smaller than 250 jam, the TSS  evaluation shows a 35%  reduction, while the SSC
evaluation shows a 32% reduction.
                                           30

-------
Table 5-6.  Particle Size Distribution Analysis Results (Particle Sizes Smaller than 250

                                        TSS              SSC
                                  Influent  Effluent  Influent  Effluent
                     Event No.1      (Ib)      (Ib)      (Ib)      (Ib)
3
4
5
6
7
8
9
10
11
12
13
14
15
Sum of the
Loads
Removal
Efficiency (%)
0.8
5.8
160
8.3
42
1.4
23
320
20
18
27
74
220
920
35
0.4
3.3
170
8.3
36
0.8
15
230
8.6
7.1
25
58
40
600

0.9
7.2
140
7.8
42
1.3
16
520
15
22
24
66
270
1,100
32
0.5
4.2
150
8
40
0.9
14
400
3.9
6.5
23
51
55
770

                   1. Data were not available for Events 1 or 2.

5.3.2   Particle Size Distribution with Coulter Counter Data

While the analysis discussing the behavior of particles > 250 |j,m is most relevant to the vendor's
performance claim, samples were  also analyzed using the Coulter  Counter for particle sized
between 0.8 and 240 |j,m. Particle size distribution analyses were completed on individual storm
events.

Coulter Counter Analysis - Single Storm Event

For the purposes of this evaluation, the storm of November 16, 2005  was arbitrarily selected for
evaluation purposes  to demonstrate  how the  Coulter  Counter analyses were completed.  A
comparison between the influent and effluent samples for the storm  is shown in Figure  5-1 for
analysis of the number of particles of a particular diameter, and in Figure 5-2 for analyzing the
volume of particles  of a particular  diameter. The volume graph is most  comparable to the
traditional sieve analysis  (because of the relationship to mass through density) and it is the one
that will be discussed.

-------
                                 Storm of November 16, 2005
                  100
                                                10
                                     Particle Diameter (\*m)
                                                             100
Figure 5-1.  Coulter analysis comparison by  count of the Terre Kleen™ influent  and
             effluent total particle count for Event 8.
                                 Storm of November 16. 2005
                                     Particle Diameter (|,irn)


Figure 5-2.  Coulter analysis comparison by volume of the Terre  Kleen™ influent and
             effluent particle volume for Event 8.
                                          32

-------
Figure 5-1 shows that when  analyzing  for a total  particle  count, the dso (the particle size
corresponding to the 50th percentile of the cumulative volume of particles in the sample) of the
influent and effluent samples are essentially the same. The dso by particle count is approximately
1 |j,m, further emphasizing the small size of the particles in the stormwater runoff at the test site.
However, when analyzing the results by volume (Figure 5-2), which is most directly relatable to
traditional sieve analysis, the  dso shifts from approximately  20 to 25  |j,m for the influent to
approximately 4.2 |j,m for the  effluent, indicating a substantial removal of the larger particles.
This is confirmed on the number count graph (Figure 5-1) by the shift to the "left" of the effluent
at the upper end of the particle size range.

Incorporating the two particle  size analyses with the data for material > 250 |j,m results in the
graph shown in Figure 5-3. The graph shows that the dso of the influent is approximately 30 |j,m
and the effluent approximately  5 |j,m. This shift in the d50 to the smaller size ranges indicates that
removal  of both larger  and  smaller particles is occurring,  as  would  be expected  of any
sedimentation  device as long as complete  mixing of the influent occurs. Above a certain particle
size, 100% removal is anticipated, and for the smaller particles, partial removal is attained.

                              Storm of November 16, 2005
                                             10               100
                                   Particle Diameter (urn)
1000
Figure  5-3.  Complete particle size distribution for influent and effluent samples from
Event 8.
                                           33

-------
Coulter Counter Analysis - All Qualified Storm Events

The November 16, 2005 storm was arbitrarily selected as a demonstration of analysis for a single
storm event. This type of analysis was repeated for the entire  sample set for the composite
samples.  Where  duplicate composites were  available (for  two storms, there was insufficient
volume to  create replicate  samples),  each  composite was included in the calculations. The
particle size distributions for the influent and effluent samples for all sampled storm events were
calculated and adjusted for the mass above 250 |j,m to create a complete mass distribution. The
mean particle size for five points on  the sieve analysis curve (dio, d25,  dso,  d?s,  dgo) were
calculated and graphed, and error bars were  created assuming that the size of the error bar was
one standard deviation (shown only on the positive side when the error bar would exceed the
graph width).  The results of the analyses are  shown in Figure 5-4.
               Terre Kleen Testing - Harrisburg  Public Works Yard
      100
                                                                                -  d50
                                 10                     100
                                 Median Particle Size (j.irn)
1000
Figure 5-4. Particle size distribution for influent and effluent samples from all sampled
storm events using mean dio, dis, dso, dvs, and
Figure 5-4 shows there is a definite shift in the particle size distribution between the influent and
effluent,  even for  composite samples where some of the instantaneous impact of the Terre
Kleen™ may be muted due to sample compositing.. The error bars highlight the high degree of
variability in the composition of the influent and effluent samples from this site.  The dso of the
                                          34

-------
site,  calculated using the mean, shifted from approximately 80 |j,m in the influent to just over
10 |j,m in the effluent. Because of this variability in the particle size distributions between
samples, the data was reanalyzed using the median particle sizes to reduce the effect of the very
large and very  small  values  on the  data analysis.   The results of the median analyses are
summarized in  Figure 5-5, and show that the site's median particle diameter shifted from
approximately 10.5 |j,m in the influent to approximately 6.6 |j,m in the effluent.
               Terre Kleen Testing - Harrisburg  Public Works Yard
      100
       40
       20
                                  Approximately 98% of the
                                  particles removed were
                                  smaller than 200 |jm.
                                                                         Influent
                                                                         Effluent
                                 10                     100
                                 Median Particle Size (j.irn)
1000
Figure 5-5. Particle size distribution for influent and effluent samples using median dio,
 is, dso, d?5, and
The  vendor's performance  claim  stated that the Terre Kleen™ would remove 100% of the
particles 200 |j,m and larger when the device was operated at no greater than the design flow. A
review of Figure 5-5 shows that the effluent quality, as measured in these composite samples, did
not meet  this performance  claim. The composite samples  showed  that the Terre Kleen™
removed approximately 95% to 98% of the particles 200 |im and larger. This data, however, has
to be combined with the hydrologic data for the site.  The Terre Kleen™ does not contain an inlet
flow-control structure where the device only treats the flow up to a certain rate and then bypasses
everything else. The entire flow entering the device passes through it (bypass over the plates was
monitored after the primary chamber and at no time during the testing were the plates bypassed).
                                          35

-------
For many of the storms, the device treated instantaneous flow rates greater than the design flow
for between ten minutes and two hours. During those times, the higher hydraulic flow rate would
create a condition where it would not be expected that the device would remove the 200 |j,m or
larger particles. Therefore, the composite samples likely contain these larger particles.

5.3.3   Particle Size Distribution and Hjulstrom Diagram Evaluation

Section 2.2 of the verification report indicated that the vendor uses the Hjustrom diagram as the
basis for their sediment removal performance claims.   The Terre Kleen™ performance was
evaluated against the  Hjulstrom  diagram  for  events where individual  grab  sample aliquots
remained  after  samples  were composited.    This  evaluation went beyond the  test plan
requirements,  but helped to evaluate the Terre Kleen™  performance  and the relevance of the
Hjulstrom diagram to stormwater treatment and the vendor's performance claim.

In order to conduct the Hjulstrom diagram evaluation, the particle size and velocity data for the
grab samples  had to be gathered.   Particle size analysis was performed on grab samples for
events  where  sufficient sample was leftover after compositing on a per-bottle basis.  The VO
determined that the 95th percentile particles for the Terre Kleen™ field results should be used for
plotting on  the Hjulstrom diagram because the 95th percentile particle size more accurately
represents the upper end  of the particle size  distribution in the water,  and that only 3 to 5
particles represent the remaining mass between  the 95th and 100th percentile.

The instantaneous horizontal water velocity through the laminar plates in the Terre Kleen™ was
calculated based on the effluent flow hydrograph using the following equation (AWWA, 1990):


                                        v = -Q—                                    (5-3)
                                            Nwb

                    where:

                    v = horizontal water velocity (or hydraulic loading rate)
                    Q =  flow rate (calculated from independent level and velocity
                    measurements in the effluent pipe)
                    w =  horizontal width between laminar plates
                    b = length of one plate in the direction perpendicular to flow
                    N =  number of sedimentation cells

The 95th percentile largest  sediment particle and horizontal water velocity  data were then plotted
on the  Hjulstrom diagram  provided by the vendor.  The revised Hjulstrom diagram is shown in
Figure 5-6.
                                           36

-------

                           •Ł
               81 326?  mCO.	*"•'
                              MMI'I •!>•*< M «>* itWHv^d ftff tt Ih, 4»r .1 >««••*•«
~f^p~niff~ jg 	
BE ...•
•,.
MB nix
L .;
^17; 1 JB
M:P?I' :6*
BJ! !>««
•. i-
sea 1313*
z^i", :«K
1*73 Ofit
J5 	 ftjftM
r*fln • ' .
59W JdDiI'J

tofi-.

UQG9 *t'
"J '
\ &7 inn
:'. • .
; ';_ ;;
t BOW *
l3"*' 1
' f
. •-
i ' i
•'•
026&3 ^
•.:. •
- 4
•
'• '
own at
fiTff n :
•


P*nlC)*Stt«D*«rlp«/
;^
x














! I
i S
MM








J>
^

^/























^^


/
















IJ








^


/










•i














Jl


j^
?










EDIf








0 *3 4 j
1






'' >

^^

















^
^














IBITATION |








,» 1
:








S




*-•
X'





















!
-
i




+*"
^>

























,••
0*

























*~~
^~

























,-±±
*****






















n J i I !
i
1
iMML STW*$



•J »
F— ^-






















I
I
a



— ^
__-






















!
i
.



^^
—






















i
-



— •






















•
r



»^-^— — >






















if


BOCKS
                                                 -th
Figure 5-6.  Hjulstrom diagram plotting the 95   percentile particle size remaining in solution versus the  horizontal water

velocity through the plates.
                                                               37

-------
Figure 5-6 shows approximately half of the data points in the sedimentation zone, and the other
half in the transportation zone.  If the  Hjulstrom diagram  was precisely predicting the  site
conditions, the data points would lie in the transportation zone, just above the line adjusted for
hydraulic radius  and inclination.   Several reasons could be  used to explain the variation,
including, but not limited to:

    •   Water temperature and its impact on water viscosity;
    •   Non-ideal settling conditions, including turbidity associated with deeper flow conditions
       and sediment cohesion caused by clay particles and hydrocarbons;
    •   The presence of irreducibly fine silt particles in the water;
    •   Variations in particle densities;
    •   Differences  in flow depth between the Terre Kleen™ and the flume  on which  the
       Hjulstrom Diagram is based; or
    •   Sampling inaccuracies.

5.4    Retained Solids Analysis

During January 2006, after an observation noting substantial  negative removals were occurring
in the Terre Kleen, the unit was inspected and sediment depths measured. It was determined that
the unit needed to be cleaned. The City of Harrisburg agreed to  provide the sewer vacuum truck
to perform the needed cleaning. Prior to the cleaning, samples were collected in several locations
in the device. These samples were composited and shipped to  an outside laboratory for a particle
size distribution and chemical (Toxicity Characteristic Leachate Procedure [TCLP]) analysis.

5.4.1   Particle Size Distribution of Retained Solids

A particle size distribution analysis of the sediment retained in the Terre Kleen™ was performed
on a composite sample of the solids. The results of the analysis for particle size distribution are
shown in Figure 5-7.

As with all sedimentation devices, for any given flow rate, the device should have a particle size
for which 100% removal  will  occur.  This  does not  mean that particles smaller than that
100%-removal size will not occur, just that they will not be completely removed.  Figure  5-7
shows that 80% of the material captured in the sediment storage areas was smaller than the 200-
|j,m particle for which 100% removal was  claimed. The analysis also indicates that the  Terre
Kleen™ is capable of removing and retaining particles smaller than 200 |j,m.
                                           38

-------
   OJ
   N
   CO
     c
   03
   —•
   i_
   JSS
   "co
   E
   c/)
      100
       80
60
40
       20
                     Sieve Analysis on Captured Material
                     Terre Kleen Sediment Storage Areas
                                              \
80% of retained
solids were smaller
than 200 [im.
                           10               100
                                     Particle Size (jim)
                                                     1000
                  10000
Figure 5-7.  Particle size distribution for material captured in the sediment storages areas
of the Terre Kleen™.

5.4.2   TCLP A nafysis of Retained Solids

A composite sample  of the retained solids  was  also was analyzed for metals content in
accordance with the guidance for determining if the collected material is a hazardous waste. The
test selected for this analysis was the TCLP, a  test designed to simulate the behavior of a waste
material in contact with acids and acid rain leachate  in a landfill. The results are reported in
Table 5-7. As expected because of the high organic  content of the solids at the site and the
resulting high sorption affinity of the metals for the solids, the disposal solids were not hazardous
in accordance with the hazardous waste regulations. It is important to note that these results are
site-specific and are dependent on the metals found as  sources on the site (with the exception of
mercury, where most of the mercury in runoff  is from  airborne deposition). It is anticipated that
these results, in general, would be seen at  other installation locations for the Terre  Kleen™.
However, when installed at a site with known  specific problems of dissolved or colloidal-sized
metals  in the runoff or of large metal pieces in the runoff, the captured solids should  be tested
prior to disposal to confirm the appropriateness of municipal landfill disposal.
                                          39

-------
Table 5-7. Results for Cleanout Solids

           n      ...            T^T n n   m    n\      Regulatory Hazardous
           Parameter           TCLP Result (mg/L)       __,  x  _ ;,   .x ,   _,
    	     v  6  '	Waste Limit (mg/L)
     Arsenic                           <0.01                     5.0
     Barium                           <0.01                     100
     Cadmium                         <0.01                     1.0
     Chromium                         0.10                      5.0
     Copper                            1.12                      NA
     Lead                              0.69                      5.0
     Mercury                          <0.01                     0.2
     Nickel                             0.27                      NA
     Selenium	<0.01	LO	

    NA: Not applicable.
                                        40

-------
                                     Chapter 6
                           QA/QC Results and Summary

The Quality Assurance Project Plan (QAPP) in the test plan identified critical measurements and
established several  QA/QC  objectives. The verification  test procedures  and data collection
followed the QAPP. QA/QC  summary results are reported in this section, and the full laboratory
QA/QC results and supporting documents are presented in Appendix D.

6.1    Laboratory/Analytical Data QA/QC

6.1.1   Bias (Field Blanks)

Field blanks were collected  at both the inlet and outlet samplers to evaluate the potential for
sample  contamination through the  automatic sampler,  sample collection bottles, splitters,  and
filtering devices. The field blanks were analyzed for TSS and SSC only. All samples were below
the method detection limit of 5 mg/L, indicating that the  samplers were capable of pulling up
clean samples. Because of the nature of the influent at the maintenance yard  site with the oils and
greases,  the  sampler tubing was  replaced and the  sampler inlet was cleaned periodically
throughout the project.

6.1.2   Replicates (Precision)

Precision measurements were performed by the collection and analysis of duplicate samples. The
relative percent difference (RPD) recorded from the sample analyses was calculated to evaluate
precision. RPD is calculated using the following formula:

                                                                                 (6-1)
                              %RPD = \    -    | x 100%
                                            A!

       where:
       AI =  Concentration of compound in sample
       A2 =  Concentration of compound in duplicate
       A =  Mean value of AI and X2

The  RPD data show an acceptable level of field precision,  with a few parameters outside
generally  accepted  limits.  In  most circumstances  where the  RPD  values  are  high,  the
concentrations were near or below method detection limits.
                                          41

-------
Field precision: To address the concern of the ability of the sampler to  provide  repeatable
samples, a sampler calibration procedure was performed prior to installing the samplers in the
field. A Sil-Co-Sil  106-250 mixture of 200 mg/L was created and  15  replicate samples were
collected by the sampler and by hand, grabbing a sample immediately after the sampler collected
an aliquot. The results of this sampler calibration are shown in Table 6-1. The results show that
the sampler has a higher variability associated with it, but one that is in the acceptable range of
error, as measured by the coefficient of variation (COV) (equal to the standard deviation divided
by the mean and which provides a measure of the variability relative to the sample average). Part
of these differences also may be attributed to the hand mixing of the  solution between sampling
intervals and to the potentially slightly different sampling heights in the water column.


Table 6-1. Sampler Calibration for TSS using Sil-Co-Sil Mixture


                                                     TSS (mg/L)
      Sample Number	Hand	Sampler
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
Average
Std. Dev.
COV
170
200
190
200
160
210
210
210
200
200
180
210
190
230
200
197
16.9
0.09
200
180
150
230
190
210
180
160
220
200
360
150
200
180
170
199
48.8
0.25
Field duplicates were collected to  monitor the overall  precision of the sample collection
procedures,  including  sample splitting. Duplicate inlet  samples were  collected  during two
different storm events to evaluate precision in the sampling process and analysis. The duplicate
samples were  processed,  delivered to the laboratory, and analyzed in the same manner as the
regular samples.  Summaries of the field duplicate data are presented in Table 6-2.

In addition,  for periodic storms, not all sample bottles collected were used in the compositing.
These individual bottles were periodically analyzed for the same constituents  as the composite
samples. The results of these analyses are shown in Table 6-3 as a comparison between the field
composite sample and the average of the three per-bottle  analyses.  This is a second method of
                                           42

-------
verifying that quality control was maintained.  This method was performed because many of the
qualifying events had most or all of the  collected bottles from the sampler used to make the
composite.  As can be seen, the RPDs are generally  in accordance with the desired  replication
between the field duplicates. Differences are only seen for the storms where the concentration in
the individual bottles varied greatly over the storm. This variance across an individual storm was
not unexpected  since the variable nature of rainfall and intensity on the site will affect the
TSS/SSC concentration of the runoff.

Table 6-2.  Field Duplicate Sample Relative Percent Difference Data Summary
Analyte
TSS
Units
mg/L
Repl
1,322
Event 1
Rep 2
1,316
RPD
0.45
Repl
564
Event 2
Rep 2
768
RPD
31
Table 6-3.  Comparison of Composite Concentration with Per-Bottle Average

„    , „ ,                   Composite SSC      Average SSC of Per Bottle
Event Date                      /   /T \               c    i  /•   /T \
                               (mg/L)	Samples (mg/L)
12/25/2005 - Influent
12/25/2005 - Effluent
1 1/29/2005 - Influent
1 1/29/2005 - Effluent
1 1/22/2005 - Influent
1 1/22/2005 - Effluent
11/16/2005 -Influent
11/16/2005 -Effluent
10/22/2005 - Influent
10/22/2005 - Effluent
10/13/2005 - Influent
10/13/2005 - Effluent
9/17/2005 - Influent
9/17/2005 - Effluent
4/3/2006 - Influent
4/3/2006 - Effluent
680
110
1600
990
280
170
220
150
200
170
140
140
3,800
500
780
540
550
190
1800
1200
280
170
250
130
300
120
160
110
3,100
600
1400
650
19
73
13
21
0
0
14
13
50
29
14
21
18
20
79
20
Laboratory precision: As part of their QA/QC program, PSH analyzed duplicate samples from
the cone splitter for every storm for which there was sufficient sample volume. Summaries of the
laboratory duplicate data are presented in Table 6-4. Laboratory spikes were discussed as part of
the sampler calibration  (see Table  6-1).  As can  be seen from an  analysis of that data, the
precision of the sampling of the automatic sampler and the analysis combined are an average of
                                           43

-------
25%. The data show that the quality of sampling and analysis was maintained throughout the
course of the project.

6.1.3  Accuracy

Method accuracy  was  determined and  monitored using  a combination  of MS/MSD and
laboratory control  samples (known concentration in blank  water).  This information was also
pulled from the sampler calibration data. The MS/MSD information showed that the accuracy
achieved by the automatic sampler and the full analytical procedures was 2%. The MS/MSD data
are evaluated  by calculating the deviation  from perfect recovery (100%), while laboratory
control data are evaluated by calculating the absolute value of deviation from the laboratory
control concentration.

The  balance used for TSS and  SSC analyses was calibrated routinely with weights that were
NIST traceable. The laboratory maintained calibration records. The temperature of the  drying
oven was also monitored using a  thermometer that was calibrated  with an NIST traceable
thermometer.
Table 6-4. Laboratory Duplicate Sample Relative Percent Difference Data Summary

                                                                 Standard
                           Average     Maximum    Minimum     Deviation    Objective
  Parameter    Count        (%)          (%)         (%)          (%)         (%)
TSS
SSC
30
30
11
15
75
84
0
0
16
19
0-30
0-30
6.1.4  Representativeness

The field procedures were designed to ensure that representative samples were collected of both
influent and effluent stormwater. Field duplicate samples and  supervisor oversight provided
assurance that  procedures  were being  followed. The challenge in sampling  stormwater 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
stormwater, and redundant  methods  of evaluating key  constituent loadings in the stormwater
were utilized to compensate for the variability of the laboratory data.

The  laboratories  used  standard  analytical methods, with written SOPs for each method,  to
provide a consistent approach  to  all  analyses. Sample handling, storage,  and  analytical
methodology were reviewed 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 actual stormwater conditions.
                                           44

-------
6.1.5   Completeness

Completeness is a measure of the number of valid samples and measurements that are obtained
during a test period. Completeness will be measured by tracking the number of valid data results
against the specified requirements of the test plan. The goal for this data quality objective was to
achieve 80% completeness for flow and analytical data. The data quality objective was exceeded,
with discrepancies noted below:

•  The flow data (15 events, influent and effluent monitoring per event) is complete for all of
   the monitored events, except for the effluent flow data on the 14th storm. This resulted in the
   flow data being greater than 95% complete.
•  Duplicate samples for TSS and SSC were not analyzed for Events 8 and 9  due to insufficient
   sample volume collected.
•  Sieved TSS and SSC were not analyzed for Events 1 and 2. This analytical parameter was
   not in the original test plan but was added to account, by mass, for the fraction outside of the
   range  of the Coulter Counter. It had been assumed  that few particles larger than 250  |j,m
   would be found in the influent to the device because of the piping problems and creation of a
   miniature detention pond in the influent pipe upstream of the samplers. It was assumed that
   these larger particles would have settled out/been filtered out prior to the Terre Kleen™.

These  issues are appropriately flagged in the analytical reports and the data used in the final
evaluation of the Terre Kleen™ device.
                                           45

-------
                                     Chapter 7
                      Operation and Maintenance Activities

7.1    System Operation and Maintenance

The vendor designed the device to require periodic, but infrequent maintenance. During device
installation, Mr. Jim Close of the City of Harrisburg asked what the maintenance interval of the
device would be since the device would be maintained by the City. The design maintenance
interval was indicated to be a minimum of one year. The device was cleaned prior to the start of
testing in March 2005.

PSH personnel periodically inspected the device during the test period. If there was a question
about the device maintenance during one of these visits, representatives of THCP were contacted
and they  made a site visit with  the PSH personnel. The device was cleaned prior to the start of
testing in March, 2005. A review of the storm event records in January 2006 showed that two
late January storms had substantial negative removals. Therefore, the decision was made to clean
the device at the end of January 2006. This maintenance activity consisted of using  a sewer
vactor truck from the City of Harrisburg to dewater and remove sediment from the device and to
approximate depth  of sediment in the device.  The  TK09 unit was designed to  store 74 ft3 of
sediment. Approximate  depths of sediment  were  measured and  were  in accordance with
measurements taken during the start-up part of the  project. Sediment depths prior to pump-out
were between 50%  and 75% of the maximum design sediment depth, measured at several points
in the device. A picture of the device before and after maintenance is shown in Figures 7-1 and
7-2, respectively.
                   (a)
(b)
Table  7-1. (a) Initial cleanout of the sedimentation  chamber,  (b)  Bottom  of primary
chamber after dewatering and during sediment cleanout.
                                          46

-------
Table 7-2.  Primary chamber nearing the end of cleanout.

THCP  indicates that the sedimentation rate is the primary factor for determining maintenance
frequency,  and that a maintenance schedule should be based on  site-specific  sedimentation
conditions.  Observations by the TO during pumping indicated  that the Terre  Kleen™ was
relatively easy to pump out. The device is constructed so that the plate section can be tilted
against the primary chamber headwall to open up the floor of the device below the plates for
easy cleaning access. This was relatively easy at the test site because a traditional lid with round
manholes (for access) were never installed. Lid design improvements are being considered by
THCP to improve access to the sediment removal areas in future installations.
                                          47

-------
                                     Chapter 8
                                     References
1.      American Public Health Association. Standard methods for the examination of water and
       wastewater, 18th edition. American Public Health Association, American Water Works
       Association, Water Pollution Control Federation, Alexandria, VA, 1995.
2.      American Water  Works Association. Handbook for Community Water Systems:  Water
       Quality and Treatment, Fourth Edition.  McGraw-Hill, 1990.
3.      McCuen, R.H. Hydrologic Analysis and Design, Third Edition. Prentice-Hall, Inc., Upper
       Saddle River, NJ, 2005.
4.      National Oceanic and Atmospheric Administration (NOAA).  Technical Paper No. 40
       Rainfall Frequency Atlas of the United States. Washington, DC, 2000.
5.      NRCS (Natural Resources Conservation Service). Urban hydrology for small watersheds
       (Technical Release 55). Natural Resources Conservation Service, 1986.
6.      PA DOT (Pennsylvania Department of Transportation). Field Manual  of Pennsylvania
       Department   of   Transportation:  Storm   Intensity-Duration-Frequency  Charts.
       Pennsylvania Department of Transportation and the Federal Highway Administration,
       1986.
7.      Penn State Harrisburg. Environmental Technology Verification Test Plan for Terre Hill
       Concrete Products:  Terre Kleen. City  of Harrisburg Public Works  Yard, Harrisburg,
       Pennsylvania. November 2004.
8.      U.S. Environmental Protection Agency. Handbook of sampling and sample preservation
       of water and wastewater. EPA 600/4-82-029 (including later additions under this report
       number). U.S. Environmental Protection Agency, Washington, D.C., 1982.
9.      U.S. Environmental  Protection Agency. Methods for chemical  analysis  of water and
       wastes. EPA 600/4-79-020  (including  later additions and revisions under  this  report
       number). U.S.  Environmental Protection Agency,  Washington, D.C. (1979; revised
       1983).
10.    U.S. Environmental  Protection Agency. Quality criteria for water. EPA  440/5-86-001.
       U.S. Environmental Protection Agency, Washington, D.C. , 1986.
11.    United States Environmental Protection Agency. Methods and Guidance for Analysis of
       Water, EPA 821-C-99-008, Office of Water, Washington, DC,  1999.
                                          48

-------
                              Appendices
A    Design and O&M Guidelines
B    Verification Test Plan
C    Event Hydrographs and Rain Distribution
D    Analytical Data Reports with QC
                                  49

-------