June 2005
NSF05/10/EPADWCTR
EIMS ID: 136223
EPA/600/R-05/073
Environmental Technology
Verification Report
Removal of Arsenic in Drinking Water
ADI International Inc.
ADI Pilot Test Unit No. 2002-09
withMEDIAG2®
Prepared by
NSF International
Under a Cooperative Agreement with
U.S. Environmental Protection Agency
-------
THE ENVIRONMENTAL TECHNOLOGY VERIFICATION
PROGRAM
ET
U.S. Environmental Protection Agency NSF International
ETV Joint Verification Statement
TECHNOLOGY TYPE: ARSENIC ADSORPTION MEDIA FILTER USED IN
DRINKING WATER TREATMENT SYSTEMS
APPLICATION: REMOVAL OF ARSENIC IN DRINKING WATER
TECHNOLOGY NAME: ADI PILOT TEST UNIT NO. 2002-09 WITH MEDIA G2®
COMPANY: ADI INTERNATIONAL INC.
ADDRESS: SUITE 300 PHONE: (506)452-9000
1133 REGENT STREET FAX: (506) 459-3954
FREDERICTON, NB E3B 3Z2 CANADA
WEB SITE: www.adi.ca
EMAIL: mjm@adi.ca
The U.S. Environmental Protection Agency (EPA) supports 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
(consisting of buyers, vendor organizations, and permitters), and with the full participation of individual
technology developers. The program evaluates the performance of innovative technologies by developing
test plans that are responsive to the needs of stakeholders, conducting field or laboratory tests (as
appropriate), collecting and analyzing data, and preparing peer-reviewed reports. All evaluations are
conducted in accordance with rigorous quality assurance protocols to ensure that data of known and
adequate quality are generated and that the results are defensible.
NSF International (NSF), in cooperation with the EPA, operates the Drinking Water Systems (DWS)
Center, one of seven technology areas under the ETV Program. The DWS Center recently evaluated the
performance of an adsorption media filter system for the reduction of arsenic in drinking water. This
verification statement provides a summary of the test results for the ADI Pilot Test Unit No. 2002-09 with
MEDIA G2® system. Gannett Fleming, Inc., an NSF-qualified field testing organization (FTO),
performed the verification testing. The verification report contains a comprehensive summary of the
verification test.
05/10/EPADWCTR The accompanying notice is an integral part of this verification statement. June 2005
VS-i
-------
ABSTRACT
Verification testing of the ADI International Inc. Pilot Test Unit No. 2002-09 with MEDIA G2® arsenic
adsorption media filter system was conducted at the Hilltown Township Water and Sewer Authority
(HTWSA) Well Station No. 1 in Sellersville, Pennsylvania from October 8, 2003 through May 28, 2004.
The source water was groundwater from Well No. 1, one of HTWSA's three groundwater supply wells.
The treatment unit feed water for the verification test was withdrawn from an on-site chlorine detention
tank, which contained groundwater that had been disinfected with sodium hypochlorite. Verification
testing was conducted under manufacturer-specified operating conditions. The feed water, with an
average total arsenic concentration of 21 |jg/L and a pH of 7.6, was treated with sulfuric acid to lower the
pH to 6.4 prior to the treatment unit. When operated under the manufacturer's specified conditions for
this site and at the design flow rate of 1.7 gpm, the ADI International Inc. Pilot Test Unit No. 2002-09
with MEDIA G2® system reduced the total arsenic concentration from an average of 21 ug/L in the feed
water to an average of 7 |jg/L in the treated water.
TECHNOLOGY DESCRIPTION
The following technology description was provided by the manufacturer and has not been verified.
MEDIA G2® is an iron-based adsorption treatment technology for removing arsenic from drinking water
supplies, specifically groundwater. MEDIA G2® arsenic adsorption media consists of an inorganic,
natural substrate to which iron (ferric hydroxide) has been chemically bonded. The iron attracts metallic
ions in water and binds them to the substrate by chemisorption. The arsenic adsorption filter pilot unit
used in this test consisted of one vessel containing MEDIA G2® adsorption media which was operated in
a downflow mode. Arsenic is removed by the technology by adsorption onto the filter media as water
passes through the media. Over time, as the media becomes saturated with arsenic, the concentration of
arsenic in the treated water begins to increase. Before the treated water arsenic concentration reaches the
pre-determined maximum allowable contaminant level (breakthrough), the media is either replaced or
regenerated on-site. ADI has stated that MEDIA G2® can be regenerated four to five times, with a loss in
capacity of approximately 10% following each regeneration.
MEDIA G2® is a registered trade mark of ADI International Inc. and is protected by US Patent No.
6,200,482. MEDIA G2® adsorption media is certified under NSF/ANSI Standard 61 for water treatment
plant applications. MEDIA G2® treatment units can be used for groundwater supplies of any size and
require limited manpower and operating skills. The filter system can operate continuously or
intermittently. The filter tank is freestanding, and filter components, which are modular in nature, can be
installed by a qualified plumber. The filter system requires only a level surface capable of supporting its
weight, sustained ambient temperature above 35°F, a feed water pressure between 20 and 125 psi, and
flow rate control.
VERIFICATION TESTING DESCRIPTION
Test Site
The verification testing site was the HTWSA Well No. 1 in Sellersville, Pennsylvania. The source water
was groundwater from Well No. 1, which was first disinfected with sodium hypochlorite. Well No. 1 is
one of three wells currently used to supply the HTWSA water distribution system. The feed water quality
was particularly variable for a groundwater supply. During the verification test, the turbidity ranged from
0.15 NTU to 7.6 NTU and averaged 0.70 NTU. The feed water iron concentration ranged from 47 ug/L
to 1,120 ug/L and averaged 180 ug/L. The feed water manganese concentration ranged from 77 ug/L to
1,070 ug/L and averaged 140 ug/L. The feed water was characterized as having a high level of hardness,
05/10/EPADWCTR The accompanying notice is an integral part of this verification statement. June 2005
VS-ii
-------
270 mg/L as CaCO3, and a high degree of buffering as indicated by an alkalinity of 120 mg/L as CaCO3.
The raw water pH was relatively stable at 7.6, but the feed water pH varied due to the operation of the
acid feed pump. It ranged from 5.7 to 7.1, with an average of 6.4. The feed water total arsenic
concentration ranged from 12 ug/L to 63 ug/L and averaged 21 ug/L.
Methods and Procedures
Operations, sampling, and analytical procedures were performed in a manner that ensured the quality of
the data collected and provided an accurate evaluation of the treatment system under field conditions.
The verification test consisted of three main phases. The first phase, the Integrity Test, evaluated the
reliability of equipment operation under the environmental and hydraulic conditions at the well station
site during the initial two weeks of testing. The second phase, the Capacity Test, evaluated the capacity
of the arsenic adsorption system with respect to arsenic. The third phase of the test monitored the
performance of the system for one month following regeneration.
The Integrity Test ran for 13 full days plus eight hours, during which the field test operator was on-site
twice per day to monitor the test equipment, collect data, and collect water samples for analysis. The
Capacity Test began in conjunction with the Integrity Test on October 8, 2003 and continued through the
media regeneration on April 30, 2004. One month of post-regeneration operation began on April 30,
2004 and continued through May 28, 2004. The treatment system was operated continuously,
independent of the well operations, using water supplied from the well station's pressurized chlorine
detention tank. Flow rate, production volume, and pressure were monitored and recorded twice per day.
Raw, feed (before and after addition of sulfuric acid), and treated water samples were analyzed for pH,
temperature, turbidity, alkalinity, calcium, magnesium, hardness, free available chlorine, and fluoride by
the field test operator. Samples were collected and delivered to the Pennsylvania Department of
Environmental Protection Laboratory to be analyzed for silica, sodium, aluminum, iron, manganese,
chloride, sulfate, and total phosphorus. Arsenic samples were collected and sent to NSF's laboratories for
analysis. A total of 14 sets of arsenic samples were speciated during the test to determine the relative
concentration of soluble arsenic compared to total arsenic, and, with respect to the soluble arsenic, the
relative amounts of arsenic III and arsenic V.
Complete descriptions of the verification testing results and quality assurance/quality control procedures
are included in the verification report.
VERIFICATION OF PERFORMANCE
System Operation
The verification test was conducted under the manufacturer's specified operating conditions. Contact
time is a critical parameter for arsenic adsorption efficiency and is dependent upon maintaining the flow
rate within the design range of 1.7 gpm ±0.1 gpm. A pressure-reducing valve was used to reduce the
pressure from the chlorine detention tank from 110 psi to 50 psi to make throttling the flow rate easier for
the operator. A relatively constant flow rate was maintained, with minimal flow rate adjustments
required. The system was operated continuously, 24 hours each day, for the entire test. The filter unit
was manually backwashed and rinsed 15 times throughout the test, based on the accumulation of filter
bed headless.
Water Quality Results
The results of total arsenic analyses are shown in Figure VS-1. During the Capacity Test, the feed water
total arsenic concentration averaged 21 |Jg/L, with 13 ug/L in the soluble state. Pretreatment with
hypochlorite completely converted 1he feed water soluble arsenic to the arsenic V species. The treated
05/10/EPADWCTR The accompanying notice is an integral part of this verification statement. June 2005
VS-iii
-------
water total arsenic concentration averaged 7 |j,g/L during the Capacity Test, all of which was in the
soluble state. For calculation of the media capacity to remove arsenic from the feed water, 430,000
gallons were treated from October 8, 2003 through April 22, 2004 during the Capacity Test. The treated
water volume represents 25,000 media bed volumes, based on the calculated bed volume of 2.3 cubic feet
and an empty bed contact time of ten minutes. Based on the feed and treated water total arsenic
concentrations during the Capacity Test, the capacity of the media for this system, through April 22,
2004, was 470 |jg arsenic per gram of media.
One media regeneration was performed during the verification test. As shown in Figure VS-1, treated
water arsenic concentrations were elevated for several hours following the media regeneration. However,
the post-regeneration treated water arsenic concentration (April 30, 2004 through May 28, 2004) returned
to a level similar to that observed at the beginning of the Capacity Test, averaging 4 ug/L, which indicates
that the media regeneration was successful.
70
65
60
55
50
45
40
35
30
25
Date of Media Regeneration
10/31/03
11/30/03
12/30/03
1/29/04
Date
4/28/04
"Raw~* Feed(ST2)'
-Treated(ST3) I
Figure VS-1. Capacity Test Arsenic Concentration.
The addition of sulfuric acid prior to the treatment unit reduced the pH of the raw water from an average
of 7.6 to 6.4 in the feed water. The pH reduction corresponded with a 21% reduction in alkalinity. The
sulfate concentration increased from an average of 100 mg/L in the raw water to 160 mg/L in the feed
water, following the addition of sulfuric acid. The feed water pH appeared to have a significant impact on
the treatment unit's ability to remove arsenic. The highest treated water arsenic concentrations occurred
when the feed water pH was highest. The manufacturer indicated that the feed water pH should be
maintained between 6.5 and 6.8 for optimum arsenic removal, but difficulties encountered with the acid
feed pump operation resulted in several periods during the verification test when the pH was above this
range. As an example of the correlation, a decrease in feed water pH from 7.1 to 6.2 on the ninth day of
the test resulted in a 70% decrease in the treated water arsenic concentration. Thereafter, correlations in
treated water arsenic with the feed water pH were not as significant but continued to occur. At the request
05/10/EPADWCTR
The accompanying notice is an integral part of this verification statement.
VS-iv
June 2005
-------
of the manufacturer, testing was carried out at reduced pH from April 23, 2004 through April 30, 2004.
During the reduced pH operation, the treated water arsenic concentration averaged 6 ug/L.
Feed water calcium and hardness concentrations were reduced only slightly through the adsorption filter.
The average feed water iron and manganese concentrations during the Capacity Test, 180 ug/L and
140 ug/L, respectively, were significantly reduced by the adsorption filter. The treated water iron
concentration averaged 68 ug/L and the treated water manganese concentration averaged 16 ug/L.
Turbidity was also reduced by the adsorption filter during the Capacity Test, from an average of
0.70 NTU in the feed water to 0.30 NTU in the treated water. The silica concentration increased by an
average of 15%, from a feed concentration of 28 mg/L to a treated water concentration of 33 mg/L.
Sodium, fluoride, chloride, aluminum, and sulfate concentrations were generally unaffected by the
adsorption filter.
Operation and Maintenance Results
The verification test began on October 8, 2003 and ended on May 28, 2004. The treatment unit operated
manually, including backwash cycles, throughout the test. The majority of operator time and attention
was spent on water quality and equipment testing. Equipment operation required minimal operator
attention overall, with the exception of the sulfuric acid metering pump, which required frequent re-
priming and feed rate adjustment to maintain the feed water pH within the manufacturer's stated
operating range. Periodic manual filter backwashes each required 1.5 to 2 hours of operator time, and
media regeneration required approximately five hours. Fifteen manual filter backwashes and one media
regeneration were performed during the verification test. The backwash water was relatively turbid and
contained elevated concentrations of iron, manganese, aluminum, and arsenic. Arsenic in the backwash
water was primarily in particulate form, which indicates the removal of participate material from the
filter, not desorption of arsenic from the media. The treated water arsenic concentration returned to
approximately that of the new media following the media regeneration, which indicates a successful
regeneration. However, a spike in the treated water arsenic concentration occurred when the unit was
returned to service following the media regeneration. Modification of the media regeneration procedures
and increased on-site monitoring of the treated water arsenic concentration may be required to prevent
returning a unit to service with an elevated treated water arsenic concentration immediately following
regeneration. Other than monitoring the metering pump and performing filter backwashes, regular
operator attention was primarily required to verify, adjust, and maintain a constant flow rate.
Consumables and Waste Generation
Electrical power was required only for the metering pump and a solenoid valve. The solenoid valve was
provided to automatically shut off the feed water supply in the event of a power outage to prevent water
from entering the treatment unit without pH adjustments. Wastewater from each filter backwash and
rinse was discharged to a sanitary sewer adjacent to the well station. The total water usage for each
backwash and rinse was approximately 200 gallons, for a total backwash and rinse water usage of 2,800
gallons. The backwash and rinse water usage represents 0.5% of the total throughput of 520,000 gallons
during the test, including the Integrity, Capacity, and post-regeneration phases.
The media regeneration, which was performed once during the verification test following seven months
of operation, required three bed volumes (50 gallons) of 1% caustic soda, 20 gallons of 0.5% sulfuric acid
solution, and rinse water.
Quality Assurance/Quality Control
NSF provided technical and quality assurance oversight of the verification testing as described in the
verification report, including an audit of nearly 100% of the data. NSF personnel also conducted a
05/10/EPADWCTR The accompanying notice is an integral part of this verification statement. June 2005
VS-v
-------
technical systems audit during 1he verification test to ensure the testing was in compliance with the test
plan. A complete description of the QA/QC procedures is provided in the verification report.
Original Signed by
Sally Gutierrez
8/19/05
Original Signed by
Robert Ferguson
8/30/05
Sally Gutierrez Date
Director
National Risk Management Research Laboratory
Office of Research and Development
United States Environmental Protection Agency
Robert Ferguson
Vice President
Water Systems
NSF International
Date
NOTICE: Verifications are based on an evaluation of technology performance under specific,
predetermined criteria and the appropriate quality assurance procedures. EPA and NSF make no
expressed or implied warranties as to the performance of the technology and do not certify that a
technology will always operate as verified. The end-user is solely responsible for complying with
any and all applicable federal, state, and local requirements. Mention of corporate names, trade
names, or commercial products does not constitute endorsement or recommendation for use of
specific products. This report is not an NSF Certification of the specific product mentioned
herein.
Availability of Supporting Documents
Copies of the ETV Protocol for Equipment Verification Testing for Arsenic Removal
dated April 2002, the verification statement, and the verification report (NSF Report
#05/10/EPADWCTR) are available from the following sources:
(NOTE: Appendices are not included in the verification report. Appendices are available
from NSF upon request.)
1. ETV Drinking Water Systems Center Manager (order hard copy)
NSF International
P.O. Box 130140
Ann Arbor, Michigan 48113-0140
2. NSF web site: http://www.nsf. org/etv (electronic copy)
3. EPA web site: http://www. epa.gov/etv (electronic copy)
05/10/EPADWCTR
The accompanying notice is an integral part of this verification statement.
VS-vi
June 2005
-------
June 2005
Environmental Technology Verification Report
Removal of Arsenic in Drinking Water
ADI International Inc.
ADI Pilot Test Unit No. 2002-09 with MEDIA G2
Prepared for:
NSF International
Ann Arbor, Michigan 48105
Prepared by:
Gannett Fleming, Inc.
Harrisburg, Pennsylvania 17106-7100
Under a cooperative agreement with the U.S. Environmental Protection Agency
Jeffrey Q. Adams, Project Officer
National Risk Management Research Laboratory
U.S. Environmental Protection Agency
Cincinnati, Ohio 45268
-------
Notice
The U.S. Environmental Protection Agency (EPA), through its Office of Research and
Development, has financially supported and collaborated withNSF International (NSF) under
Cooperative Agreement No. R-82833301. This verification effort was supported by the Drinking
Water Systems (DWS) Center, operating under the Environmental Technology Verification
(ETV) Program. This document has been peer-reviewed, reviewed by NSF and EPA, and
recommended for public release.
-------
Foreword
The EPA is charged by Congress with protecting the Nation's land, air, and water resources.
Under a mandate of national environmental laws, the Agency strives to formulate and implement
actions leading to a compatible balance between human activities and the ability of natural
systems to support and nurture life. To meet this mandate, EPA's research program is providing
data and technical support for solving environmental problems today and building a science
knowledge base necessary to manage our ecological resources wisely, understand how pollutants
affect our health, and prevent or reduce environmental risks in the future.
The National Risk Management Research Laboratory (NRMRL) is the Agency's center for
investigation of technological and management approaches for preventing and reducing risks
from pollution that threaten human health and the environment. The focus of the Laboratory's
research program is on methods and their cost-effectiveness for prevention and control of
pollution to air, land, water, and subsurface resources; protection of water quality in public water
systems; remediation of contaminated sites, sediments and ground water; prevention and control
of indoor air pollution; and restoration of ecosystems. NRMRL collaborates with both public
and private sector partners to foster technologies that reduce the cost of compliance and to
anticipate emerging problems. NRMRL's research provides solutions to environmental
problems by: developing and promoting technologies that protect and improve the environment;
advancing scientific and engineering information to support regulatory and policy decisions; and
providing the technical support and information transfer to ensure implementation of
environmental regulations and strategies at the national, state, and community levels.
This publication has been produced as part of the Laboratory's strategic long-term research plan.
It is published and made available by EPA's Office of Research and Development to assist the
user community and to link researchers with their clients.
Sally Gutierrez, Director
National Risk Management Research Laboratory
in
-------
Contents
Section Page
Verification Statement VS-i
Title Page i
Notice ii
Foreword iii
Table of Contents iv
Abbreviations and Acronyms x
Acknowledgements xii
Chapter 1 Introduction 1
1.1 ETV Purpose and Program Operation 1
1.2 Testing Participants and Responsibilities 1
1.2.1 NSF International 2
1.2.2 Field Testing Organization 2
1.2.3 Manufacturer 3
1.2.4 Analytical Laboratories 3
1.2.5 Pennsylvania Department of Environmental Protection 4
1.2.6 U.S. Environmental Protection Agency 4
1.3 Verification Testing Site 5
1.3.1 Feed Water 5
1.3.2 Pilot Filter Discharges 6
Chapter 2 Equipment Description and Operating Processes 8
2.1 Equipment Description 8
2.1.1 Basic Scientific and Engineering Concepts of Treatment 8
2.1.2 Generic Arsenic Adsorption Media Description 8
2.1.3 Capacity 9
2.1.4 Intermittent Operation 10
2.1.5 MEDIA G2® Description 10
2.2 ADI's Arsenic Adsorption Media Test Unit Installation 12
2.2.1 Filter System Components 14
2.2.2 Physical Construction of the Test Unit 14
2.2.3 Equipment Performance Range and Known Limitations of Equipment 14
2.2.4 Drawings of Equipment 15
2.2.5 Design Criteria 18
2.2.6 Photographs of Equipment 21
2.2.7 Data Plate 21
2.3 Operating Process 22
2.3.1 Operator Requirements 23
2.3.2 Required Consumables 23
2.3.3 Rates of Waste Product! on 24
2.3.3.1 Backwash 24
2.3.3.2 Regeneration 24
iv
-------
2.3.4 Licensing Requirements Associated with Equipment Operation 24
Chapter 3 Methods and Procedures 25
3.1 Experimental Design 25
3.1.1 Objectives 25
3.1.2 Equipment Characteristics 25
3.1.2.1 Qualitative Factors 25
3.1.2.2 Quantitative Factors 26
3.1.2.3 Raw and Feed Water Quality 26
3.2 Equipment Operations and Design 27
3.3 Field Test Equipment 27
3.4 Communications, Documentation, Logistics, and Equipment 27
3.5 Equipment Operation and Water Quality Sampling for Verification Testing 28
3.6 Recording Data 28
3.7 Recording Statistical Uncertainty for Assorted Water Quality Parameters 29
3.8 Verification Testing Schedule 30
3.9 Task 1: System Integrity Verification Testing 30
3.9.1 Introduction 30
3.9.2 Experimental Objectives 30
3.9.3 Work Plan 30
3.9.4 Analytical Schedule 31
3.9.5 Evaluation Criteria and Minimum Reporting Requirements 35
3.10 Task 2: Adsorption Capacity Verification Testing 35
3.10.1 Introduction 35
3.10.2 Experimental Objectives 36
3.10.3 Work Plan 36
3.10.4 Analytical Schedule 36
3.10.5 Evaluation Criteria and Minimum Reporting Requirements 39
3.11 Task 3: Documentation of Operating Conditions and Treatment Equipment
Performance 41
3.11.1 Introduction 41
3.11.2 Experimental Objectives 42
3.11.3 Work Plan 42
3.11.4 Schedule 42
3.11.5 Evaluation Criteria 42
3.12 Task 4: Data Management 42
3.12.1 Introduction 42
3.12.2 Experimental Objectives 43
3.12.3 Work Plan 43
3.13 Task 5: Quality Assurance/Quality Control 44
3.13.1 Introduction 44
3.13.2 Experimental Objectives 44
3.13.3 Work Plan 44
3.13.4 Analytical Methods 44
3.13.5 Samples Shipped Off-Site for Analysis 45
-------
3.13.6 Tests and Data Specific to Adsorption Media Type Evaluated 47
3.14 Operations and Maintenance 47
3.14.1 Operations 47
3.14.2 Maintenance 47
Chapter 4 Results and Discussion 48
4.1 Introduction 48
4.2 Equipment Installation, Startup, and Shakedown 48
4.3 Task 1: System Integrity Verification Testing 50
4.3.1 Experimental Objectives 50
4.3.2 Preliminary Water Quality Analyses 50
4.3.3 Integrity Test Operational Data 53
4.3.4 Integrity Test On-site Water Quality Analyses 55
4.3.5 Integrity Test Laboratory Water Quality Analyses 61
4.3.6 Integrity Test Laboratory Arsenic Analyses 65
4.3.7 Integrity Test Equipment Operation 70
4.4 Task 2: Adsorption Capacity Verification Testing 70
4.4.1 Experimental Objectives 70
4.4.2 Capacity Test Operational Data 70
4.4.3 Capacity Test On-site Water Quality Analyses 72
4.4.4 Capacity Test Laboratory Water Quality Analyses 81
4.4.5 Capacity Test Arsenic Analyses 99
4.4.6 Capacity Test Equipment Operation 109
4.5 Capacity Test Backwash Water Quality, Quantity, and Flow Rate 109
4.6 Media Regeneration Ill
4.7 Task 3: Documentation of Operating Conditions and Treatment Equipment 113
4.7.1 Introduction 113
4.7.2 Experimental Objectives 114
4.8 Task 4: Data Management 114
4.9 Task 5: Quality Assurance/Quality Control 114
4.9.1 Introduction 114
4.9.2 Data Quality Indicators 115
4.9.2.1 Representativeness 115
4.9.2.2 Accuracy 115
4.9.2.2.1 Field Equipment Calibrations 115
4.9.2.2.2 Split Samples 116
4.9.2.2.3 Performance Evaluation Samples for Water Quality
Testing 117
4.9.2.2.4 Spike Sample Analyses 118
4.9.2.3 Precision 118
4.9.2.4 Statistical Uncertainty 119
4.9.2.5 Completeness 119
VI
-------
Chapter 5 References 121
Chapter 6 Vendor Comments 122
Tables Page
1-1 Feed (ST2) Water Quality During Integrity Testing 7
2-1 Manufacturing and Procedures Specific to the MEDIA G2 Adsorptive Media 11
2-2 MEDIA G2® Adsorptive Media Specifications 12
2-3 Equipment Design Criteria 18
3-1 Field Analytical and Calibration Equipment 27
3-2 On-site Equipment Operating Parameter Monitoring and Data Collection Schedule 32
3-3 Water Quality Sampling Schedule - System Integrity Verification Testing 33
3-4 Arsenic Sampling Plan 34
3-5 Backwash Wastewater and Rinse Water Monitoring, Sampling, and Analyses 35
3-6 Water Quality Sampling Schedule - Media Adsorptive Capacity Verification Testing ...38
3-7 Water Quality Sampling Schedule - Post-Regeneration Media Verification Testing 40
3-8 Regeneration Waste Monitoring, Sampling and Analysis 41
3-9 Schedule for Observing and Recording Equipment Operation and Performance Data 42
3-10 Water Quality Sampling Protocol 46
4-1 Preliminary Arsenic Speciation 49
4-2 Preliminary On-site Water Quality Analyses (9/15/03 to 9/19/03) - After Tank
Cleaning 52
4-3 Preliminary Laboratory Water Quality Analyses (9/15/03 to 9/19/03) - After Tank
Cleaning 53
4-4 Integrity Test Operational Data 54
4-5 Integrity Test On-site Water Quality Analyses 57
4-6 Integrity Test Laboratory Water Quality Analyses 62
4-7 Integrity Test Laboratory Arsenic Analyses 66
4-8 Capacity Test Operational Data 71
4-9 Post-Regeneration Operational Data 71
4-10 Capacity Test On-site Water Quality Analyses 73
4-11 Post-Regeneration On-site Water Quality Analyses 74
4-12 Capacity Test Laboratory Water Quality Analyses 82
4-13 Post-Regeneration Laboratory Water Quality Analyses 83
4-14 Capacity Test Laboratory Arsenic Analyses 100
4-15 Post-Regeneration Laboratory Arsenic Analyses 101
4-16 Capacity Test Backwash and Rinse Water Characteristics Ill
4-17 Regeneration Wastewater Quality 113
4-18 Regeneration Wastewater Arsenic Concentration 113
4-19 Media Regeneration Wastewater Characterization 113
4-20 Field Instrument Calibration Schedule 116
4-21 Split Samples (May 25, 2004) 117
Figures Page
2-1 Schematic of MEDIA G2® Arsenic Removal Pilot Unit 16
vn
-------
2-2 Schematic of MEDIA G2® Arsenic Removal Pilot Unit 17
2-3 ADI International Inc. Pilot Test Unit. No. 2002-09 with Media G2® as installed at
the HTWSA Well Station No. 1 21
4-1 Integrity Test Filter Loss of Head, Feed, and Treated Pressure 54
4-2 Integrity Test Temperature 58
4-3 Integrity Test pH 58
4-4 Integrity Test Turbidity 59
4-5 Integrity Test Alkalinity 59
4-6 Integrity Test Fluoride 60
4-7 Integrity TestFAC 60
4-8 Integrity Test Sodium 63
4-9 Integrity Test Silica 63
4-10 Integrity Test Iron 64
4-11 Integrity Test Manganese 64
4-12 Integrity Test Raw Total, Soluble and Speciated Arsenic 67
4-13 Integrity Test Feed (ST2) Total, Soluble and Speciated Arsenic 67
4-14 Integrity Test Total Arsenic 68
4-15 Integrity Test Affect of pH on Arsenic Removal 68
4-16 Integrity Test Treated Total, Soluble and Speciated Arsenic 69
4-17 Integrity TestFAC and Arsenic Removal 69
4-18 Capacity Test Feed Pressure, Treated Pressure, and Filter Bed Headloss 72
4-19 Capacity Test pH 75
4-20 Capacity Test Temperature 76
4-21 Capacity Test Turbidity 77
4-22 Capacity Test Alkalinity Concentration 78
4-23 Capacity Test Fluoride Concentration 79
4-24 Capacity Test FAC 80
4-25 Capacity Test Calcium, Magnesium, and Hardness 81
4-26 Capacity Test Sodium Concentration 85
4-27 Capacity Test Silica Concentration 86
4-28 Capacity Test Aluminum Concentration 87
4-29 Capacity Test Raw Water Iron Concentration 88
4-30 Capacity Test Feed Water Iron Concentration 89
4-31 Capacity Test Feed (ST2) Total and Soluble Iron Concentration 90
4-32 Capacity Test Treated Water Iron Concentration 91
4-33 Capacity Test Raw Water Manganese Concentration 92
4-34 Capacity Test Manganese Concentration 93
4-35 Capacity Test Feed (ST2) and Treated Manganese Concentration 94
4-36 Capacity Test Treated Water Manganese Concentration 95
4-37 Capacity Test Chloride Concentration 96
4-38 Capacity Test Sulfate Concentration 97
4-39 Capacity Test Phosphorus Concentration 98
4-40 Capacity Test Raw Arsenic Concentration 102
4-41 Capacity TestFeed (ST2) Arsenic Concentration 103
4-42 Capacity Test Treated Water Arsenic Concentration 104
4-43 Capacity Test Treated Water Arsenic Concentration (0-25 jig/L Scale) 105
Vlll
-------
4-44 Capacity Test Total Arsenic Concentration 106
4-45 Capacity Test Arsenic Concentration and pH 107
4-46 Capacity Test Arsenic and FAC 108
Appendices
A ADI Operations Manual
B NSF/ANSI 61 Certification and Media MSDS
C Equipment Photographs
D ADI's Procedure for Media Replacement
E Arsenic Speciation Protocol
F On-Site Arsenic Analysis Procedure
G First Integrity Test Data
H Preliminary Arsenic Speciation Data
I Preliminary Water Quality Analyses, Logbook Copies, and Analytical Reports
J Operational Data, On-Site Water Quality Data, and Logbook Copies
K PADEP Laboratory Water Quality Data, Laboratory Test Reports, and Sample
Submission Forms
L NSF Laboratory Arsenic Data, Test Reports, and Chain of Custodies
M Wastewater TCLP Analytical Reports
N Media Gradation Analyses
O Analytical Performance Evaluation Checks
IX
-------
Acronyms and Abbreviations
ANSI
AWWA
CAWET
°C
EBCT
EPA
ETV
°F
ft
FTO
g
gpm
HOPE
HTWSA
ISE
ITS
L
Ib
m
MCL
mgd
mg/L
mL
MDL
MSDS
N/A
ND
NEMA
NIST
NRMRL
NSF
NTIS
NTU
O&M
PADEP
PE
PRV
psi
PSTP
PVC
QA
QC
QA/QC
QAPP
American National Standards Institute
American Water Works Association
California Waste Extraction Tests
Degrees Celsius
Empty Bed Contact Time
U. S. Environmental Protection Agency
Environmental Technology Verification
Degrees Fahrenheit
Feet
Field Testing Organization
Gram
Gallons per Minute
High Density Polyethylene
Hilltown Township Water and Sewer Authority
Ion Selective Electrode
Industrial Test Systems, Inc.
Liter
Pound
Meter
Maximum Contaminant Level
Million Gallons per Day
Milligram per Liter
Milliliter
Method Detection Level
Material Safely Data Sheets
Not Applicable
Non-Detect
National Electrical Manufacturers Association
National Institute of Standards and Technology
National Risk Management Research Laboratory
NSF International
National Technical Information Service
Nephelometric Turbidity Unit
Operation and Maintenance
PA Department of Environmental Protection
Performance Evaluation
Pressure Reducing Valve
Pounds per Square Inch
Product Specific Test Plan
Poly Vinyl Chloride
Quality Assurance
Quality Control
Quality Assurance/Quality Control
Quality Assurance Project Plan
-------
Acronyms and Abbreviations (continued)
SAC Standard Analysis Code
SCADA Supervisory Control and Data Acquisition
SMCL Secondary Maximum Contaminant Level
SM Standard Methods for the Examination of Water and Wastewater
SOP Standard Operating Procedure
TCLP Toxicity Characteristic Leaching Procedure
TSTP Technology Specific Test Plan
|j,g/L Microgram per Liter
XI
-------
Acknowledgements
The Field Testing Organization (FTO), Gannett Fleming, Inc., was responsible for all elements
in the testing sequence, including collection of samples, calibration and verification of
instruments, data collection and analysis, data management, data interpretation, and the
preparation of this report.
Gannett Fleming, Inc.
P.O. Box 67100
Harrisburg, PA 17106-7100
Phone: (717) 763-7212, Ext. 2109
Fax: (717)763-1808
Contact: William Allis, Project Manager
E-mail: wallis@gfnet.com
The laboratory selected for analysis of the water quality parameters, with the exception of
arsenic, a U.S. Environmental Protection Agency (EPA) accredited and Pennsylvania
Department of Environmental Protection (PADEP) certified laboratory, was:
PADEP Laboratories
Inorganic Services Division
Mailing Address:
P.O. Box 1467
Harrisburg, PA 17105-1467
Express Mail Address:
1500 North 3rd Street
Harrisburg, PA 17102
Phone: (717) 705-2197
Fax: (717)783-1502
Contact: Ted Lyter, Inorganic Services Division Chief
E-mail: plyter@state.pa.us
Regeneration wastewater toxicity analyses were performed by:
TriMatrix Laboratories, Inc.
5555 Glenwood Hills Parkway, SE
Grand Rapids, MI 49588
Phone: (616) 975-4500
Fax: (810) 220-2803
Contact: Michael W. Movinski, Vice President, Sales and Marketing
Email: mmtrimatrix@comcast.net
xn
-------
Arsenic analyses were performed by the NSF laboratory.
NSF International
789 N. Dixboro Rd.
Ann Arbor, MI 48105
Phone: (734) 769-8010
Fax: (734)769-0109
Contact: Bruce Bartley, Project Manager
Email: bartley@nsf.org
The manufacturer of the equipment was:
ADI International Inc.
Suite 300
1133 Regent Street
Fredericton, NB E3B 3Z2 Canada
Phone: (506) 452-9000
Fax: (506) 459-3954
Contact: Michael McMullin, P. Eng.
E-mail: mjm@adi.ca
Gannett Fleming wishes to thank the following participants:
NSF, especially Bruce Bartley, Angela Beach and Dale Scherger, for providing guidance and
program management.
PADEP, especially Ted Lyter and Dennis Neuin, for providing laboratory services.
David "Butch" Erwin, Operations Manager, Hilltown Township Water and Sewer Authority
(HTWSA).
xni
-------
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; with stakeholder
groups consisting 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 responsive to the needs of stakeholders, conducting field or
laboratory tests (as appropriate), collecting and analyzing data, and preparing peer-reviewed
reports. All evaluations are conducted in accordance with rigorous quality assurance protocols to
ensure that data of known and adequate quality are generated and that the results are defensible.
EPA has partnered with NSF International (NSF) under the ETV Drinking Water Systems
(DWS) Center to verify performance of small drinking water systems that serve small
communities. A goal of verification testing is to enhance and facilitate the acceptance of small
drinking water treatment equipment by state drinking water regulatory officials and consulting
engineers, while reducing the need for testing of equipment at each location where the
equipment's use is contemplated. NSF meets this goal by working with manufacturers and NSF-
qualified Field Testing Organizations (FTOs) to conduct verification testing under the approved
protocols. It is important to note that verification of the equipment does not mean 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 FTO.
The DWS Center evaluated the performance of the ADI International Inc. (ADI) Pilot Test Unit
No. 2002-09 with MEDIA G2® system, which is an adsorption media filter for the reduction of
arsenic in drinking water. The verification test evaluated the ability of the adsorptive media to
remove arsenic from drinking water. This document provides the verification test results for the
ADI Pilot Test Unit No. 2002-09 with MEDIA G2® system.
1.2 Testing Participants and Responsibilities
The ETV testing of the ADI International Inc. Pilot Test Unit No. 2002-09 with MEDIA G2®
was a cooperative effort among the following participants:
NSF International
Gannett Fleming, Inc.
-------
ADI International Inc.
Pennsylvania Department of Environmental Protection (PADEP)
U.S. Environmental Protection Agency
Hilltown Township Water and Sewer Authority
The following is a brief description of all ETV participants and their roles and responsibilities.
1.2.1 NSF International
NSF is an independent, not-for-profit testing and certification organization dedicated to public
health and safety and to the protection of the environment. Founded in 1946 and located in Ann
Arbor, Michigan, NSF has been instrumental in the development of consensus standards for the
protection of public health and the environment. NSF also provides testing and certification
services to ensure products bearing the NSF Name, Logo and/or Mark meet those standards.
EPA partnered with NSF to verify the performance of drinking water treatment systems through
EPA's ETV Program.
NSF provided technical oversight of the verification testing and conducted an audit of the field
analytical, data gathering, and recording procedures. NSF provided review of the Product
Specific Test Plan (PSTP) as well as this report.
Contact Information:
NSF International
789 N. Dixboro Rd.
Ann Arbor, MI 48105
Phone: (734) 769-8010
Fax: (734)769-0109
Contact: Bruce Bartley, Project Manager
Email: bartley@nsf.org
1.2.2 Field Testing Organization
Gannett Fleming, Inc., a consulting engineering firm located in Harrisburg, Pennsylvania,
conducted the verification testing of the ADI International Inc. arsenic removal system. Gannett
Fleming is an NSF-qualified FTO for the ETV DWS Center.
Gannett Fleming was responsible for conducting the Integrity Test for 14 calendar days (13 full
days plus 8 hours) and for conducting the Capacity Test. The Integrity Test evaluated the
reliability of the equipment under field conditions, while the Capacity Test produced operational
and water quality data for the system through the pre-defined arsenic breakthrough
concentration. The test also included one media regeneration and one month of post-
regeneration monitoring. Gannett Fleming provided all needed logistical support, established a
communications network, and scheduled and coordinated activities of all participants. Gannett
Fleming was responsible for ensuring that the testing location and feed water conditions were
such that the verification testing could meet its stated objectives. Gannett Fleming prepared the
-------
PSTP; oversaw the pilot testing; managed, evaluated, interpreted, and reported on the data
generated by the test; and evaluated and reported on the performance of the technology.
The Gannett Fleming field technician conducted the on-site analyses and data recording activities
during the test. Gannett Fleming's Project Manager provided oversight of the daily tests.
Contact Information:
Gannett Fleming, Inc.
P.O. Box 67100
Harrisburg, PA 17106-7100
Phone: (717) 763-7212, Ext. 2109
Fax: (717)763-1808
Contact: William Allis, Project Manager
E-mail: wallis@gfnet.com
1.2.3 Manufacturer
The treatment system is the ADI Pilot Test Unit No. 2002-09 with MEDIA G2® manufactured by
ADI International Inc. The manufacturer was responsible for supplying a field-ready arsenic
adsorption media filter system equipped with all necessary components, including treatment
equipment, instrumentation and controls, and an operations and maintenance (O&M) manual.
The manufacturer was responsible for providing logistical and technical support, as needed, as
well as providing technical assistance to the FTO during operation and monitoring of the
equipment undergoing field verification testing.
Contact Information:
ADI International Inc.
Suite 300
1133 Regent Street
Fredericton, NB E3B 3Z2 Canada
Phone: (506) 452-9000
Fax: (506) 459-3954
Contact: Michael McMullin, P.Eng.
E-mail: mjm@adi.ca
1.2.4 Analytical Laboratories
The PADEP Laboratories performed all of the laboratory water quality analyses, excluding
arsenic.
Contact Information:
PADEP Laboratories
Inorganic Services Division
Mailing Address:
P.O. Box 1467
Harrisburg, PA 17105-1467
-------
Express Mail Address:
1500 North 3rd Street
Harrisburg, PA 17102
Phone: (717) 705-2197
Fax: (717)783-1502
Contact: Ted Lyter, Inorganic Services Division Chief
E-mail: plyter@state.pa.us
Regeneration wastewater toxicity analyses were performed by:
Contact Information:
TriMatrix Laboratories, Inc.
5555 Glenwood Hills Parkway, SE
Grand Rapids, MI 49588
Phone: (616) 975-4500
Fax: (810) 220-2803
Contact: Michael W. Movinski, Vice President, Sales and Marketing
Email: mmtrimatrix@comcast.net
NSF laboratories performed all laboratory arsenic water quality analyses.
1.2.5 Pennsylvania Department of Environmental Protection
PADEP's mission is to protect Pennsylvania's air, land, and water from pollution and to provide
for the health and safety of its citizens through a cleaner environment. PADEP is the state
agency largely responsible for administering Pennsylvania's environmental laws and regulations.
Its responsibilities include: reducing air pollution; making sure Pennsylvania's drinking water is
safe; protecting water quality in Pennsylvania's rivers and streams; making sure waste is handled
properly; managing the Commonwealth's recycling programs; and helping citizens prevent
pollution and comply with the Commonwealth's environmental regulations. PADEP is
committed to general environmental education and encouraging effective public involvement in
setting environmental policy.
PADEP provided laboratory water quality analyses, excluding arsenic, and review of the test
plan and final report.
1.2.6 U.S. Environmental Protection Agency
The EPA, through its Office of Research and Development, has financially supported and
collaborated with NSF under Cooperative Agreement No. CR-82833301. This verification effort
was supported by the DWS Center operating under the ETV Program. This document has been
peer reviewed, reviewed by NSF and the EPA, and recommended for public release.
-------
1.3 Verification Testing Site
The verification test site was the Hilltown Township Water and Sewer Authority (HTWSA) Well
Station No. 1 located off Brookside Drive in Hilltown Township, Bucks County, PA. Well
Station No. 1 has a permitted capacity of 145 gallons per minute (gpm) and supplies a portion of
HTWSA's 1,065 connections, with a population served of 3,200. The frequency and duration of
Well Station No. 1 pump operation depends on the distribution system demand. HTWSA
indicates that the Well Station No. 1 cumulative daily well pump run time ranges from 8 to 20
hours per day at a flow rate of 145 gpm.
HTWSA also has two other sources of supply, Well Nos. 2 and 5. Chlorine in tie form of
sodium hypochlorite and Calciquest, a brand of polyphosphate sequestrant, are normally fed at
all three well stations.
The MEDIA G2® arsenic adsorption media filter was installed inside the Well Station No. 1
building, a masonry block building located off Brookside Drive in the Pleasant View housing
development. The building is heated to a minimum temperature of 60°F. During this test, a
continuous flow of chlorinated water from the sample tap located on the chlorine detention tank
of Well No. 1 was diverted to the MEDIA G2® arsenic adsorption media filter. Normally, water
from this sample tap would also contain two chemicals fed at the well station: sodium
hypochlorite for disinfection, and polyphosphate for sequestration and corrosion control.
However, HTWSA agreed to terminate the addition of polyphosphate for the duration of the
ETV test because sequestrants could possibly interfere with the arsenic adsorption removal
process. In addition, HTWSA has indicated that it has not observed a significant improvement in
water quality since the sequestrant feed program was initiated. The treated water from the
arsenic adsorption media filter was discharged (via the station floor drain) to an existing storm
water culvert. At the request of the PADEP, the backwash, rinse, and regenerant wastewaters
were discharged to an existing sanitary sewer adjacent to the building.
Well No. 1 operates intermittently and is controlled through the HTWSA Supervisory Control
and Data Acquisition (SCADA) system, which is interconnected with Well Station Nos. 2 and 5,
and the system's one million gallon finished water, ground-level storage tank. High and low
water level sensors in the finished water storage tank, set at 78 feet (ft) and 71 ft, respectively,
activate/deactivate the well pumps located at each well station. Booster pumps, located in the
distribution system, increase the pressure to a constant 115 pounds per square inch (psi). Prior to
the installation of the SCADA system, Well No. 1 operated off high and low pressure settings on
the hydropneumatic tank located within Station No. 1. The hydropneumatic tank has since been
converted to a chlorine detention tank; this tank no longer has any control features associated
with the well pump.
1.3.1 Feed Water
The source water for the verification test was chlorinated ground water from HTWSA's Well
Station No. 1 chlorine detention tank. The Well No. 1 source water and the treatment system
feed water were generally of poor quality, with a highly variable turbidity that averaged 0.70
Nephelometric Turbidity Units (NTU) and a very high level of hardness. The feed [sample tap
-------
#2 (ST2)] water's average manganese concentration of 140 micrograms per liter (p,g/L) is more
than two times the Secondary Standard for drinking water. The feed (ST2) water's total arsenic
concentration averaged 21 ng/L. The source water's total arsenic concentration was less than the
current maximum contaminant level (MCL) of 50 |J,g/L, but exceeds the future limit of 10 |J,g/L,
which will become effective in January 2006. According to the manufacturer's performance
objectives, a pH adjustment of the source water was required to achieve the manufacturer's
equipment operation specifications. A summary of the feed (ST2) water quality information is
presented in Table 1-1. Additional feed water quality data are presented in Chapter 4. The
source water quality appears to degrade in the on-site chlorine contact tank, as indicated by
higher concentrations of arsenic, iron, and manganese in the feed water than in the raw water.
This is discussed further in Chapter 4, Section 4.3.2.
Under normal operations, there were four sample locations during the test:
• Raw water, sampled from a tap on the well discharge pipe prior to any chemical addition
and prior to the chlorine detention tank;
• Feed water (ST1), a sample tap located immediately downstream of the chlorine injection
point and chlorine detention tank;
• Feed (ST2) water, a sample tap located downstream of the chlorine injection point,
chlorine detention tank, and sulfuric acid feed point (immediately before entering the
arsenic adsorption filter); and
• Treated water (ST3), a sample tap located immediately downstream of the arsenic
adsorption filter.
1.3.2 Pilot Filter Discharges
An existing floor drain served to collect treated water from the arsenic adsorption media filter to
avoid having to re-pump the water into the distribution system. The floor drain was piped
outside to an existing storm water culvert. The arsenic adsorption media filter backwash and
regeneration wastewaters were discharged to a sewer manhole via a garden hose that was routed
through a louvered vent in the building.
Spent media was properly disposed of in a municipal landfill. The disposal of the media for
manifesting purposes was the responsibility of the manufacturer.
Backwash and regeneration wastewater flow rate, volume, and duration were monitored for each
manually initiated backwash during the test. Backwash and rinse wastewater quality parameters
were sampled to evaluate the quantity and quality of water discharged to the sanitary sewer.
These data are presented in detail in Chapter 4.
-------
Table 1-1. Feed (ST2) Water Quality During
Number of Mean/
Parameter Units Samples Median'1'
Arsenic
PH
FAC
Temperature
Turbidity
Alkalinity
Calcium
Magnesium
Hardness
Fluoride
Sodium
Silica
Aluminum
Iron
Manganese
Chloride
Sulfate
Total
Phosphorus
Mg/L
units
mg/L
°C
NTU
mg/L as
CaCO3
mg/L as
CaCO3
mg/L as
CaCO3
mg/L as
CaCO3
mg/L
mg/L
mg/L
Mg/L
Mg/L
Mg/L
mg/L
mg/L
mg/L
121
218
218
201
203
98
29
27
29
43
41
41
41
96
96
29
29
29
21
6.43
0.85
13.2
0.70
121
254
22
272
0.30
25.1
28.5
208
180
140
36.9
155
0.011
Verification
Minimum
12
5.70
0.05
12.0
0.15
62
218
4
238
0.13
22.7
25.7
<200
47
77
36.1
111
O.010
Testing
Maximum
63
7.09
2.17
14.3
7.6
148
296
40
320
0.65
29.0
40.2
539
1120
1070
37.6
202
0.016
Standard
Deviation
8
N/A
0.24
0.22
0.65
14
20.7
11
23.1
0.08
1.41
2.28
53
158
133
0.37
17.4
0.002
95%
Confidence
Interval
19-
23
N/A
0.81-
13.1-
0.60-
117-
245-
17-
262-
0.27-
24.5-
27.7-
<200-
143-
109-
36.8-
147-
0.011-
0.89
13.2
0.80
124
264
27
282
0.33
25.6
29.4
228
217
171
37.1
163
0.012
'-1-1 The median is reported for the pH data, not the mean.
N/A = Not applicable (statistics not calculated for pH).
FAC = Free available chlorine.
-------
Chapter 2
Equipment Description and Operating Processes
2.1 Equipment Description
The equipment tested was ADI International Inc.'s arsenic adsorption media filter system. The
model tested was the ADI Pilot Test Unit No. 2002-09 with MEDIA G2®. The major system
components included: a pressure filter tank, filter media, feed water pipe, treated water pipe, raw
water sample tap, two feed water sample taps, treated water sample tap, and chemical feed
pumps. The system configuration and components are described in more detail in the following
sections.
2.1.1 Basic Scientific and Engineeri ng Concepts of Treatment
The conceptual treatment process for the arsenic adsorption media filter is based on passing
arsenic-contaminated feed water through a bed of media that has a strong affinity for arsenic.
2.1.2 Generic Arsenic Adsorption Media Description^
Arsenic occurs in water in two valence states (arsenic III and arsenic V). The toxicity of arsenic
varies, depending upon its concentration and valence. The arsenic valence state can change
while in aqueous solution, depending upon the oxidation-reduction potential of the water and/or
its pH. Consequently, the objective of arsenic removal treatment is to remove all of the arsenic
— regardless of its valence.
Adsorption is the attachment of the adsorbate (arsenic) to the surface of a porous adsorbent, such
as media grains. The adsorption capacity and effectiveness of the arsenic removal media
depends on a number of factors, such as pH, competing anions, and available sites for adsorption
on the media. An adsorptive media's surface area s a function of its available porosity. An
adsorptive treatment media contains an extensive network of fine (small diameter) pores that
extend throughout the body of a grain of media.
The arsenic ion requires time to migrate into a pore within the grain of the adsorbent. As the
surface area of each adsorbent grain becomes saturated with arsenic ions, the time required for
additional adsorption becomes longer. Other factors that determine the capacity and
effectiveness of adsorbent media are accessibility of the pore sites for arsenic ions, competing
ions for pore sites, concentration of arsenic in the feed water, pH of the feed water, and flow
characteristics of the feed water that convey the arsenic into the bed of adsorbent media.
The adsorptive media is normally in a packed bed contained in a pressure vessel. The water to
be treated typically flows in the downflow mode through the treatment bed. Gravity flow is
feasible, but if pH adjustment is employed, gravity flow is not as effective because the pressure
required to retain the carbon dioxide in solution does not occur under gravity flow conditions.
Therefore, the free carbon dioxide is released, which results in the pH rising to higher than the
'-'-' From Chapter 6 - Adsorptive Media Processes for the Removal of Arsenic of the EPA/NSF ETV Protocol for Equipment
Verification Testing for Arsenic Removal
-------
desired level. As the feed water flows through the adsorptive media, the arsenic ions are
adsorbed onto the available adsorption sites. As the water flows through the bed, the arsenic
concentration decreases until it is no longer detectable.
As the feed water continues to flow through the treatment bed, the media that first comes in
contact with the feed water becomes saturated with arsenic ions. A treatment band then
progresses through the treatment bed until breakthrough occurs. At that point, traces of arsenic
appear in the treated water. As flow continues, the treatment band progresses through the media
until the bed is saturated; the arsenic concentration in the treated water is then the same as that in
the feed water. Since the arsenic concentration in the treated water is the contaminant of
concern, it must be controlled to the desired level.
There are various methods of sequencing multiple treatment beds (parallel and/or series
arrangements), which allow use of the entire (or almost the entire) adsorptive media capacity.
When the adsorptive media becomes saturated with arsenic ions, it is removed from service for
regeneration or disposal.
Normally, the economic feasibility of the adsorptive process requires reuse of the treatment
media. This is accomplished by means of chemical regeneration, which requires adjustment of
pH (or other methods) to a level at which adsorptive conditions no longer exist. At these pH
levels, the adsorptive treatment media desorbs the adsorbate. The arsenic is released and flushed
from the adsorptive media as a high concentration arsenic wastewater.
When regeneration is complete, the pH of the media is adjusted to the desired treatment pH, at
which point the media is reused for a subsequent treatment cycle. During regeneration, some
adsorptive media may be consumed (through attrition); if that occurs, replacement adsorptive
media should be added to the treatment bed. In small treatment systems and/or in treatment
systems in which the arsenic concentration in the feed water is not excessively high, economic
feasibility might dictate replacement of spent media in lieu of regeneration.
Historically, the adsorptive media that has demonstrated the most cost-effective, reliable
performance has been granular activated alumina. Other adsorptive media, such as bone char
and synthetic bone char (tri-calcium phosphate), have also been used, but have not performed as
effectively as activated alumina.
2.1.3 Capacity
The capacities and performance of different adsorptive media vary. Some types of adsorptive
media may be capable of regeneration, while others may not. Adsorptive media that have
regeneration capability may also vary in performance during subsequent treatment runs. The
arsenic removal capacity diminishes until it is determined that adsorptive media replacement is
required. Other types of adsorptive media experience attrition during each regeneration,
requiring the addition of makeup adsorptive media prior to commencement of the next arsenic
removal treatment run. The latter type of adsorptive media may not experience reduction of
arsenic removal capacity during subsequent treatment runs.
-------
2.1.4 Intermittent Operation
In full-scale arsenic adsorptive media water treatment systems, operation may be intermittent.
The smaller the system, the higher the probability that the operation of the treatment system will
experience more frequent starts and stops.
The performance of adsorptive media is not degraded when operated on an intermittent basis. In
fact, after a shutdown of the process, the arsenic adsorption media generally exhibits a short
period of improved performance during the time immediately after the restart of treatment.
Performance then returns to the level occurring at the time of the treatment process shutdown.
2.1.5 MEDIA G2® Description
MEDIA G2® arsenic adsorption media consists of an inorganic, natural substrate (calcined
diatomite) upon which iron (ferric hydroxide) is chemically bonded. It is the iron that attracts
the metallic ions in water and binds them to the substrate by chemisorption. Although it was
developed specifically for adsorbing arsenic, ADI claims that MEDIA G2® will also adsorb iron,
manganese, zinc, cadmium, lead, copper, and uranium. The adsorption capacity for arsenic is
800 |j,g to 2,400 |j,g of arsenic per gram of media, depending on the operating pH and initial
arsenic concentration in the raw water.
The arsenic adsorption pilot unit used in this test consists of one vessel containing MEDIA G2®
adsorption media that was operated in a downflow mode. As the media becomes saturated with
arsenic, the concentration of arsenic in the treated water begins to increase. Before this
concentration reaches the pre-determined maximum allowable contaminant level (breakthrough),
the media is either replaced or regenerated on-site.
ADI has stated that MEDIA G2® can be regenerated four to five times, with a loss in capacity of
approximately 10% following each regeneration. Eventually it becomes more economical to
replace the media rather than continue to regenerate it, due to the cumulative loss in arsenic
adsorption capacity.
Previous research and pilot tests conducted by ADI have shown that MEDIA G2® systems work
well within the pH range of 5.0 to 7.5. However, the lower the pH, the better the chance for
extending the life of the media because its adsorption capacity increases with decreasing pH.
However, for most applications it is desirable to operate the system in the pH range of 6.5 to 6.8,
as indicated in the ADI Operations Manual in Appendix A.
After pH, the most critical parameter is the contact time in the adsorption vessels. The MEDIA
G2® arsenic adsorption filter is sized for a ten-minute empty bed contact time (EBCT).
MEDIA G2® is a registered trademark of ADI International Inc. and is protected by U.S. Patent
No. 6,200,482. MEDIA G2® adsorption media is certified to NSF/American National Standards
Institute (ANSI) Standard 61 for water treatment plant applications, as indicated in Appendix B.
The performance of MEDIA G2® was also verified under ETV Canada's Environmental
Technology Verification Program in March 2001.
10
-------
Table 2-1 presents information about manufacturing and procedures specific to the MEDIA G2
®
arsenic adsorption media. Table 2-2 presents the specifications of the MEDIA G2 .
Table 2-1. Manufacturing and Procedures Specific to the MEDIA G2® Adsorptive Media
Item Manufacturing/Procedures
Raw material used to make
the adsorptive media
Method of Manufacture
Preconditioning Procedure
(Pilot Test Unit No. 2002-09)
Regeneration Procedure
(Pilot Test Unit No. 2002-09)
Regeneration Results
Calcined diatomite substrate and iron (ferric hydroxide)
Chemical processes: ferric hydroxide is chemically bonded to
the calcined diatomite media (proprietary process)
Thermal processes: proprietary
Sizing/Screening methods: proprietary
Packaging methods: proprietary
Wetting requirements:
a) Place media in filter vessel and backwash at a rate of 3.2 gpm
for 30 to 60 minutes to remove fines.
b) Rinse with acidified water (pH 4.0-5.0) at a filtration rate of
1.7 gpm until pH of the filter effluent water is reduced to 6.5.
See ADI Operations Manual in Appendix A for further details.
Waste: discharged to nearby sewer easement
Backwash: see ADI Operations Manual, pages 3 and 4, in
Appendix A.
Chemical process: meter -50 gallons of 1% caustic soda to the
filter to regenerate the media, followed by neutralization of the
media by feeding 0.5% sulfuric acid solution until a filter
effluent pH of less than 7.0 occurs (but not less than 5.0). See
ADI Operations Manual, pages 6 and 7, in Appendix A, for
additional details.
Return to treatment mode: rinse following acid neutralization
until the pH of the effluent water is within one pH unit of the
pilot feed water (ADI's target pH in the effluent water is 6.5 to
6.8). Media is ready for normal operation following rinse.
Waste: see ADI Operations Manual, Appendix A, for treatment
and disposal of wastewater.
Adsorption media capacity: reduced by 10% with each
regeneration.
Number of regenerations: MEDIA G2® reportedly can be
regenerated 4 to 5 times in place. After this, it is more
economical and practical to replace the media.
Waste: see ADI Operations Manual in Appendix A for
discussion on disposal of regenerants and spent media.
11
-------
Table 2-2. MEDIA G2® Adsorptive Media Specifications
Chemical Constituents
Base material
Processed
Iron, % by weight
Physical Properties
Bulk density
Hardness
Attrition
Voids
Pore size
Pore volume
Abrasion loss
Moisture (weight)
Sieve sizes, US sieve series
Particle size
Effective size
Uniformity coefficient
Arsenic adsorption capacity
Ionic preference series
Approvals
MEDIA G2® Material Safety Data Sheets
(MSDS)
Mined calcined diatomite graded
and coated with ferric hydroxide
5 to 30
47 pounds (lbs)/cubic feet (ft3)
210 Ibs/sq in
No data available
No data available
No data available
No data available
No data available
No data available
No data available
No data available
0.32mm
1.8 to 2.0
800 - 2,400 ug arsenic per gram of media
no data available
Certified to NSF/ANSI 61 (See Appendix B)
Toxicity Characteristic Leaching Procedure
(TCLP)
See Appendix B
2.2 ADI's Arsenic Adsorption Media Test Unit Installation
ADI's filter installation at Hilltown in Well Station No. 1 was a pilot test unit, with a footprint of
less than one square foot. The pilot unit has a capacity of only 1.7 gpm, although it is
hydraulically configured and operated to simulate any size system that employs pressure filter
vessels.
The depth of media in the pilot filter is about 3 ft and the diameter of the filter is 12 inches,
which results in a hydraulic loading rate of 2.16 gpm/sf and an EBCT of 10.3 minutes. The
hydraulic capacity of a full-scale system is determined by the size and number of vessels.
Smaller systems generally use vertical pressure vessels, while larger systems [over 5 million
gallons per day (mgd)] use horizontal vessels.
The feed water was obtained from an existing tap on the well station chlorine detention lank,
located in the building. The pressure at this location was a constant 115 psi. ADI installed a
pressure-reducing valve (PRV) on the feed line to reduce pressure to the pilot filter unit to 50 psi.
An electric solenoid valve was installed on the line feeding the pilot filter unit to positively shut
off the water flow to the unit in the event of a power outage at the station. The solenoid valve
and chemical metering pumps received power from a constant 120-volt circuit. In response to a
12
-------
loss of power, the solenoid valve would have closed, to prevent the passage of water through the
pilot filter unit without pH adjustment by the addition of sulfuric acid.
An electronic, battery-operated flow meter measured the rate of flow to the test unit. Its stated
accuracy is ±1% of full scale, or ±0.05 gpm. A Neptune Trident mechanical totalizer meter
measured both the feed water supply and backwash water supply to the test unit. Because
backwash water for the test unit was supplied by the well, which contained arsenic, sampling for
arsenic in the treated water for verification purposes was done at least one hour after forward
flow was re-initiated.
HTWSA feeds sodium hypochlorite for disinfection to the raw water just upstream of the
chlorine detention tank at a dose sufficient to produce a free chlorine residual of approximately
0.8 milligrams per liter (mg/L) in the water entering the distribution system. ADI stated that the
target chlorine residual in the pilot unit effluent should be 0.7 to 1.0 mg/L. Therefore, additional
chlorine was not supplied to the feed water as part of the arsenic adsorption treatment process.
The portion of chlorinated water that was diverted to the pilot filter unit was dosed with sulfuric
acid to decrease the pH to less than 7.0 (preferably within the range of 6.5 to 6.8 specified by
ADI). Two sample taps (ST1 and ST2) were located immediately upstream and downstream,
respectively, of the acid feed point.
The test unit was supplied with two graduated chemical batch tanks and two metering pumps for
feeding sulfuric acid, and either sodium hypochlorite or caustic soda into the water. Since the
well water was already chlorinated prior to being diverted to the pilot filter unit, ADI's sodium
hypochlorite feed system served as a spare and was available for use in the event that HTWSA's
chlorine feed system was out of service; it also served to meter caustic soda during the
regeneration. The metering pumps operated at a fixed rate. They were plugged into electrical
outlets that provide continuous power to operate in conjunction with the continuous arsenic
adsorption treatment unit.
A sulfuric acid solution was added to lower the pH from the approximate raw water pH of 7.6 to
a target range of 6.5 to 6.8. Periodic adjustments of the pump speed were required to keep the
pH within the target range. The sulfuric acid usage was calculated daily from the volume
changes in the batch tank. All chemicals were metered into the pilot filter unit from diluted
solutions. (See ADI's Operations Manual in Appendix A for an example preparation of sulfuric
acid solution.)
The filter unit was fitted with inlet and outlet pressure gauges for measuring pressure drop
through the media bed. The filter was to be backwashed when the pressure drop reached 10 psi
or following four weeks of continuous operation, whichever occurred first.
Analysis of backwash water is discussed in detail in Chapter 4. ADI noted that backwash water
will always exhibit a reddish-orange color, even after hours of backwashing. ADI indicated that
this was normal and that the backwash period should not exceed the time given in the Operations
Manual. After each backwash, the filter was rinsed at the service flow rate for 15 minutes.
During the filter rinse, the reddish-orange color disappeared in the rinse water within a few
minutes.
13
-------
Treated water from the test unit was discharged to the well station floor drain, which conveys the
treated water to a stream culvert adjacent to the building. PADEP gave permission for this
discharge. Treated water samples were collected at Sample Tap 3 (ST3), as shown in Figures
2-1 and 2-2.
2.2.1 Filter System Components
The following equipment was provided by ADI, specifically for the verification test; they are not
normally included with the arsenic adsorption media filter:
• Two chemical metering pumps for metering sulfuric acid and caustic soda (for
regeneration only);
• One solenoid valve to automatically isolate the filter unit in the event of a power outage
(Well Station No. 1 does not have a back-up power supply);
• One electronic flow meter for monitoring flow rate;
• One mechanical totalizer meter for monitoring accumulative production through the filter
unit;
• Two chemical batch and feed tanks (sulfuric acid and caustic soda);
• One pressure reducing valve to lower the chlorine detention tank discharge pressure from
115 psi to 50 psi for ease in regulating the flow through the test unit;
• Two pressure gauges for measuring filter influent pressure and filter effluent pressure,
and calculating pressure drop across the filter unit; and
• Eight ball valves for manual operation of the filter unit flows, including service,
backwash, rinse, regeneration, and sampling.
2.2.2 Physical Construction of the Test Unit
The filter vessel is constructed of fiberglass, with a pressure rating of 125 psi. Rigid piping is
Schedule 80 polyvinyl chloride (PVC). Flexible piping is reinforced, clear plastic tubing.
Manually operated valves are all PVC ball valves.
2.2.3 Equipment Performance Range and Known Limitations of Equipment
ADI has stated that their MEDIA G2® ADI Pilot Test Unit No. 2002-09 requires specific water
quality conditions to minimize interference from other ions and to maximize arsenic removal.
Based on previous lab and pilot-scale tests by ADI, as well as experience with ten, full-scale
installations, ADI has made the following statements regarding interferences to MEDIA G2®:
• Adsorption capacity is not affected by chloride and sulfate ions at concentrations up to
250 mg/L. Chloride and sulfate ions in excess of 250 mg/L may reduce the MEDIA G2®
capacity for arsenic;
• MEDIA G2® does not adsorb silica when the operating pH is at or below 7.2;
• The presence of naturally-occurring iron and manganese concentrations up to 2.0 mg/L
and 0.8 mg/L, respectively, may enhance performance of MEDIA G2® for arsenic
removal, but more frequent backwashing may be required due to increased pressure drop.
Both iron and manganese will be reduced to well below the MCLs;
14
-------
• MEDIA G2® does not remove fluoride by any measure;
• Chlorine and ozone, when used as oxidants or for disinfection, reportedly have no effect
on the integrity of MEDIA G2®. Preoxidation is necessary where a portion of the arsenic
exists as arsenic III; oxidation converts arsenic III to arsenic V, which is more easily
removed by MEDIA G2®. (ADI prefers chlorine residual within the filter bed.) As to
whether an oxidant will improve the performance of MEDIA G2® directly, there are no
available data;
• MEDIA G2® will remove arsenic III and arsenic V with preoxidation of arsenic III to
arsenic V;
• Adsorption media is appropriate for groundwater not under the influence of surface
water;
• Although MEDIA G2® has performed effectively over a pH range of 5.0 to 7.5 in
previous applications, the optimal pH range for most applications is 6.5 to 6.8; and
• The manufacturer states that the process is appropriate for "smaller" systems. It is also
appropriate for "larger" systems up to 5.0 mgd.
The equipment flow range and maximum system pressure are presented in Table 2-3. The filter
tank material rating is 125 psi.
2.2.4 Drawings of Equipment
®
Figures 2-1 and 2-2 present two schematics of the MEDIA G2 Arsenic Removal Pilot Unit.
15
-------
1
*
i
D
*
1
I
tl
Ł
1
I
.
\\
s
•
'
Is
IS,!
an si
is
<
[8
* „ *
i * E l| lily
s y -B I y| | i a| |
» " K 3 iw rf 1- f IP I I « t §1 *
i! 1111 si is g i si § g S li rf S
S s I 3 B s el ul E; i BB t a K SE KB e
t Ł E
© 0® © ©© © ©
I
5 1
S S
«
0. I
a 5
VJIWS
n
a
S
-
ti
j
-
u
t
a
a
j
"
u G
i
S
0
;
=-
Figure 2-1. Schematic of MEDIA G2® Arsenic Removal Pilot Unit.
16
-------
9
is i
* *3 B
S Hb s
s SB 2E
" - S|.
a i fe «1
I 5 *^ *b
j i
A
d
J
•
4
ill
j
3* 1
If «i
11
P--
o-
i
.•:
i
c
—r-c=;
!!
ij2iBfe6Ł
slsfilsii
f>ii
H
V^ \
' ''' ' \
: .
.-> •:.''.;'•.:,:<:•>.
~*\t
& i
5 g
> i
3 3
*—MS
uj g g
fe i I -I
E > u j
i H i
s i d a
0
Figure 2-2. Schematic of MEDIA G2® Arsenic Removal Pilot Unit.
17
-------
2.2.5 Design Criteria
Table 2-3 presents design criteria for the arsenic adsorption process and appurtenances.
Table 2-3. Equipment Design Criteria
Model No. 2002 Pilot Unit MEDIA G2®
No. of Filter Units
Filter Tank Dimensions
Diameter (ID)
Height (vessel only)
Mode of Operation
Operating Capacity and Service Flow Rate
Hydraulic Loading Rate
EBCTatl.Vgpm
Initial pressure drop (clean media bed)
Filter Media
Depth
Freeboard above media
Volume
Weight
Support gravel layer
Media expansion during backwash
Filter Tank Material
Control
Backwash
Frequency criteria
Flow Rate
Duration
Rinse
Duration
Regeneration
Flow Rate
Caustic Soda Feed Rate
Caustic Soda Feed Duration
1
12 in
54 in
downflow
1.7 gpm
2.16 gpm/sf
10.3 minutes
2psi
35.4 in
12.6 in
-2.3 ft3, or -17 gal (uncompacted)
109 Ibs (based on 47 lbs/ft3)
Note: 109 Ibs of media was weighed on-site
by ADI prior to installation in the filter vessel
6 in
See backwash expansion curve in Appendix
A, Operations Manual.
Fiberglass rated at 125 psi
Manual
Pressure drop of 10 psi, or every four weeks,
whichever occurs first
3.2 gpm during initial startup
3.9 gpm after initial startup
15 minutes
-15 minutes (until rinse water is clear)
1.2 gpm (forward flow)
Rate that will allow a 1% solution to enter
the pilot unit
Time required to feed 3 bed volumes of 1%
solution
18
-------
Table 2-3. Equipment Design Criteria (Continued)
Regeneration (continued)
Neutralization
Sulfuric Acid Feed Rate
Sulfuric Acid Feed Duration
Rinse
Flow Rate
Duration
Pressure Gauges
Manufacturer
Pressure Range
Accuracy
Totalizer Meter
Manufacturer
Type
Series
Flow Meter
Manufacturer
Type
Model
Range
Accuracy and repeatedly
Power supply
Size
Enclosure
Feed Water Throttling Valve
Manufacturer
Type
Material of Construction
Size
Control
Pressure Reducing Valve
Manufacturer
Model No.
Series
Static set pressure
Adjustable pressure range
Size
Solenoid Valve
Manufacturer
Type
Maximum Service Pressure
Size
Power requirement
Rate that will allow a 0.5% solution to enter
the pilot unit
Time required to reduce the pH of the
effluent below 7.0 (ENSURE THAT THE
pH DOES NOT DROP BELOW 5.0)
1.7gpm
Time required for pH of effluent water to be
within one unit of the pH of influent water
Lyn Car
0-100 psi
±0.5%
Neptune Trident
Positive displacement
T10
Blue-White Industries, Ltd.
Electronic -impeller
F2000/RTSB-50P2-GM2
0.5 to 5.0 gpm
±1% of full scale
4 AA batteries or AC transformer
0.5 in
National Electrical Manufacturers
Association (NEMA) 4X
Hayward Industrial Products, Inc.
Ball valve
PVC
0.75 in
Manual
Conbraco Industries Inc
36C
200
50 psig
25 to 75 psig
0.75 in
Hayward Industrial Products Inc.
PVC True Union
120 psig
0.75 in
1.6 amps; 20V/AC; 50/60 Hz; 19 watts
19
-------
Table 2-3. Equipment Design Criteria (Continued)
Chemical Feed Tanks
Number
Manufacturer
Type
Volume
Dimensions
Diameter
Height
Chemical Metering Pumps
Sulfuric Acid
Manufacturer
Series
Series No.
Type
Capacity at Max Pressure
Max Pressure
Capacity at 1A Max Pressure
1A Max Pressure
Capacity at Static Set Pressure
Power Requirement
Caustic soda
Manufacturer
Series
Series No.
Type
Capacity at Max Pressure
Max Pressure
Capacity at 1A Max Pressure
1/2 Max Pressure
Capacity at Static Set Pressure
Power Requirement
AGO Container Systems
Graduated - poly
100 L
20 in
30 in
Prominent Fluid Controls
beta/4
7-BT4A1601PPE200BDO
Electronic solenoid diaphragm pump
1.1 L/hr
253 psig
1.4 L/hr
126 psig
-1.7 L/hr
0.7 amps/115 V/50-60 Hz
Prominent Fluid Controls
gamma/L
GALa 1602NPB900UD
Electronic solenoid diaphragm pump
1.4 L/hr
253 psig
1.7 L/hr
126 psig
-2.6 L/hr
0.7 amps/100-230 V/50-60 Hz
20
-------
2.2.6 Photographs of Equipment
A photograph of the equipment installed at the water treatment plant is shown in Figure 2-3.
Additional photographs are included in Appendix C.
»®
Figure 2-3. ADI International Inc. Pilot Test Unit. No. 2002-09 with Media G2 as installed
at the HTWSA Well Station No. 1.
2.2.7 Data Plate
A data plate was installed on the arsenic adsorption media filter tank that provided the following
information:
Equipment name: MEDIA G2 Arsenic Removal Unit
Model No.: 2002-09
Electrical Requirements: 120V, 60Hz for powering chemical metering pumps and
solenoid valve
Maximum pressure: 125 psi
21
-------
Manufacturer's Name and Address:
ADI International
1133 Regent Street
Suite 300
Fredericton, NB E3B 3Z2
Canada
Additional Information:
Serial Number: N/A
Service flow: 1.7 gpm, continuous or start/stop
Unit installed for NSF and EPA Environmental Technology Verification Program.
Call 506-452-9000 for more information.
Warning and Caution Statements:
Testing in progress, please do not disturb.
This unit is designed to operate with minimum and maximum inlet pressures of 30 and
125 psi, respectively.
2.3 Operating Process
This modular filter system consists of one pressurized filter vessel designed for operation in the
downflow mode. The filter does not require electricity to operate, although appurtenances
require electricity. The filter system can operate either intermittently or continuously at the
service flow rate of 1.7 gpm. Specific operating criteria are used to determine when a backwash
should be conducted. A backwash is manually initiated and operated after either four weeks of
operation or when there is a pressure drop of 10 psi across the filter, whichever comes first.
The cumulative flow and the flow rate through the filter unit were monitored with one accessory
electronic flow meter and one mechanical totalizer meter, each located on the feed side of the
unit prior to the sulfuric acid injection point. The flow meter also monitored backwash, rinse,
and regeneration water flow rates. The flow rate was throttled with a non-integral PVC ball
valve located on the treated water side of the filter unit. The collection of backwash and rinse
wastewaters for volume determination and water quality analyses occurred once per month
during the Capacity Test. The collection of regeneration wastewater for volume determination
and water quality analyses occurred once at the end of the Capacity Test. The difference in feed
water and treated water pressure readings provided loss of head across the filter unit.
Grab samples for on-site and laboratory analyses were collected from a raw water sample tap on
the well discharge pipe prior to chlorine addition (not shown on Figure 2.1), at a sample tap prior
to sulfuric acid addition (ST1), at the feed water sample tap (ST2) located immediately upstream
of the pilot filter unit, and from the treated water sample tap (ST3), located downstream of the
pilot filter unit. Samples from these taps were collected following the opening of their respective
valves and a flush period of about five seconds.
22
-------
2.3.1 Operator Requirements
The pilot filter unit operated continuously using the chlorinated well water bled from the chlorine
detention tank. Operator attention during the verification test mainly consisted of monitoring the
equipment, conducting process water quality analyses, refilling chemical tank(s), and confirming
operation in accordance with the test plan. Frequent adjustment of the sulfuric acid feed rate was
required to maintain the required feed water pH.
The system must be backwashed manually following four weeks of operation or a pressure drop
of 10 psi, whichever comes first. Operator initiation is required. The operator also manually
re-initiates service operation of the filter following a backwash period. For the purpose of the
Integrity Test, the filter was manually backwashed at the end of the two-week Integrity Test.
Each manually initiated backwash required one to two hours of operator time. Operator time
included setup, 30 minutes of backwash time, 30 minutes of rinse time, on-site water quality
analyses, sample collection for laboratory water quality analyses, documentation, and equipment
cleanup. The manually initiated backwash, monitoring, and data collection was requested by
PADEP as a special condition of the test plan and is not a general equipment operation
requirement.
Spent MEDIA G2® media must be regenerated by the operator by feeding an alkaline caustic
soda solution to the media. An increase in pH above 10 with the addition of the alkaline solution
causes the previously adsorbed arsenic to solubilize and release from the media. Following this
step, the operator is required to feed sulfuric acid to the filter to neutralize the high pH caused by
the caustic solution. Media regeneration was performed once during the Capacity Test. The
combination of filter backwash and regeneration required more than five hours of operator time,
including setup and sample collection.
2.3.2 Required Consumables
Due to the water quality chemistry of the Well No. 1 supply, the system requires chemicals
during normal treatment operations, as well as additional chemicals during regeneration. Raw or
treated water is also required for each filter backwash. The required consumables are:
• MEDIA G2® media: 2.3 ft3, or 109 Ibs (uncompacted), replaced following four
regenerations;
• Sulfuric Acid: 15 gallon carboy of 94% F^SO/i Frequency of replacement is site-specific.
At HTWSA's Well Station No. 1, the rate of acid usage was 1 gallon for about every
25,000 gallons of treated water. At this rate, the frequency of replacement for a 15-gallon
carboy of 94% H2SO4 acid is about six months;
• Sodium Hypochlorite: based on a required chlorine dose of 1.2 mg/L to achieve ADI's
target free chlorine residual of 0.7 mg/L to 1.0 mg/L, approximately 0.75 Ibs of chlorine
(0.75 gallons of 12.5% sodium hypochlorite) is required for every 75,000 gallons of
treated water;
• Caustic Soda: 50 gallons of 1% caustic soda were used during the media regeneration;
and
23
-------
• Electricity: power consumption for the metering pumps and solenoid valve.
2.3.3 Rates of Waste Production
2.3.3.1 Backwash Approximately 103 gallons of backwash wastewater and 90 gallons of rinse
wastewater were generated from each backwash. Filter backwashes are required following every
four weeks of continuous operation or following a 10 psi filter bed pressure drop, whichever
comes first. The backwash and rinse wastewater was discharged to the sanitary sewer adjacent
to the building. Backwash water quantity and water quality characteristics are described in more
detail in Chapter 4.
2.3.3.2 Regeneration. Regeneration of the spent media produced approximately 50 gallons of
regenerant wastewater. A portion of this regenerant wastewater was collected for TCLP analysis.
In addition, wastewater is generated following regeneration due to rinsing of the filter bed until
the treated water is within one pH unit of the feed water. Both of these wastewaters were
discharged to the sanitary sewer during the media regeneration performed during the Capacity
Test.
Spent MEDIA G2 media must be disposed in a manner that complies with all state and federal
regulations for ultimate waste disposal. ADI has stated that spent MEDIA G2® media is suitable
for disposal in a landfill.
2.3.4 Licensing Requirements Associated with Equipment Operation
States usually require a specific grade of waterworks operator permit in order to operate a filter
process on a public water supply. However, this requirement did not apply for the ETV since all
of the treated water was discharged to waste.
In Pennsylvania, a C9 license is required to operate a full-scale version of this treatment
technology for the Hilltown Water and Sewer Authority. "C" refers to a capacity of 1.0 mgd or
less, and "9" refers to inorganics removal.
24
-------
Chapter 3
Methods and Procedures
3.1 Experimental Design
This verification test was developed to provide verifiable information related to the performance
of the ADI International Inc. ADI Pilot Test Unit No. 2002-09. Field operations, sampling, and
analytical methodologies were performed in a manner which assured the quality of the data
collected provided an accurate evaluation of the treatment system under the field conditions.
The results of the verification test were reported in two phases. The Phase 1 Report included the
results of testing designed to evaluate the reliability of the equipment operation under the
environmental and hydraulic conditions at the Well Station No. 1 site during the initial two
weeks of testing. In addition to the Phase 1 results, this Phase 2 report includes the results of
testing designed to evaluate the capacity of the arsenic adsorption system to remove arsenic from
the feed water.
3.1.1 Objectives
The objectives of this phase of the verification testing were to:
• Produce data to meet the Data Quality Objectives shaped by the manufacturer's
performance objectives;
• Present data on the impact of variations in feed water quality such as turbidity, arsenic,
pH, silica, fluoride, iron, and manganese on equipment performance;
• Evaluate the logistical, human, and economic resources necessary to operate the
equipment;
• Evaluate the reliability, ruggedness, cost factors, range of usefulness, and ease of
operation of the equipment; and
• Evaluate the arsenic adsorption capacity of the equipment under field conditions.
3.1.2 Equipment Characteristics
3.1.2.1 Qualitative Factors. The equipment was operated in such a way as to maintain its
operating parameters within the manufacturer's recommendations. The operating range for pH is
a critical parameter for arsenic adsorption efficiency. Contact time is also a critical parameter for
arsenic adsorption efficiency and is dependent on maintaining the flow rate within the design
range. The nature and frequency of the changes required to maintain the operating conditions are
used in the qualitative evaluation of the equipment.
Frequent and significant adjustments would indicate relatively lower reliability and higher
susceptibility to environmental conditions, and also the degree of operator experience that may
be required. Note that the system appurtenances that required adjusting, such as valves and
metering pumps, are likely to be of higher quality and automatically controlled on a full- scale
installation. The effect of operator experience on the treatment results was also evaluated.
25
-------
The modular nature of the filter components, similar to a residential ion exchange water softener,
makes equipment installation easy and straightforward. The equipment can be installed by a
qualified plumber. This also makes the equipment easy to move and reinstall at another location,
if necessary. The filter unit is freestanding, requiring only a level surface capable of supporting
135 Ibs and maintenance of an ambient temperature above 40°F.
3.1.2.2 Quantitative Factors. The following factors were quantified for site-specific conditions,
based on data collected during this testing program:
• Rate of media exhaustion;
• Quantity of sulfuric acid;
• Quantity of sodium hypochlorite;
• Quantity of caustic soda;
• Frequency of media replacement;
• Backwash and rinse water quantity and quality;
• Backwash and rinse duration and frequency;
• Regenerant water quantity and quality;
• Regeneration duration and frequency;
• Estimated labor hours for operation and maintenance; and
• Chemical tank batching frequency and volume.
These quantitative factors were evaluated for the purpose of assessing equipment performance
and developing operation and maintenance cost factors.
3.1.2.3 Raw and Feed Water Quality. Well Station No. 1 finished water supplied the feed
water for the adsorption media filter unit. Raw and finished water quality data (following the
chlorine detention tank and all chemical additions) are presented in Chapter 4. The raw water is
of relatively poor quality for a groundwater supply; the arsenic concentration exceeds the
promulgated MCL and the manganese concentration exceeds the existing secondary maximum
contaminant level (SMCL). The water is high in total and calcium hardness. Sulfate levels are
moderately high in this supply, although less than the level stated by ADI as a potential limiting
factor for MEDIA G2®'s capacity for arsenic removal. A relatively high conductivity level, due
in part to the hardness and sulfate concentration, indicates a high level of dissolved ions.
The feed water (Well Station No. 1 finished water) quality was analyzed prior to testing when
the sequestrant was being added to minimize manganese precipitation; the manganese
concentration could create aesthetic problems for consumers if allowed to "plate out". HTWSA
agreed to terminate the addition of the sequestrant for the duration of the verification test, stating
that it has not provided significant improvement in water quality. The feed water had high levels
of hardness and alkalinity, slightly alkaline pH, and an arsenic concentration double the
promulgated MCL of 10 |ig/L.
The feed water turbidity was relatively high for a groundwater supply. Since the raw water
samples had generally lower turbidity, the higher feed water turbidity is likely due to the
oxidation of iron and manganese in the chlorine detention tank.
26
-------
3.2 Equipment Operations and Design
The test plan was developed based on the ETV Protocol For Equipment Verification Testing For
Arsenic Removal. Chapter 5 in the protocol, the Equipment Verification Testing Plan -
Adsorptive Media Processes for the Removal of Arsenic, includes information on the
requirements for the verification test and other documents used in the preparation of this report.
This chapter also specifies the procedures that were used to ensure the accurate documentation of
both equipment performance and treated water quality. Strict adherence to these procedures
resulted in verifiable performance of the equipment.
3.3 Field Test Equipment
Table 3-1 presents the analytical and calibration equipment that were used on-site.
Table 3-1. Field Analytical and Calibration Equipment
Equipment
Manufacturer/Model/Specs
Turbidimeter
pH/Ion Selective Electrode (ISE) Meter
Thermometer
Arsenic Field Test Kit
Colorimeter
Dead weight pressure gauge tester
Burettes for analytical titrations
Stopwatch and "bucket"
Hach Model 2100P Portable Ratio™ Optical System
(meets or exceeds EPA Method 180.1 criteria)
Orion Model 290A with Triode pH Electrode Model 91-
578N (resolution 0.1/0.01/0.001, accuracy ± 0.005); and
Fluoride Combination Electrode Model 96-09
(reproducibility ± 2%)
Miller & Weber [range 032° C; National Institute of
Standards and Technology (NIST) traceable]
Industrial Test Systems, Inc. (ITS) Model QUICK Low
Range II (optimum accuracy below 6 ug/L)
Hach Model DR/850; wavelength range 520, 610 nm;
wavelength accuracy ±1 nm
Amthor Testing Instrument Co. Inc. (Type No. 460;
range 0-6000 psi)
50 mL capacity with 0.1 mL subdivisions and 1000 mL
reagent reservoir
Digital stopwatch and 2.0 L graduated cylinder with 10
mL increments for rotameter, totalizer meters, and
control module drive water calibration checks; fifty-
gallon container for backwash wastewater flow
calibration
3.4 Communications, Documentation, Logistics, and Equipment
Gannett Fleming was responsible for coordinating communication among all verification testing
participants. All field activities were thoroughly documented in:
• Field Logbook;
27
-------
• Field Data Sheets;
• Photographs;
• Laboratory Chain-of-Custody Forms, Submission Sheets, and Reports; and
• Laptop Computer.
Gannett Fleming was responsible for maintaining all field documentation. A bound field logbook
was used to record all water treatment equipment operating data. Each page was sequentially
numbered and labeled with the project name and number. Completed pages were signed and
dated by the individual responsible for the entries. Errors have one line drawn through them and
are initialed and dated.
Laboratory submission forms accompanied all samples shipped to the PADEP and NSF
laboratories. Copies of laboratory submission forms or chain-of-custody forms for all samples
were provided at the time of the quality assurance/quality control (QA/QC) inspection and are
included in the verification report appendices.
3.5 Equipment Operation and Water Quality Sampling for Verification Testing
The field activities conformed to requirements included in the PSTP that was developed and
approved for this verification testing. The sampling and sample analyses that occurred during
this verification testing program were performed according to the procedures detailed by Gannett
Fleming in the PSTP.
Any unanticipated or unusual situations that altered the plans for equipment operation, water
quality sampling, or data quality were discussed with the NSF technical lead and PADEP. Any
deviations from the approved final PSTP were documented.
During routine operation, the following were documented daily:
• Number of hours the arsenic adsorption media filter was operated;
• Number of hours the operator was working at tasks at the well station related to the
operation of the arsenic adsorption media filter; and
• Description of tasks performed during arsenic adsorption media filter operation.
3.6 Recording Data
The following information was recorded on-site:
• Experimental run number;
• Water type (raw, feed—prior to and after sulfuric acid addition—and treated);
• Wastewater type (backwash, rinse, regenerant);
• Hours of operation (since previous monitoring period) and total hours;
• Feed water flow rate;
• Feed water total production;
• Feed water pressure;
• Treated water pressure;
28
-------
• Feed water temperature;
• Treated water temperature;
• Raw, feed, and treated water turbidity;
• Raw, feed (prior to and after sulfuric acid addition), and treated water pH;
• Raw, feed (prior to and after sulfuric acid addition), and treated water chlorine residual;
• Raw, feed, and treated water arsenic concentration (qualitatively with field test kit);
• Raw, feed, and treated calcium, magnesium, and hardness;
• Raw, feed, and treated alkalinity;
• Raw, feed, and treated fluoride;
• Occurrence of a backwash;
• Backwash water flow rate;
• Backwash duration;
• Total volume of backwash wastewater;
• Rinse water flow rate;
• Rinse duration;
• Total volume of rinse water;
• Occurrence of a regeneration;
• Regeneration water flow rate;
• Regeneration duration;
• Total volume of regenerant water;
• Sulfuric acid metering pump rate;
• Sulfuric acid dose;
• Sulfuric acid tank level;
• Caustic soda metering pump rate during regeneration;
• Caustic soda dose during regeneration; and
• Caustic soda drum level during regeneration.
3.7 Recording Statistical Uncertainty for Assorted Water Quality Parameters
For the analytical data obtained during verification testing, 95% confidence intervals were
calculated by Gannett Fleming for arsenic data and for all other water quality data where the
sample set contained eight or more values, with the exception of pH.
The consistency and precision of water quality data were evaluated with the use of the
confidence interval. A confidence interval describes a population range in which any individual
population measurement may exist with a specified percent confidence. The following formula
was employed for confidence interval calculation:
confidence interval = X ± tn -1, \.-
where: X is the sample mean;
S is the sample standard deviation;
n is the number of independent measures included in the data set;
t is the t distribution value with n-1 degrees of freedom; and
a is the significance level, defined for 95% confidence as: 1 - 0.95 = 0.05.
29
-------
According to the 95% confidence interval approach, the a term is defined to have the value of
0.05, thus simplifying the equation for the 95% confidence interval in the following manner:
95% confidence interval = X + tn - 1,0.975 (s /Jn)
Results of these calculations were expressed as the sample mean plus or minus the width of the
confidence interval.
3.8 Verification Testing Schedule
Verification testing activities included equipment set up and shakedown, equipment integrity
verification testing, and adsorptive capacity testing. The test schedule was developed to
encompass all of these activities.
The Integrity Test began on October 8, 2003. The Integrity and Capacity Tests were initiated
simultaneously. The Integrity Test ran for a two-week (13 full days plus 8 hours) period, ending
October 21, 2003. The Capacity Test continued to run until May 28, 2004.
3.9 Task 1: System Integrity Verification Testing
3. 9. 1 Introduction
During Task 1, Gannett Fleming evaluated the reliability of the equipment operation under the
environmental and hydraulic conditions at the HTWSA Well Station No. 1 site. The adsorption
media filter was operated for Integrity Test purposes within the operational range presented in
the equipment design criteria.
3. 9. 2 Experimental Objectives
The experimental objectives for the Integrity Test phase of the verification testing were to:
• Evaluate equipment operational reliability under field conditions;
• Document feed water quality and arsenic concentration; and
• Collect operational and water quality data under field conditions.
3.9.3 Work Plan
The vendor used a platform scale to weigh the media prior to installation into the filter vessel.
ADI's procedure for media replacement is included in Appendix D. The installed weight of the
media is specified in Table 2-3. The protocol for media conditioning includes placing media in
the filter vessel and backwashing at a rate of 3.2 gpm for 30 to 60 minutes to remove fines, and
rinsing with acidified water (pH 4.0 to 5.0) at the normal service flow rate (1.7 gpm) until the pH
of the filter outlet water is reduced to less than 6.5 (See Appendix A, ADI Operations Manual,
pages 1 to 4, for additional details).
30
-------
The monitoring and on-site data collection schedules are presented in Tables 3-2 and 3-3. Grab
samples for on-site and laboratory analyses were collected based on the sampling schedule
presented in Table 3-3. The raw water, feed (ST1), feed (ST2), and treated (ST3) water sample
taps were flushed for at least five seconds prior to sample collection. A sampling plan for
arsenic that includes the Integrity Test is presented in Table 3-4. Seven days of the daily raw,
feed, and treated water arsenic samples were speciated during the Integrity Test, as specified in
Table 3-4. The protocol for arsenic speciation is presented in Appendix E. Daily and weekly
samples collected for on-site analysis were analyzed immediately after collection. Sample
collection and handling procedures followed Standard Methods 3010 B.
Daily and weekly samples were collected for laboratory analysis during the daily two-hour
monitoring period. All of the samples were collected by the Gannett Fleming field technician in
appropriate sample bottles prepared with preservatives, as required, specific to the analytical
methods to be used. The samples were stored and shipped in accordance with appropriate
procedures and holding times, as specified by the PADEP and NSF laboratories. The methods
used for on-site and laboratory analytical procedures are presented in Table 3-3. The water
quality sampling protocol is described in Section 3.13.5.
One backwash was manually initiated by the field engineer at the end of the system Integrity
Test. Backwash water flow rate, duration, volume, and water quality were monitored following
the methods and schedule presented in Table 3-5.
3.9.4 A nalytical Schedu le
The arsenic adsorption media filter system operational data was monitored following the
procedures and at the frequencies prescribed in the test plan, as summarized below and in Table
O O
3-2.
• Operational Data Collection
o Feed water production was monitored twice per day at the mechanical totalizer
meter, located on the feed water pipe;
o Feed water flow rate was monitored twice per day and adjusted, as needed, with
the flow meter and ball valve located on the treated water pipe. Flow rate was
recorded twice per day, before and after adjustment. The flow rate was set at 1.7
gpm±0.1 gpm;
o Feed water pressure was monitored and recorded twice per day at the pressure
gauge located on tie feed water pipe. The specified minimum and maximum
operating pressures for the filter vessel are 30 psi and 125 psi, respectively. A
PRV maintained the feed water pressure at approximately 50 psi during the test;
o Treated water pressure was monitored and recorded twice per day at the pressure
gauge located on the treated water pipe. This was performed at the same time as
the feed water pressure measurement. The difference between these
measurements represents the pressure drop through the pilot filter unit;
o The sulfuric acid chemical batch tank level was checked and recorded daily (see
Table 3-2). The tank was refilled, as needed, with the time and quantity of refill
noted; and
-------
o The sulfuric acid metering pump feed rate was monitored and adjusted, based on
the draw down in the batch tank, to maintain ADI's goal of sustaining the pH as
close as possible to 6.5 in the feed (ST2) and treated (ST3) water.
Table 3-2. On-site Equipment Operating Parameter Monitoring and Data Collection
Schedule
Parameter
Monitoring Frequency
Monitoring Method
Feed water production
Feed water flow rate
Feed water pressure
Treated water pressure
Chemical feed:
tank volume and pump
metering rate
Chemicals used
Check and record twice per day
Check and record twice per day
(adjust when 5% above or below
target; record before and after
adjustment)
Check and record twice per day
Check and record twice per day
Check and record once per day
As needed
Feed water totalizer meter
Feed water flow meter
Feed water pressure gauge
Treated water pressure
gauge
Measure with measuring
tape depth of chemical
remaining and, as required,
quantity of chemical refill
Record name of chemical,
supplier, commercial
strength, dilution used for
making batch solution
Water Quality Data Collection
o
o
o
o
The water quality of the raw water, feed water (both prior to and after sulfuric
acid addition), and treated water was characterized by analysis of the water
quality parameters listed in Tables 3-3 and 3-4;
Samples were collected during the two-hour monitoring period;
All "on-site analyses" were analyzed on-site; and
The water quality analyses presented in Tables 3-3 and 3-4 were conducted to
provide state drinking water regulatory agencies with background data on the
quality of the raw water being treated and the quality of the feed and treated
water.
32
-------
Table 3-3. Water Quality Sampling Schedule - System Integrity Verification Testing
Parameter
On-site Analyses
Arsenic
Sampling
Frequency
(4)
Test Streams Sampled
Raw Water, Adsorptive Media
Standard EPA Hach
Method^ Method^ Method3'
(See Appendix F)
pH
Temperature
Turbidity
Alkalinity
Calcium
Magnesium
Hardness
Fluoride
FAC
Laboratory Analyses
Feed and Treated Water
Twice Daily Raw Water, ST 1 , Adsorptive Media
Feed and Treated Water
Daily Adsorptive Media
Feed and Treated Water
Daily Raw Water, ST 1(5), Adsorptive Media
Feed and Treated Water
Daily Raw Water, Adsorptive Media
Feed and Treated Water
Weekly Raw Water, Adsorptive Media
Feed and Treated Water
Weekly Raw Water, Adsorptive Media
Feed and Treated Water
Weekly Raw Water, Adsorptive Media
Feed and Treated Water
Daily Raw Water1-6-1, Adsorptive Media
Feed and Treated Water
Twice Da ily Raw Water, ST 1(5), Adsorptive Media
Feed and Treated Water
4500-H+ B
2550 B
2130 B
4500-F" C
8221
8222
Calculated
(8226-
8222)
8226
8021
Arsenic ^ '
Silica
Aluminum
Iron
Manganese
Chloride
Sulfate
Sodium
Total Phosphorus
Daily
Daily
Daily
Daily
Daily
Weekly
Weekly
Weekly
Weekly
Raw Water, Adsorptive Media
Feed and Treated Water
Raw Water , Adsorptive Media
Feed and Treated Water
Raw Water(6), Adsorptive Media
Feed and Treated Water
Raw Water, Adsorptive Media
Feed1-8-1 and Treated Water
Raw Water-6-1, Adsorptive Media
Feed(8) and Treated Water
Raw Water, Adsorptive Media
Feed and Treated Water
Raw Water, Adsorptive Media
Feed and Treated Water
Raw Water, Adsorptive Media
Feed and Treated Water
Raw Water, Adsorptive Media
Feed and Treated Water
200.8
200.7
200.7
200.7
200.7
300.0
300.0
200.7
365.1
APHA, AWWA and WPCF (1998). Standard Methods for Examination of Water and Wastewater. 20th ed. Washington, D.C.
APHA.
EPA Methods Source: EPA Office of Ground Water and Drinking Water. EPA Methods are available from the National Technical
Information Service (NTIS).
Hach Water Analysis Handbook (1992). Hach Company, Loveland, Colorado.
See Table 3-4. An arsenic field test kit was used for periodic qualitative arsenic checks.
A sample was collected once per day from ST1, the sample tap located on the chlorinated water, prior to sulfuric acid addition.
The raw water was collected and analyzed weekly.
The NSF laboratory performed laboratory arsenic analyses. The PADEP Laboratory analyzed all other laboratory analyses during
the Integrity Test.
The feed water was checked for the soluble fractions of iron and manganese, in addition to the total concentrations of these metals, requiring
filtration through 0.22 um filter paper.
33
-------
Table 3-4. Arsenic Sampling Plan
Test Period
Laboratory Analyses
Integrity Test
Capacity Test
Capacity Test
Post-Regeneration
A7cvnfi r»ci"h rvn
Sample
Sources
raw, feed,
treated
raw(2),feed,
treated
raw (2), feed,
treated
raw(2),feed,
treated
Sample
Frequency
daily
weekly
daily
3x/week(4);
weekly
Sampling
Period
13 days
8 hours
first
6 months(1)
final
2 months(1)
one 4 week
period
No. of Days
Samples
Speciaterf1'
7
monthly(3)
monthly(3)
lx(3)
Hold
Samples
none
none
12 per
week
N/A
Total No.
Analyses
84
84
72
19
On-site Qualitative Analyses^
Integrity Test
Capacity Test
Capacity Test
Post-Regeneration
Verification
feed, treated
feed, treated
feed, treated
feed, treated
weekly
weekly
3xper
week
weekly
13 days
8 hours
first 6
months(1)
final 2
months(1)
one 4 week
period
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
4
48
48
8
(3)
The sampling period was based on the manufacturer's performance objectives.
Three samples per week, if after the results of daily sampling during the Integrity Test indicated that raw and
feed total arsenic levels may vary significantly.
This was considered the minimum number of samples speciated; if arsenic results not anticipated occurred, such
as premature breakthrough or significant variation in feed arsenic level, more frequent arsenic speciation would
have occurred.
For the first week, weekly thereafter.
Method procedure presented in Appendix F.
34
-------
Table 3-5. Backwash Wastewater and Rinse Water Monitoring, Sampling, and Analyses
Backwash and Rinse
Wastewater Monitoring or
Parameter Sample Type Frequency Method
Flow Rate
Volume
Duration
Turbidity
pH
Arsenic
Manganese
Iron
Aluminum
Sodium
Alkalinity
FAC
yes
yes
yes
grab(1)
grab(1)
grab(1)
grab(1)
grab(1)
grab(1)
grab(1)
grab(1)
grab(1)
every backwash
every backwash
every backwash
monthly
monthly
monthly
monthly
monthly
monthly
monthly
monthly
monthly
"bucket"(2) and
stopwatch
graduated
container®
stopwatch
SM 2130-B
SM 4500-H+
EPA 200. 8
EPA 200.7
EPA 200.7
EPA 200.7
EPA 200.7
Hach 8221
Hach 8021
( ' Grab samples were collected from a continuously mixed batch tank using a 2-liter beaker. All wastewaters
were collected in a 50-gallon container.
^ The "bucket" was a 50-gallon container for calibrating backwash and rinse flow rates. Increments in liters were
marked on the sides of this container, based on incrementally filling the container beforehand with a 2-liter
graduated cylinder.
3.9.5 Evaluation Criteria and Minimum Reporting Requirements
Tables and time series plots were produced to present all feed water and treated water quality
data that varied over time during the Integrity Test. These plots are presented in Chapter 4. The
Integrity Test was performed to demonstrate the initial ability of the adsorptive media to reduce
the feed water arsenic concentration in the treated water. All water quality parameters,
operational parameters, backwash flow rates, and quantities were also tabulated and/or plotted,
as appropriate. A plot of feed and treated water pressure and system headloss is presented in
Chapter 4. System headloss information may be used to infer power requirements for a system
that pumps directly through the treatment unit. No direct measurement of power was possible,
since the system does not directly require electricity.
3.10 Task 2: Adsorptive Capacity Verification Testing
3.10.1 Introduction
The objectives of the Capacity Test were to produce quality operational and water quality data
until what ADI has defined as the breakthrough arsenic level for their arsenic adsorption system.
The performance of the adsorptive media is a function of the feed water quality, contact time,
rest time, and type of adsorptive media used. Arsenic breakthrough is highly dependent on the
concentration and adsorptive characteristics (isotherm) of the arsenic to be treated by the
adsorptive media. Design and EBCT helps define the performance of a given media for a given
feed water quality.
35
-------
The Capacity Test was performed once through arsenic breakthrough. Following breakthrough,
MEDIA G2 was regenerated and returned to operation for four weeks of continuous operation.
Regeneration was accomplished by first performing a backwash of the pilot unit, followed by
chemical regeneration of MEDIA G2® by passing three bed volumes (approximately 50 gallons
total) of 1% caustic soda through the filter bed. Subsequently, the filter bed was neutralized with
a 0.5% sulfuric acid solution. Details of the regeneration procedure are provided in Section 6.0
of ADI's Operations Manual (see Appendix A).
3.10.2 Experimental Objectives
The experimental objective was to provide equipment operating and water quality data relative to
the adsorptive media capacity to remove arsenic from the feed water.
3.10.3 Work Plan
Task 2 (Adsorption Capacity Verification Testing) began simultaneously with Task 1 (System
Integrity Verification Testing). The operating conditions were as stated under the Work Plan for
Taskl (Section 3.9.3).
3.10.4 AnalyticalSchedule
• Operational Data Collection
o Feed water production was monitored twice per day at the mechanical totalizer
meter, located on the feed water pipe;
o Feed water flow rate was monitored twice per day at the flow meter and adjusted,
as needed, with the ball valve located on the treated water pipe. Flow rate, before
and after adjustment, was recorded twice per day. The flow rate was set at 1.7
gpm±0.1 gpm;
o Feed water pressure was monitored and recorded twice per day at the pressure
gauge located on the feed water pipe. Minimum and maximum operating
pressures for the filter vessel are 30 psi and 125 psi, respectively. A PRV
maintained the feed water pressure at approximately 50 psi;
o Treated water pressure was monitored and recorded twice per day at the pressure
gauge located on the treated water pipe. This was performed at the same time as
the feed water pressure measurement. The difference between these
measurements represents the pressure drop through the pilot filter unit;
o The sulfuric acid chemical batch tank level was checked and recorded daily. The
tank was refilled as needed, with the time and quantity of refill noted;
o The sulfuric acid metering pump feed rate was monitored and adjusted based on
the treated water pH to maintain ADI's goal of sustaining the pH within the range
of 6.5 to 6.8 in the treated water; and
o Fluoride, silica, and aluminum samples were collected weekly during Task 2.
More frequent sampling of these parameters would have occurred if data collected
daily during Task 1 showed that the concentration of the parameter fluctuated or
was at a higher concentration than expected.
36
-------
Water Quality Data Collection
o The water quality of the raw water; adsorptive media feed and treated water (pre-
and post-regeneration); backwash wastewater; rinse wastewater; and regeneration
wastewater were characterized by the analysis of the water quality parameters
listed in Tables 3-2, 3-4, 3-5, 3-6, 3-7, and 3-8. The sampling frequency was
intended to provide sufficient water quality data to effectively characterize the
breakthrough profile of arsenic, to develop representative backwash and
regenerant wastewater quality profiles, and to produce quality operational and
water quality data for a minimum of four weeks of continuous operation
following regeneration of the media.
o Grab samples of backwash, rinse, and regenerant wastewaters were collected for
the water quality analyses at the frequency presented on Table 3-5 and 3-8. The
wastewaters were collected separately; each was mixed to maintain a relatively
homogenous suspension during sample collection.
37
-------
Table 3-6. Water Quality Sampling Schedule - Media Adsorption Capacity Verification
Testing
Parameter
Sampling
Frequency
Test Streams Sampled
Standard EPA Hach
Method00 Method'2' Method'3'
On-Site Analyses
Arsenic w
pH Daily
Temperature Daily
Turbidity Daily
Alkalinity 3/Week
Calcium Weekly
Magnesium Weekly
Hardness Weekly
Fluoride Weekly
FAC Daily
Laboratory Analyses
Adsorptive Media
Feed and Treated Water
Raw Water, ST1(5),
Adsorptive Media Feed and
Treated Water
Adsorptive Media
Feed and Treated Water
Raw Water, ST1(5),
Adsorptive Media Feed and
Treated Water
Raw Water, Adsorptive
Media
Feed and Treated Water
Adsorptive Media
Feed and Treated Water
Adsorptive Media
Feed and Treated Water
Adsorptive Media
Feed and Treated Water
Adsorptive Media
Feed and Treated Water
ST1( , Adsorptive Media
Feed and Treated Water
(See Appendix F)
4500-H+ B
2550 B
2130 B
8221
4500-F'C
8222
Calculated
(8226-
8222)
8226
8021
Arsenic (6)
Silica
Aluminum
Iron
Manganese
Chloride
Sulfate
Sodium
Total Phosphorus
Weekly(/)
Weekly
Weekly
Weekly
Weekly
Weekly
Weekly
Weekly
Weekly
Raw Water, Adsorptive
Media
Feed and Treated Water
Adsorptive Media
Feed and Treated Water
Adsorptive Media
Feed and Treated Water
Adsorptive Media
Feed and Treated Water
Adsorptive Media
Feed and Treated Water
Adsorptive Media
Feed and Treated Water
Adsorptive Media
Feed and Treated Water
Adsorptive Media
Feed and Treated Water
Adsorptive Media
Feed and Treated Water
200.8
200.7
200.7
200.7
200.7
300.0
300.0
200.7
365.1
APHA, AWWA and WPCF (1995). Standard Methods for Examination of Water and Wastewater. 19th ed.
Washington, D.C. APHA.
EPA Methods Source: EPA Office of Ground Water and Drinking Water. EPA Methods are available from the
National Technical Information Service (NTIS).
Hach Water Analysis Handbook (1992). Hach Company, Loveland, Colorado.
See Table 3-4. An arsenic field test kit was used for periodic qualitative arsenic checks.
A sample was collected three times per week from ST1, the sample tap located on the chlorinated water, prior to
sulfuric acid addition.
The NSF laboratory performed laboratory arsenic analyses.
See arsenic sampling plan in Table 3-4.
38
-------
3.10.5 Evaluation Criteria and Minimum Reporting Requirements
The results of the Capacity Test are presented in Chapter 4 and include the following:
• Record of Arsenic Removal
o An arsenic breakthrough curve showing adsorptive media treated water
concentrations versus bed volumes treated was plotted. Feed water arsenic
concentrations were included on the same plot.
o A spreadsheet table tabulating arsenic feed water concentrations and the average
feed water arsenic concentration.
• Process Control
o The adsorptive media feed water and treated water arsenic, pH, FAC, pressure,
water production, and flow rate were tabulated and used to calculate incremental
feed and treated water production, differential pressure, and cumulative arsenic
removed. The average, standard deviation, and confidence interval were included
for each parameter when appropriate.
• Record of Chemical Consumption
o Gallon(s) of chemicals consumed per 1,000 gallons of treated water were
calculated. The calculated data aids in generating operating cost factors for the
treatment system.
39
-------
Table 3-7. Water Quality Sampling Schedule - Post-Regeneration Media Verification
Testing
Sampling Standard EPA Hach
Parameter Frequency Test Streams Sampled Method1' Method2' Method3'
On-Site Analyses
Arsenic (4)
pH Daily(5)
Temperature Daily
Turbidity Daily
Alkalinity 3/Week
Calcium 3/Week
Magnesium 3/Week
Hardness 3/Week
Fluoride 3/Week^
FAC Daily
Laboratory Analyses
Arsemc(7) 3/Week(8)(9)
Silica 3/Week00
Aluminum 3/Week00
Iron 3/Week(8)(5)
Manganese 3/Week-8-1
Chloride 3/Week00
Sulfate 3/Week00
Sodium 3/Week00
Total 3/Week00
Phosphorus
Adsorptive Media Feed and Treated
Water
Raw Water, ST1(6), Adsorptive Media
Feed and Treated Water
Adsoiptive Media Feed and Treated
Water
Adsorptive Media Feed and Treated
Water
Raw Water, Adsorptive Media Feed and
Treated Water
Adsorptive Media Feed and Treated
Water
Adsorptive Media Feed and Treated
Water
Adsorptive Media Feed and Treated
Water
Adsorptive Media Feed and Treated
Water
ST1(6), Adsorptive Media Feed and
Treated Water
Raw Water, Adsorptive Media Feed and
Treated Water
Adsorptive Media Feed and Treated
Water
Adsorptive Media Feed and Treated
Water
Adsorptive Media Feed and Treated
Water
Adsorptive Media Feed and Treated
Water
Adsorptive Media Feed and Treated
Water
Adsorptive Media Feed and Treated
Water
Adsorptive Media Feed and Treated
Water
Adsorptive Media Feed and Treated
Water
(See Appendix F)
4500-Ff B
2550 B
2130 B
8221
8222
Calculated
(8226-8222)
8226
4500-F' C
8021
200.8
200.7
200.7
200.7
200.7
300.0
300.0
200.7
365.1
~ APHA,AWWA and WPCF (1995). Standard Methods for Examination of Water and Wastewater. 19th ed. Washington,
B.C. APHA.
(2) EPA Methods Source: EPA Office of Ground Water and Drinking Water. EPA Methods are available from the National
Technical Information Service (NTIS).
(3) Hach Water Analysis Handbook (1992). Hach Company, Loveland, Colorado.
^ See Table 3-4. An arsenic field test kit was used for periodic qualitative arsenic checks.
(5) Samples were collected 3x during the first 12 hours following regeneration.
(6) A sample was collected three times per week from ST1, the sample tap located on the chlorinated water, prior to sulfuric
acid addition.
^ The NSF laboratory performed laboratory arsenic analyses. The PADEP Laboratory analyzed all other laboratory analyses
during the post-regeneration Test.
(8) During the first week following regeneration, followed by weekly sampling thereafter.
^ See arsenic sampling plan in Table 3-4.
40
-------
Table 3-8. Regeneration Wastewater Monitoring, Sampling and Analyses
Backwash, Rinse and
Regeneration Wastewater
Monitoring or
Parameter
Flow Rate
Volume
Duration
Volume of Caustic
Soda
Volume of
Sulfuric Acid
Turbidity
pH
Arsenic
Manganese
Iron
Aluminum
Sodium
Alkalinity
FAC
TCLP(5)
Sample Type
yes
yes
yes
yes
VPS
J VO
grab(2)
grab(3)
grab(4)
grab(2)
grab(2)(4)
grab(2)(4)
grab(2)
grab(2)
grab(2)
grab(2)
Frequency
Ix
Ix
Ix
Ix
Ix
Ix
Ix
Ix
Ix
Ix
Ix
Ix
Ix
Ix
Ix
Method
"bucket"(1) and
stopwatch
Graduated
container
Stopwatch
Graduated
container
Graduated
container
SM 2130-B
SM 4500-H+
EPA 200. 8
EPA 200.7
EPA 200.7
EPA 200.7
EPA 200.7
Hach 8221
Hach 8021
SW-846
EPA 1311
- ' The "bucket" was a 50-sallon container for calibrating backwash, rinse and regeneration flow rates, and th
(3)
(4)
(5)
volume of caustic soda fed. Increments in liters were marked on the sides of this container based on
incrementally filling the container beforehand with a 2-liter graduated cylinder.
Grab samples were collected from a continuously mixed batch tank using a 2-liter beaker. All wastewaters
were collected in a 50-gallon container.
Samples for pH analysis were collected every 5 minutes during the regeneration rinse to evaluate the efficiency
of media pH adjustment and to assure that the media is conditioned to within the pH "window" specified by
ADI.
Samples for laboratory analysis of arsenic, iron and aluminum were collected every 10 minutes during the
regeneration and rinse to evaluate the efficiency of regeneration.
TriMatrix Laboratories, Inc. performed the TCLP analyses.
3.11 Task 3:
Performance
Documentation of Operating Conditions and Treatment Equipment
3.11.1 Introduction
During each day of verification testing, arsenic adsorption media filter operating conditions were
documented, including the rate of head loss gain. The volumetric flow rate through an
adsorptive media vessel is a critical parameter, and must be thoroughly monitored and
documented. Adsorptive media performance is affected by the EBCT, which varies directly with
the volumetric flow rate through the vessel.
41
-------
3.11.2 Experimental Objectives
The objective of this task was to accurately and fully document the operating conditions and
performance of the equipment.
3.11.3 Work Plan
During each day of verification testing, treatment equipment operating parameters were
monitored and recorded on a routine basis. This included documenting a complete description of
all applicable data.
3.11.4 Schedule
Table 3-9 presents the schedule that was followed for observing and recording equipment
operation and performance data.
Table 3-9. Schedule for Observing and Recording Equipment Operation and Performance
Data
Operational Parameter Action
Feed water flow rate Check and record in logbook twice per day, adjust when
>5% above or below target. Record before and after
adjustment.
Filter system feed water and treated water Record in logbook initial clean bed feed water and
pressures treated water pressure at the start of the run, and
thereafter record twice per day.
Tasks performed during equipment Record in logbook tasks performed on a daily basis.
operation
Number of hours per day operator attends Record number of hours required by operator to
to all tasks related to the treatment process accomplish all tasks.
Totalizer meter readings Record totalizer meter readings twice daily.
3.11.5 Evaluation Criteria
The data developed from this verification test were used to evaluate the performance of the
adsorption media filter. An objective evaluation of the difficulty of operations was based on the
assessment of time required for process monitoring and hydraulic control.
3.12 Task 4: Data Management
3.12.1 Introduction
The data management system that was used in this verification involved computer spreadsheet
software and manual recording of system operating parameters.
42
-------
3.12.2 Experimental Objectives
The objective of this task was to establish a viable structure for the recording and transmission of
field testing data by Gannett Fleming, such that sufficient and reliable data were produced fir
verification purposes.
3.12.3 Work Plan
The following procedures were implemented for data handling and data verification by Gannett
Fleming. The field testing operator recorded operating and water quality data and calculations by
hand in a laboratory logbook, using the following protocol:
• All daily measurements were recorded in the logbook;
• The logbook was permanently bound with consecutively numbered pages;
• The logbook indicated the starting and ending dates that apply to entries in the logbook;
• All pages had appropriate headings to avoid entry omissions;
• All logbook entries were made in black water-insoluble ink;
• All corrections in the logbook were made by drawing one line through the erroneous
information and were initialed by the field testing operator; and
• Pilot operating logs included a description of the adsorptive media equipment, description
of test run(s), names of visitors, and a description of any problems or issues, etc. Such
descriptions were provided in addition to experimental calculations and other items.
The original logbook was stored on-site. The logbook was photocopied at least once per week
and copies forwarded to the Gannett Fleming project engineer. This protocol not only eased
referencing the original data, but offered protection of the original record of results.
The database for this verification testing program was set up in the form of custom-designed
spreadsheets. The spreadsheets were capable of storing and manipulating each monitored water
quality and operational parameter from each task, each sampling location, and each sampling
time. All data from the laboratory notebooks and data log sheets were entered into the
appropriate spreadsheets. Data entry was conducted off-site by the designated data-entry
technician. All recorded calculations were also checked at this time. Following data entry, the
spreadsheet was printed out and the printout was checked against the handwritten data sheet by
another individual. Any corrections were noted on the hard copies and corrected on the screen,
and then a corrected version of the spreadsheet was printed out. Each step of the verification
process was initialed by the field testing operator or supervisor performing the entry or
verification step.
Each experiment (i.e. each test run) was assigned a run number that was then tied to the data
from the experiment through each step of data entry and analysis. As samples were collected
and sent to the PADEP and NSF laboratories, the data were tracked by use of a system of run
numbers. Data from the PADEP and NSF laboratories was received and reviewed by the field
testing operator. These data were entered into the data spreadsheets, corrected, and verified in
the same manner as the field data.
43
-------
3.13 Task 5: Quality Assurance/Quality Control
3.13.1 Introduction
Quality assurance and quality control for the operation of the arsenic adsorption media filter and
the measured water quality parameters was maintained during the verification testing program,
as described in this section.
3.13.2 Experimental Objectives
The objective of this task was to maintain strict QA/QC methods and procedures during this
verification. Maintenance of strict QA/QC procedures was important in that if a question arose
when analyzing or interpreting data collected for the arsenic adsorption media filter, it would be
possible to verify exact conditions at the time of testing.
3.13.3 Work Plan
Equipment flow rates were verified and equipment operation verification was recorded on a
routine basis. A routine daily walk-through during testing was established to verify that each
piece of equipment or instrumentation was operating properly. The items listed below are in
addition to any specified checks outlined in the analytical methods.
It was extremely important that system flow rates were maintained at set values and monitored
frequently. Doing so allowed a constant and known EBCT to be maintained in the adsorptive
media. Adsorptive media performance is directly affected by the EBCT, which in turn is
proportional to the volumetric flow rate through the media. Therefore, an important QA/QC
objective was the maintenance of a constant volumetric flow rate through the adsorptive media
by frequent monitoring and documentation. Documentation included calculating an average and
standard deviation of recorded flow rates through the adsorptive media.
The following weekly QA/QC checks were performed by the field testing operator to assure
representative data:
• In-line rotameter (cleaned any foulant buildup, as needed, and verified flow rate
volumetrically);
• In-line totalizer meter (cleaned any foulant buildup, as needed, and verified the
production rate volumetrically);
• Tubing (verified good condition of all tubing and connections, replaced as necessary);
and
• Chemical fenk volumes (calculated the change for weekly time of filter operation to
confirm calibration of each metering pump).
3.13.4 Analytical Methods
The analytical methods utilized in this study for on-site and laboratory monitoring of raw water,
and adsorptive media feed and treated water quality, are summarized in Tables 3-3 and 3-5.
44
-------
Arsenic analyses were the most critical for the entire verification test. Minimum analytical
turnaround time was required to achieve optimum process control. This method required that
ultra-pure (optimum) grade nitric acid be used, not reagent grade acid, to avoid the trace amounts
of arsenic that can be present in reagent grade nitric acid.
Arsenic analyses were also performed on-site for qualitative purposes using the Model QUICK
Low Range II field test kit from ITS. The arsenic field test kit has an optimum accuracy below 6
|ig/L and has a reaction time of less than 15 minutes. The complete method procedure is
presented in Appendix F.
At the end of the Capacity Test, TCLP analyses were performed on the regenerant wastewater by
TriMatrix Laboratories, Inc. using SW-846 and EPA Method 1311. TriMatrix Laboratories, Inc.
used Method SW-846 601 OB for As, Ba, Cd, Cr, Cu, Pb, Se, Ag, and Zn; and Method SW-846
7470A for Hg.
3.13.5 Samples Shipped Off-Site for Analysis
Samples for inorganic analysis by the PADEP Laboratory including chloride, sodium, sulfate,
silica, aluminum, total phosphorus, iron, and manganese, were collected and preserved in
accordance with Standard Methods procedure 3010 B, paying particular attention to the sources
of contamination as outlined in Standard Methods procedure 3010 C. After collection and
during shipment, the samples were maintained at a temperature of 2° to 8°C. The samples
collected for analysis by the PADEP Laboratory were dropped off at the Bucks County
Department of Health located in the Neshaminy Manor Center in Doylestown. The Bucks
County Department of Health ships water samples to the PADEP Laboratory on a daily basis,
Monday through Thursday. Any samples collected Friday through Sunday were kept refrigerated
until they could be shipped on Monday. The laboratory kept the samples between 2° to 8°C until
initiation of analysis. The samples collected for arsenic analysis by the NSF laboratory were
shipped by Gannett Fleming at the determined frequencies, without being preserved or packed in
ice, as per NSF instructions. The PADEP and NSF laboratories processed the samples for
analysis (logged in the samples) within 24 hours of receiving the samples. Table 3-10 presents
the sampling protocol that was followed during the verification test for samples analyzed by the
PADEP Laboratory.
45
-------
Table 3-10
Parameter
Laboratory
Aluminum,
Silica,
Sodium,
Iron and
Manganese
Arsenic
Sulf ate and
Chloride
Total
Phosphorus
. Water Quality Sampling Protocol
PADEP Sample ID Protocol- Sample Submission Sheet
Sample
Bottle
125 mL
HOPE3'
125 mL
Trace
Clean
500 mL
HOPE
125 mL
HOPE
Sample
Volume
125 mL
100 mL
250 mL
100 mL
Sample
Preservation
Nitric acid to
pH <2.0; iced
N/A(4)
Iced
Sulfuric acid
to pH <2.0;
iced
Sequence
Number™
Sample
Hold SAC(1)
Time Feed Treated No.(2)
6 101 102 109
months
6 N/A N/A N/A
months
28 days 201 202 109
28 days 201 202 109
NSF
Test Tracking ID
Bottle
Cap Collector Date/Time
D)(2) No.(2) Collected Integrity Capacity
M 1749 •/ I II
N/A N/A S I II
N/A 1749 •/ I II
P 1749 S I II
SAC: Standard Analysis Code.
'-2-1 Information also required on sample bottle.
(3) High Density Polyethylene.
(4) Samples preserved with nitric acid (to pH <2.0) upon arrival to NSF laboratory.
46
-------
3.13.6 Tests and Data Specific to Adsorptive Media Type Evaluated
ADI's MEDIA G2® adsorptive media used for this testing is described by data on the adsorptive
media type, characteristics, and tests, listed in Tables 2-1, 2-2, and 2-3.
3.14 Operations and Maintenance
The following are recommendations for criteria to be included in the Operations Manual for
adsorptive media removal of arsenic, as described in the Technology Specific Test Plan (TSTP)
within the ETV Protocol.
3.14.1 Operations
ADI provided a customized O&M manual (which included installation instructions) that
provided information needed to operate the equipment at the HTWSA Well No. 1 site. ADI's
Operations Manual is included in Appendix A.
3.14.2 Maintenance
ADI International Inc. provided readily understood information on the required or recommended
maintenance schedule for each piece of operating equipment including, but not limited to:
• manual valves;
• solenoid valve;
• pressure reducing valve;
• on-line measuring instruments;
• chemical metering pumps; and
• pressure gauges (Both pressure gauges required occasional removal for cleaning, which
was not mentioned in the Operations Manual).
ADI International Inc. provided readily understood information on the required or recommended
maintenance schedule for non-mechanical or non-electrical equipment including, but not limited
to:
• adsorptive media vessel; and
• feed lines.
47
-------
Chapter 4
Results and Discussion
4.1 Introduction
The verification test of the ADI International Inc. ADI Pilot Test Unit No. 2002-09 was
conducted in two phases, which included an Integrity Test and a Capacity Test. The two-week
(13 full days plus 8 hours) Integrity Test was conducted from October 8 through October 21,
2003. The Capacity Test was initiated in conjunction with the Integrity Test and continued until
May 28, 2004. The test included on-site media regeneration and one month of post-regeneration
operation. The verification test site was the HTWSA Well Station No. 1, located in Sellersville,
Pennsylvania. The well station and arsenic adsorption media filter were described in detail in
Chapter 2.
Due to unrepresentative feed water quality during the original Integrity Test, the Integrity Test
that began on October 8, 2003 was a retest. It was determined that the well station chlorine
detention tank had accumulated a significant amount of sediment, which would become
suspended whenever well pump operation was initialized. On September 10, 2003, HTWSA staff
removed the tank from service and flushed all sediment from the tank. Data collected during the
first Integrity Test, which was performed from August 12 through August 25, 2003, is not
presented in this report but is included in Appendix G for reference.
This chapter presents a summary of water quality and operational data, including preliminary
arsenic analyses, water quality immediately following chlorine detention tank cleaning,
equipment startup, results of the Integrity Test, results of the Capacity Test, and a discussion of
the results. The results and discussion encompass the concentration and speciation of arsenic in
the raw, feed (ST2), and treated (ST3) water; other water quality analyses of raw, feed (ST1),
feed (ST2), and treated (ST3) water; the quantity and rate of treated water production; backwash
and rinse wastewater quantity and quality; media regeneration summary; and equipment
operation characteristics. QA/QC procedures are also presented.
4.2 Equipment Installation, Startup, and Shakedown
The equipment was installed by ADI International Inc. personnel in September 2002. Initial
arsenic speciation tests on the feed (ST2) water and treated water (ST3) were performed in
December 2002, prior to the development of the PSTP. These initial arsenic tests were used to
make a preliminary assessment of the ability of the system to remove arsenic under the existing
water quality conditions at the site, and to evaluate the speciation of arsenic in the Łed and
treated water. Preliminary arsenic speciation results are presented in Table 4-1. The analytical
test reports and sample submission forms are included in Appendix H.
Preliminary arsenic speciation analyses indicated that 95% of the feed water total arsenic
concentration was in the soluble state. Arsenic III was detected in one of the feed water samples
at 55% of the total arsenic concentration; no arsenic III was detected in a second feed water
sample. The detection of arsenic III in the first sample may have been due to an under-dose of
chlorine. The treated water total arsenic concentration was equal to 50% of the total feed water
arsenic concentration during each sampling event, with all of the detectable arsenic occurring as
48
-------
arsenic V in the treated water. The average treated water arsenic concentration of 11 |ig/L
during the preliminary testing exceeded the proposed MCL of 10 |ig/L. Subsequent bench
testing conducted by the manufacturer determined that more effective removal of arsenic by
MEDIA G2® would occur if the existing feed water pH of 7.5 was depressed below 7.0,
preferably in the range of 6.5 to 6.8.
Table 4-1. Preliminary Arsenic Speciation
Feed (ST2)
Sample
Date
10-Dec-02
19-Dec-02
Total
Arsenic
(ug/L)
21.6
19.2
Soluble
Arsenic
(ug/L)
20.4
19.8
Arsenic III
(ug/L)
12.0
<4.0
Calculated
Arsenic V
(ug/L)
8.4
>15.8
Treated
Total
Arsenic
(ug/L)
10.7
10.9
Soluble
Arsenic
(ug/L)
10.2
11.9
Arsenic III
(ug/L)
<4.0
<4.0
Calculated
Arsenic V
(ug/L)
>6.2
>7.9
The arsenic speciation analytical method uses an anion exchange resin to separate arsenic III and
arsenic V. The anion exchange resin used for arsenic speciation during these preliminary arsenic
sampling events was subsequently determined from resin performance evaluation testing (using
known concentrations of arsenic III and arsenic V) to be only approximately 70% accurate in the
recovery of arsenic III. All subsequent arsenic speciations were performed using a new batch of
anion exchange resin, prepared by NSF. Arsenic speciation using the NSF-prepared resin
columns resulted in 100% recovery of known concentrations of arsenic III in performance
evaluation testing.
The laboratory arsenic analyses for the preliminary samples were performed at the PADEP
Laboratory, using the analytical method EPA 200.8, with a reporting limit of 4.0 |ig/L. All
arsenic analyses following the preliminary arsenic speciation were performed at the NSF
Laboratory, using the analytical method EPA 200.8, with a reporting limit of 2 |ig/L.
Performance evaluation testing results for arsenic speciation and on-site water quality analyses
are presented in Section 4.9.2.2.3.
Several physical modifications were made to the arsenic adsorption media filter system prior to
the initiation of testing. The source of feed water was moved from just upstream of the chlorine
detention tank to just downstream of the tank. For pH adjustment, a sulfuric acid feed system
was installed to inject acid into the feedwater just upstream of the arsenic adsorption media filter.
In addition, an electronic flow meter was installed just upstream of the electronic solenoid valve
on the feed water side of the filter.
The manufacturer installed new MEDIA G2® media on September 23, 2003, following the
chlorine detention tank cleaning. The media installation was witnessed by the Gannett Fleming
field technician. Installation of the media and pilot unit startup notes are summarized below.
A platform scale, calibrated on-site, was used to weigh 109 Ibs of MEDIA G2® prior to
installation of the media into the pilot filter vessel. Two bags of media were weighed
individually; each weighed 55.6 Ibs. The weight of each empty bag was 0.6 Ibs; therefore, the
tare weight of the media was 110 Ibs. One pound of media was weighed and removed prior to
49
-------
the media installation. Based on the reported media bulk density of 47 Ibs per cubic foot, the 109
Ibs of installed media should have resulted in an uncompacted bed volume of 2.3 ft3. However,
this could not be confirmed since the internal riser tube and flow distribution apparatus
contribute an unknown volume to the bed depth. Based on the filter vessel having an internal
tank diameter of 12 inches and a total tank height of 54 inches, the depth of media should be
around 35 inches, without taking into account the displaced volume due to the internal flow
distribution apparatus. According to the PSTP, "Data will be generated that will represent the
actual volume of water treated by the 2.3 cubic feet of MEDIA G2 media..."
Equipment startup and media conditioning were performed by the manufacturer and witnessed
by the Gannett Fleming field technician. The protocol for start-up is included in the ADI
Operations Manual in Appendix A. Following media installation, the filter was backwashed at a
rate of 3.2 gpm for a duration of 45 minutes. The manufacturer conditioned the new media by
using feed water dosed with sulfuric acid so that the rinse water pH was depressed to 4.5. The
media rinse and conditioning were combined in one step. The filter media was conditioned
overnight at the depressed pH of 4.5. However, the ADI Operating Manual instructions indicate
the rinse (treated) water pH need only be depressed to 6.5 during the conditioning period.
Prior to the installation of new MEDIA G2®, the totalizer meter reading was 120,181 gallons.
The totalizer meter reading at the initiation of the Integrity Test, following media conditioning,
was 124,874 gallons. This indicated that the manufacturer had used approximately 4,700 gallons
during the media conditioning period. Based on a media bed volume of 2.3 ft3, the volume of
water used during startup was equal to 272 bed volumes. Water used during the media
conditioning period was not included in the treated water volume used to assess the capacity of
the media.
4.3 Task 1: System Integrity Verification Testing
During the Integrity Test, Gannett Fleming evaluated the reliability of the arsenic adsorption
media filter equipment under the environmental and hydraulic conditions at the HTWSA Well
Station No. 1 site, with the arsenic adsorption media filter feed water supplied from the chlorine
detention tank.
4.3.1 Experimental Objectives
As established in the PSTP, the experimental objectives for the Integrity Test were as follows:
• Evaluate the equipment's operational reliability under field conditions;
• Document feed water quality and arsenic concentration; and
• Collect operational and water quality data under field conditions.
4.3.2 Preliminary Water Quality Analyses
Prior to initiation of the Integrity Test, several steps were taken to mitigate the degradation in
well water quality that occurred during the original Integrity Test. Degradation in well water
quality occurred due to the apparent resuspending of accumulated solids in the chlorine detention
50
-------
tank when the well pump operation was initiated and water from the well was discharged into the
bottom of the chlorine detention tank.
The chlorine detention tank was temporarily removed from service and sediment build-up was
removed from the tank by HTWS A. Following the tank cleaning, samples were collected for five
consecutive days (September 15-19, 2003) from both the raw water sample tap and the chlorine
detention tank sample tap (ST2), in order to evaluate the effectiveness of the tank cleaning. The
treatment system was off-line during this period of monitoring.
The samples were analyzed on-site for turbidity, pH, and FAC. The samples were also analyzed
both on-site and at the PADEP Laboratory for total and soluble fractions of iron, manganese, and
arsenic, as presented in Tables 4-2 and 4-3. Logbook copies and analytical data reports for the
preliminary water quality analyses are included in Appendix I. The results of the on-site
analyses for iron, manganese and arsenic are qualitative only, having used methods for the
analyses that are not EPA-approved. The Hach FerroVer and Periodate Oxidation methods, used
for analyzing iron and manganese, respectively, each require a predigestion step, which was not
performed. In addition, the Periodate Oxidation method is not considered accurate below 100
Hg/L. Therefore, the on-site iron, manganese, and arsenic results were not considered sufficiently
accurate for analysis of the effects of the tank cleaning and are not further discussed.
The data indicate that there was a reduction in turbidity, iron, manganese and arsenic between
the raw water tap and ST2. Turbidity and iron reduction was probably due to particulates settling
in the chlorine detention tank. The average reductions in turbidity, total iron (laboratory) and
total manganese (laboratory) were 33%, 44% and 43%, respectively. No total arsenic reduction
occurred. The average reductions in soluble fractions of raw water iron (laboratory), manganese
(laboratory), and arsenic (laboratory) were nearly 100%, 42% and 7%, respectively. Oxidation of
the soluble fraction of iron to a particulate apparently enabled significant removal to occur
through settling prior to the feed (ST2) sample location. Removal of soluble manganese can
possibly be accounted for by adsorption to the inside walls of the piping and tank, which
appeared to have developed a thin layer of manganese dioxide.
51
-------
Table 4-2. Preliminary On-site Water
After Tank Cleaning
Number of
Parameter Units Samples
Raw
PH
Turbidity
FAC
Iron - Tot
Iron - Sol
Manganese - Tot
Manganese - Sol
Arsenic - Tot
Feed (ST2)
PH
Turbidity
FAC
Iron - Tot
Iron - Sol
Manganese - Tot
Manganese - Sol
Arsenic - Tot
units
NTU
mg/L
Mg/L
Mg/L
Mg/L
Mg/L
Mg/L
units
NTU
mg/L
Mg/L
Mg/L
Mg/L
Mg/L
Mg/L
5
5
5
5
5
5
5
5
5
5
5
5
5
5
5
5
Quality Analyses (9/15/03 to 9/19/03) -
Mean(1) Minimum Maximum
7.5
0.50
0.07
340
90
100
140
13
7.5
0.35
1.00
80
30
160
100
7
7.4
0.35
0.02
90
10
<10
<10
12
7.5
0.25
0.81
70
10
100
<10
7
7.6
0.90
0.17
990
230
200
200
15
7.6
0.45
1.15
90
40
200
200
7
err
The median is reported for the pH data, not the mean.
52
-------
Table 4-3. Preliminary Laboratory Water Quality Analyses (9/15/03 to 9/19/03)
After Tank Cleaning
Number of
Parameter Units Samples Mean Minimum Maximum
Raw
Iron - Tot
Iron - Sol
Manganese - Tot
Manganese - Sol
Arsenic - Tot
Arsenic - Sol
Feed (ST2)
Iron - Tot
Iron - Sol
Manganese - Tot
Manganese - Sol
Arsenic - Tot
Arsenic - Sol
Mg/L
Mg/L
Mg/L
Mg/L
Mg/L
Mg/L
Mg/L
Mg/L
Mg/L
Mg/L
Mg/L
Mg/L
5
5
5
5
5
5
5
5
5
5
5
5
90
64
76
81
16
14
50
<10
43
47
16
13
80
<10
36
49
14
12
50
<10
36
36
16
13
120
100
91
93
17
16
50
<10
49
51
16
14
The water quality from the feed (ST2) sample tap following tank cleaning indicated that analyte
concentrations had either remained the same or decreased between the raw and feed (ST2)
locations. Therefore, the Integrity Test was initiated on October 8, 2003.
4.3.3 Integrity Test Operational Data
The arsenic adsorption media filter operated continuously, with feed water supplied from the
chlorine detention tank. Well No. 1 operated on demand, based on the water level in the storage
tank. Approximately 30,500 gallons were treated during the Integrity Test. A pressure reducing
valve installed on the chlorine detention tank sample tap was set to reduce system pressure from
115 psi to 50 psi prior to the treatment unit. Monitoring and on-site data collection were
performed, as scheduled, to verify the equipment performance.
Table 4-4 summarizes the operational data from the arsenic adsorption media filter during the
Integrity Test. Copies of the original logbook data sheets and compiled Integrity Test
operational data are included in Appendix J. The ball valve located on the treatment unit
discharge hose was throttled to control the flow rate through the arsenic adsorption media filter
in order to maintain the flow rate within the manufacturer's stated operational range.
The feed water pressure averaged 51 psi during the Integrity Test, within a range of 4 psi, which
is within the filter vessel's specified pressure range. The manufacturer indicated that the loss of
head across a clean filter bed at 1.7 gpm would be about 2 psi. However, the computed loss of
head across the treatment unit, based on the difference between the feed water and treated water
pressure gauges, was less than or equal to zero for the first six days of operation. However, the
feed water pressure gauge was found to read low by 2 psi, based on calibrations performed on
53
-------
each gauge by Gannett Fleming using a dead-weight pressure tester. An additional loss of head
of 6 psi developed between days six and seven of the testing, apparently due to the degradation
in groundwater quality as the result of a significant rainfall event. Filter bed loss of head
remained in the range of 4 psi to 6 psi for the remaining six days of the Integrity Test. The loss of
head, feed pressure, and treated pressure during the Integrity Test are presented in Figure 4-1.
Table 4-4. Integrity Test Operational Data
Before Flow After Flow Rate Feed Treated Pressure
Rate Adjustment Adjustment Pressure Pressure Differential
(gpm) (gpm) (psi) (psi) (psi)
Number of
Samples
Mean
Minimum
Maximum
Standard
Deviation
95%
Confidence
Interval
28
1.67
1.25
1.75
0.09
1.63
1.71
28
1.70
1.65
1.75
0.03
1.69
1.71
28
51
50
54
N/A
N/A
28
49
44
56
N/A
N/A
28
3
-2
6
N/A
N/A
N/A = Statistics not appropriate for these parameters.
10/6/03
10/9/03
10/12/03
10/15/03
Date
10/18/03
10/21/03
"Loss of Head ° Feed Pressure Treated Pressure I
o
10/24/03
Figure 4-1. Integrity Test Filter Loss of Head, Feed, and Treated Pressure.
54
-------
4.3.4 Integrity Test On-site Water Quality Analyses
The results of on-site water quality analyses performed during the Integrity Test are summarized
in Table 4-5. The feed and treated water temperatures were nearly identical throughout the test.
The feed water temperature varied less than 1°C during the two-week Integrity Test, as shown in
Figure 4-2. As shown in Figure 4-3, the raw water pH of 7.5 was generally depressed to within
the range of 6.5 to 7.0, as initially specified by the manufacturer. The average raw and feed
(ST1) water pH levels were relatively stable at 7.5. The addition of sodium hypochlorite had
minimal impact on pH, despite a slight increase in the alkalinity from the raw to the feed (ST1)
sample locations. Following the addition of sulfuric acid, the mean pH of 6.7 at the feed (ST2)
and treated water sample locations was within the range originally specified by the manufacturer
for optimal arsenic removal.
The raw and feed water turbidities were relatively high and variable for a groundwater supply, as
indicated on Figure 4-4. The average turbidity levels at the feed water sample taps (ST1 and
ST2) were 47% and 150% higher, respectively, than at the raw water sample tap. Several factors
likely account for this increase in turbidity. Based on the weather events recorded in the
logbook, there appeared to be a correlation between rainfall events and degradation in well water
quality, which would appear at different times at the raw and feed water sample taps. The lag
times between these sample locations could vary from 70 minutes to 17 hours depending on the
frequency of well pump operations. The combination of these factors results in turbidity
(including iron and manganese) spikes occurring at different times for these sampling locations.
Also, oxidation of raw water dissolved iron and manganese likely contributed to an increase in
the turbidity level due to the precipitation of particulates. Another factor that could have
contributed to the variation in water quality between sample locations was the on/off status of
well pump operations during sample collection.
Turbidity in the feed (ST2) was reduced by an average of 66% through the adsorption media
filter; however, the filter effluent turbidity averaged only 15% less than the raw water turbidity.
The highest treated water turbidity, of 3.8 NTU, occurred with the highest treated water iron and
manganese concentrations during the Integrity Test.
As shown in Figure 4-5, the alkalinity increased slightly from the raw water to the feed water
sample location (ST1), due to the addition of the alkaline chemical sodium hypochlorite.
Alkalinity at the feed water sample location (ST2) averaged 12% less than the raw water
alkalinity due to the addition of sulfuric acid. The feed (ST2) and treated water sample locations
had essentially the same level of alkalinity, indicating that MEDIA G2® had minimal impact on
alkalinity during the Integrity Test.
There was minimal variability between raw, feed (ST1 and ST2), and treated water fluoride
concentrations for any set of samples. As shown in Figure 4-6, recorded fluoride concentrations
were relatively variable from day to day. It is believed that some of this variation was due to
either malfunction or miscalibration of the fluoride analytical equipment. Three out of 14
fluoride data points were discarded due to suspected analytical problems and were not included
in the computed results presented in Table 4-5 or in Figure 4-6. MEDIA G2® had little, if any,
55
-------
impact on the level of fluoride in the feed water. The manufacturer indicated that fluoride would
not be removed by MEDIA G2®.
The results of analyses for calcium, magnesium, and total hardness indicate that the water is very
hard. A slight reduction in the total hardness and calcium levels appears to have occurred
through the treatment process. Only two samples were collected for calcium, magnesium, and
hardness analyses during the Integrity Test. Additional data was collected during the Capacity
Test and these water quality parameters are evaluated in more detail with the Capacity Test data
in Section 4.4.
A trace concentration of FAC was detected in the raw water, likely due to diffusion of chlorine
back from the chlorine detention tank when the well pump was off line. The two feed water
sample locations, ST1 and ST2, had similar levels of chlorine, although the chlorine
concentration level at ST2 was typically slightly lower due to the oxidant demand in the raw
water exerted during the travel time between the two locations (see Figure 4-7). Additional
chlorine depletion, averaging 0.32 mg/L, occurred between ST2 and the treated water, possibly
as a result of both further satisfying the oxidant demand of the raw water and the exertion of
oxidant demand by the media.
A complete summary of on-site water quality data and copies of the original logbook are
included in Appendix J.
56
-------
Table 4-5. Integrity Test On-site Water
Number
of
Parameter Units Samples
Raw
PH
Turbidity
Alkalinity
Fluoride
FAC
Feed (ST1)
PH
Turbidity
Alkalinity
Fluoride
FAC
Feed (ST2)
PH
Temperature
Turbidity
Alkalinity
Calcium
Magnesium
Hardness
Fluoride
FAC
Treated
PH
Temperature
Turbidity
Alkalinity
Calcium
Magnesium
Hardness
Fluoride
FAC
units
NTU
mg/L as CaCO3
mg/L
mg/L
units
NTU
mg/L as CaCO3
mg/L
mg/L
units
°C
NTU
mg/L as CaCO3
mg/L as CaCO3
mg/L as CaCO3
mg/L as CaCO3
mg/L
mg/L
units
°C
NTU
mg/L as CaCO3
mg/L as CaCO3
mg/L as CaCO3
mg/L as CaCO3
mg/L
mg/L
28
14
14
12
28
14
14
14
12
14
28
14
14
14
2
2
2
12
28
28
14
14
14
2
2
2
11
28
Quality Analyses
Mean/
Median'1' Minimum
7.5
0.65
147
0.35
0.06
7.4
0.95
156
0.36
1.25
6.7
13.2
1.6
130
260
18
278
0.35
1.13
6.7
13.4
0.55
130
247
19
266
0.35
0.80
7.4
0.15
130
0.13
0.02
7.4
0.35
146
0.13
0.91
5.9
12.9
0.35
120
258
18
276
0.13
0.82
5.9
13.0
0.15
116
246
18
264
0.13
0.29
Maximum
7.6
2.0
160
0.62
0.31
7.5
2.7
160
0.64
1.60
7.1
13.4
7.6
148
262
18
280
0.65
1.61
7.0
13.8
3.8
146
248
20
268
0.62
1.05
Standard
Deviation
N/A
0.50
7.04
0.12
0.07
N/A
0.60
3.41
0.13
0.18
N/A
0.17
2.0
11.0
N/A
N/A
N/A
0.13
0.19
N/A
0.20
0.95
10.6
N/A
N/A
N/A
0.14
0.25
95%
Confidence
Interval
N/A
0.30- 1.0
143 - 152
0.26 - 0.44
0.03 - 0.09
N/A
0.55- 1.4
153 - 158
0.26 - 0.46
1.13- 1.37
N/A
13.1 - 13.3
0.25 - 3.0
123 - 138
N/A
N/A
N/A
0.26 - 0.45
1.04- 1.21
N/A
13.2- 13.5
0-1.2
122 - 137
N/A
N/A
N/A
0.24 - 0.45
0.69 - 0.91
([) The median of the pH data is reported, not the mean.
N/A = Standard Deviation and 95% confidence intervals were not calculated for pH. Statistics not calculated for sample
sets of less than 8.
57
-------
14.0 •
I
6.0-
Date
|—•— Feed (ST2) —B- Treated (ST3) |
Figure 4-2. Integrity Test Temperature.
10
a
c.
10/8/03 10/10/03 10/12/03 10/14/03 10/16/03 10/18/03 10/20/03 10/22/03
Twice Daily Analyses
l~*-Raw """Feed (ST11 -*~Feed (ST21 Treated (ST3ll
Figure 4-3. Integrity Test pH.
58
-------
8.00
1.00
0.00
10/8/03 10/10/03 10/12/03 10/14/03 10/16/03 10/18/03 10/20/03
Date
l~*~Raw """Feed (ST11 ~*~Feed (ST21 Treated (ST311
Figure 4-4. Integrity Test Turbidity.
180
10/8/03
10/10/03 10/12/03 10/14/03 10/16/03
Daily Analyses
10/18/03 10/20/03
I-*-Raw ""-Feed (ST11 -*- Feed (ST21 Treated (ST31 I
Figure 4-5. Integrity Test Alkalinity.
59
-------
0.00
10/8/03 10/10/03 10/12/03 10/14/03 10/16/03 10/18/03 10/20/03
Date
l~*~Raw """Feed (ST11 ~*~Feed (ST21 Treated (ST311
Figure 4-6. Integrity Test Fluoride.
2.0
10/8/03 10/10/03 10/12/03 10/14/03 10/16/03 10/18/03 10/20/03 10/22/03
Date
-Raw -•- Feed (ST11 ~*~ Feed (ST21 Treated (ST311
Figure 4-7. Integrity Test FAC.
60
-------
4.3.5 Integrity Test Laboratory Water Quality Analyses
The results of water quality analyses performed at the PADEP Laboratory are summarized in
Table 4-6. Compiled data, copies of the original laboratory data reports, and sample submission
forms are included in Appendix K.
As shown in Figure 4-8, the sodium concentration at the four sample locations varied in unison
during the Integrity Test, within a range of 15%. There was no apparent sodium removal by, or
displacement from, the media. The overall variation in sodium concentrations was likely a result
of groundwater dilution from rainfall and variations in the sodium hypochlorite feed rate.
The filter media was apparently contributing silica to the treated water. The silica concentration
was initially 70% higher in the treated water than in the feed water at the beginning of the
Integrity Test. The silica concentration in the treated water decreased during the Integrity Test,
as indicated in Figure 4-9, so that by the end of the 14-day period the treated water silica
concentration was only 30% higher than in the feed water. By the end of the verification test, the
treated water silica concentration was approximately equal to the feed water silica concentration.
Concentrations of sulfate and chloride, analyzed twice during the Integrity Test, were not
significantly altered by the treatment process. Sulfate increased by 10% in the feed (ST2) water
relative to the raw water. This increase was most likely a result of the formation of sulfate by-
products following the addition of sulfuric acid for pH adjustment. Feed (ST2) and treated water
sulfate concentrations were essentially equal in the two samples. The raw water chloride level
remained the same through the treatment process. Both the chloride and sulfate feed levels were
considerably below the 250 mg/L levels that the manufacturer states may reduce the arsenic
adsorption capacity of MEDIA G2®.
The raw, feed (ST2), and treated water were analyzed twice during the Integrity Test for total
phosphorus. The analyses indicated a significant variation in concentration between sample
locations, particularly for the raw and feed water locations. This may represent a variation in
feed water quality, although it is also conceivable that the phosphorus reduction may be due to
sodium reacting with phosphorus, producing a by-product that precipitates in the detention tank.
No phosphorus was detected in the teated water, probably due to the media's affinity for
phosphorus. The effects of the phosphorus concentration by the media were further evaluated
following the Capacity Test.
There was a relative correlation in trends between the feed water turbidity, ron concentration,
and manganese concentration, presented in Figures 4-4, 4-10, and 4-11, respectively. The
correlation was particularly evident in regards to the spikes of these constituents, which peaked
on the fourth and ninth days of the Integrity Test. The peak levels of feed water iron and
manganese were greater than their secondary maximum contaminant levels (SMCLs) of
300 |ig/L and 50 |ig/L, respectively. Although the cause of the first spike is unknown, it is very
likely that the second spike in the water quality of the well supply was due to the impact of a
significant rainfall event. The treatment unit generally removed both iron and manganese below
their SMCLs, despite the variation in feed water concentrations. One treated water sample
during the Integrity Test, however, did exceed the SMCL for manganese, at 60 |ig/L. Aluminum
concentrations at all sample locations were below the detection limit of 200 |ig/L.
61
-------
Table 4-6. Integrity Test Laboratory Water Quality Analyses
Number of
Parameter Units Samples Mean Minimum Maximum
Raw
Sodium
Silica
Aluminum
Iron
Manganese
Chloride
Sulfate
Total
-L VJlCll
Phosphorus
Feed (ST2 Total)
Sodium
Silica
Aluminum
Iron
Manganese
Chloride
Sulfate
Total
-L VJlCll
Phosphorus
mg/L
mg/L
Mg/L
Mg/L
Mg/L
mg/L
mg/L
mg/L
mg/L
mg/L
Mg/L
Mg/L
Mg/L
mg/L
mg/L
mg/L
2
2
2
2
2
2
2
2
14
14
14
14
14
2
2
2
24.0
27.8
<200
147
102
35.2
101
0.063
25.1
27.8
<200
354
298
36.9
117
0.012
22.8
27.2
<200
144
99
34.2
99.9
O.010
22.8
27.2
<200
71
88
36.7
111
O.010
25.1
28.5
<200
149
104
36.1
103
0.115
27.1
28.5
<200
1120
1070
37.0
123
0.013
Standard
Deviation
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
1.56
0.40
0
342
305
N/A
N/A
N/A
95%
Confidence
Interval
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
24.1-26.0
27.6-28.1
<200 - <200
122 - 585
92 - 505
N/A
N/A
N/A
Feed (ST2 Soluble)
Sodium
Silica
Aluminum
Iron
Manganese
Treated
Sodium
Silica
Aluminum
Iron
Manganese
Chloride
Sulfate
Total
Phosphorus
mg/L
mg/L
Mg/L
Mg/L
Mg/L
mg/L
mg/L
Mg/L
Mg/L
Mg/L
mg/L
mg/L
mg/L
14
14
14
14
14
14
14
14
14
14
2
2
2
25.1
27.6
<200
75
45
25.3
38.9
<200
82
27
36.9
119
O.01
0
23.0
27.0
<200
<20
28
23.3
34.2
<200
32
<10
36.6
113
O.010
26.8
28.0
<200
202
68
26.8
47.3
<200
209
60
37.1
124
O.010
1.43
0.27
0
55
9
1.31
3.86
0
61
15
N/A
N/A
N/A
24.1-26.0
27.4 - 27.7
<200 - <200
38- 113
39-51
24.4-26.1
36.3-41.5
<200 - <200
41 - 123
17-38
N/A
N/A
N/A
N/A = Statistics not calculated for sample sets of less than 8.
62
-------
28.0
21.0
20.0
10/8/2003 10/10/2003 10/12/2003 10/14/2003 10/16/2003 10/18/2003 10/20/2003
Date
l~*~ Raw - Tot •""" Feed (ST21 Tot ~*~ Feed (ST21 Sol Treated (ST31 Tol
Figure 4-8. Integrity Test Sodium.
50.0 -i
45.0
40.0
35.0
30.0
25.0
20.0
15.0
10.0
5.0
0.0
10/8/2003 10/10/2003 10/12/2003 10/14/2003 10/16/2003 10/18/2003 10/20/2003
Date
I-*- Raw - Tot -•- Feed (ST21 Tot -*- Feed (ST21 Sol Treated (ST31Tol
Figure 4-9. Integrity Test Silica.
63
-------
1200
1000
800
600
400
200
10/8/2003 10/10/2003 10/12/2003 10/14/2003 10/16/2003 10/18/2003 10/20/2003
Date
l~*~Raw - Tot """Feed (ST21 Tot ~*~Feed (ST21 Sol Treated (ST31 Tot I
Figure 4-10. Integrity Test Iron.
1200
1000
800
I
cs
600
400
200
10/8/2003 10/10/2003 10/12/2003 10/14/2003 10/16/2003 10/18/2003 10/20/2003
Date
I-*-Raw - Tot "•"Feed (ST21 Tot -*~Feed (ST21 Sol Treated (ST31 Tot I
Figure 4-11. Integrity Test Manganese.
64
-------
4.3.6 Integrity Test Laboratory Arsenic Analyses
Raw water, feed water, and treated water samples were collected daily for arsenic analysis during
the Integrity Test. Seven of the sample sets were speciated to determine the fraction of soluble
arsenic in the arsenic III valence state relative to arsenic V. The results of the NSF laboratory
arsenic analyses are summarized in Table 4-7. Complete results of the arsenic analyses
including a summary table, analytical test reports, and chain of custody forms are included in
Appendix L. The raw data is on file at NSF.
The raw water total arsenic concentration averaged 13 |ig/L during the Integrity Test, of which 5
|ig/L was arsenic III. The average feed water arsenic concentration was 62% higher than the raw
water arsenic concentration, possibly due to resuspension of previously settled solids in the
detention tank. Arsenic III was not detected in the feed water following oxidation with sodium
hypochlorite. The average feed water total arsenic concentration was 21 |ig/L. Both the raw and
feed water arsenic concentrations, presented in Figures 4-12 and 4-13, varied widely. The
impact of the rainfall events on the well water quality may have contributed to the variations in
raw and feed water arsenic concentrations.
On average, 67% of the feed water total arsenic was removed by the MEDIA G2® filter during
the two-week Integrity Test. Total arsenic was detected in all of the treated water samples at
levels above 2 |ig/L, as presented in Figure 4-14. As indicated in Table 4-7, the treated water
arsenic concentrations during the Integrity Test averaged 7 |ig/L, with a maximum concentration
of 10 |ig/L. The lowest arsenic concentration in the treated water occurred on the first day of
operations, prior to the system reaching equilibrium, and was apparently due to the virgin
condition of the media. Interestingly, this occurred at the time of maximum displacement of
silica from the media. Average feed water-soluble arsenic and arsenic V were reduced by 45%
and 56%, respectively, in the treated water. The peak levels of total arsenic in the feed water, 38
|ig/L and 28 |ig/L, did not result in any significant increase in the treated water arsenic
concentration.
The feed water pH appeared to have a far greater impact on the removal of arsenic than the feed
water arsenic concentration. The highest treated water arsenic concentrations occurred when the
feed water pH was highest. The correlation between increasing feed water pH and increasing
treated water arsenic concentration is shown in Figure 4-15. The manufacturer indicated that the
feed water pH should be maintained between 6.5 and 6.8 for optimum arsenic removal, but
difficulties encountered with the acid feed pump operation resulted in several periods when the
pH was above this range. A decrease in feed water pH from 7.1 to 6.2 on the ninth day of the test
resulted in a 70% decrease in the treated water arsenic concentration. Thereafter, correlations in
treated water arsenic with variations in feed water pH were not as significant but continued to
occur.
Figure 4-16 shows that the majority of the treated water total arsenic concentration was in the
form of soluble arsenic, composed primarily of the arsenic V species.
65
-------
FAC concentration did not appear to significantly impact the treated water arsenic
concentrations, as shown in Figure 4-17. However, the feed water FAC concentration was
relatively stable during the Integrity Test.
Field arsenic analyses, performed using the ITS QUICK Low Range II test kit, were used to
monitor the feed and treated water arsenic concentrations on-site for operations control and for
quick qualitative results. However, the test kits are not an EPA-approved method, so the data are
not presented here. On-site arsenic analyses results are included in the logbook copies in
Appendix J.
Table 4-7. Integrity
Number of
Samples
Mean
Minimum
Maximum
Standard
Deviation
95%
Confidence
Interval
Total
Arsenic
(ug/L)
14
13
8
18
3
11- 15
Test Laboratory Arsenic Analyses
Soluble
Arsenic
(ug/L)
7
12
7
16
N/A
N/A
Raw
Arsenic
III
(ug/L)
7
5
<2
12
N/A
N/A
Feed (ST2)
Calculated
Arsenic V
(ug/L)
7
6
2
12
N/A
N/A
Total Soluble Arsenic Calculated
Arsenic Arsenic III Arsenic V
(ug/L) (ug/L) (ug/L) (ug/L)
14 7 7 7
21 11 <2 9
15 8 <2 6
38 13 <2 11
6.0 N/A N/A N/A
16-25 N/A N/A N/A
Treated
Number of
Samples
Mean
Minimum
Maximum
Standard
Deviation
95%
Confidence
Interval
Total
Arsenic
(ug/L)
14
7
2
10
N/A
N/A
Soluble
Arsenic
(ug/L)
7
6
3
12
N/A
N/A
Arsenic
III
(ug/L)
7
<2
<2
<2
N/A
N/A
Calculated
Arsenic V
(ug/L)
7
4
1
10
N/A
N/A
N/A = Statistics not calculated for sample sets of less than 8. Statistics not appropriate for treated water arsenic.
66
-------
50
45
40
35
30
.a 25
10/8/03 10/10/03 10/12/03 10/14/03 10/16/03 10/18/03 10/20/03
Date
l~*~ Total Soluble ~*~Arsenic m Arsenic VI
Figure 4-12. Integrity Test Raw Total, Soluble and Speciated Arsenic.
50
45
40
35
10/8/03 10/10/03 10/12/03 10/14/03 10/16/03
Analyses on Alternate Days
10/18/03 10/20/03
"Total """Soluble ~*~ Arsenic m Arsenic Vl
Figure 4-13. Integrity Test Feed (ST2) Total, Soluble and Speciated Arsenic.
67
-------
10/8/03
10/10/03 10/12/03 10/14/03 10/16/03 10/18/03 10/20/03
Date
l~*~Raw """FeedtSTll ~*~ Treated (ST31 I
Figure 4-14. Integrity Test Total Arsenic.
Date
6.5
• 6.0
5.0
(ST2) Tot As CZlTrtd Tot As -B- Fd (ST2) pH -*- Trtd pH |
Figure 4-15. Integrity Test - Effect of pH on Arsenic Removal.
68
-------
50
45
40
35
3 30
I
.a 25
a
0)
Ł
-------
4.3.7 Integrity Test Equipment Operation
During the Integrity Test, minimal time and/or attention was required to operate the equipment,
although significant time was spent conducting on-site analyses. The pilot adsorption media
filter is manually operated and uses electricity only for powering the feed water solenoid valve
and the chemical feed pumps for metering sulfuric acid, sodium hypochlorite, and caustic soda.
The time required for daily operation of the treatment unit included about ten minutes to monitor
the flow rate, acid tank level, totalizer, and loss of head, and to verify that there were no leaks in
the system, Permanent installation of the equipment would also require daily pH and FAC
analyses, as well as periodic on-site arsenic analyses and/or collection of samples for laboratory
analyses. FAC and pH analyses require approximately 15 to 20 minutes. On those days in which
on-site arsenic analyses with a field test kit are also performed, the total analytical time is
45 minutes.
A filter backwash was performed to fulfill the PSTP requirement of conducting a minimum of
one backwash during the Integrity Test, regardless of whether or not the criteria for backwash
had occurred. Backwash data from the Capacity Test are summarized and discussed in more
detail in Section 4.7.
4.4 Task 2: Adsorption Capacity Verification Testing
The Capacity Test began on October 8, 2003, coinciding with the initiation of the Integrity Test.
Water quality sampling and analysis, system monitoring, and data collection were performed as
scheduled in the test plan and described in Chapter 3. The filter media was regenerated
following the manufacturer's specified procedures on April 30, 2004. The test concluded on
May 28, 2004, following four weeks of post-regeneration operation. The results of the
Capacity Test, which includes data collected during the Integrity Test, are detailed in the
following sections.
4.4.1 Experimental Objectives
The experimental objective of the Capacity Test was to provide operating and water quality data
relative to the ability of the arsenic adsorption media filter system to remove arsenic from the
feed water under field conditions.
4.4.2 Capacity Test Operational Data
The treatment unit operated continuously during the Capacity Test, with feed water supplied
from the storage tank. Well No. 1 operated on demand, based on the water level in the storage
tank. Monitoring and on-site data collection were performed as scheduled to verify the
equipment performance. Table 4-8 summarizes the arsenic adsorption media filter unit
operational data during the Capacity Test, through the date of media regeneration. Table 4-9
summarizes the operational data during the four-week post-regeneration period. Copies of the
original logbook data sheets and compiled operational data are included in Appendix J. The non-
integral flow control system, consisting of a pressure regulating valve on the chlorine detention
tank sample tap and a ball valve on the filter unit discharge, maintained a relatively constant flow
rate of 1.69 gpm through the filter unit. The design flow rate was 1.7 gpm.
70
-------
Table 4-8. Capacity Test Operational Data
Number of Samples
Mean
Minimum
Maximum
Standard Deviation
95% Confidence Interval
Before Flow
Rate
Adjustment
(gpm)
398
1.68
1.25
1.77
0.04
1.68- 1.69
After Flow
Rate
Adjustment
(gpm)
398
1.69
1.57
1.77
0.03
1.69- 1.69
Feed
Pressure
(psi)
398
54
50
58
N/A
N/A
Treated
Pressure
(psi)
398
48
42
56
N/A
N/A
Pressure
Differential
(psi)
398
5
-2
11
N/A
N/A
N/A = Statistics not appropriate for these parameters.
Table 4-9. Post-Regeneration Operational Data
Number of Samples
Mean
Minimum
Maximum
Standard Deviation
95% Confidence Interval
Before Flow
Rate
Adjustment
(gpm)
58
1.70
1.60
1.75
0.02
1.69- 1.71
After Flow
Rate
Adjustment
(gpm)
58
1.70
1.66
1.75
0.02
1.69- 1.71
Feed
Pressure
(psi)
58
55
54
56
N/A
N/A
Treated
Pressure
(psi)
58
49
45
52
N/A
N/A
Pressure
Differential
(psi)
58
6
2
10
N/A
N/A
N/A = Statistics not appropriate for these parameters.
The equipment operated continuously 24 hours per day. The feed water pressure was maintained
by the pressure-regulating valve to within the manufacturer's recommended pressure limits of 30
to 125 psi and averaged 54 psi. The filter bed headloss accumulated significantly as a function
of run time, as shown in Figure 4-18. A pattern of accumulating head loss between backwashes
and a reduction in filter bed headloss following each filter backwash is apparent in the figure.
Particulate iron and manganese likely contributed significantly to the accumulating headloss.
The headloss across the treatment unit averaged 5 psi. The clean-bed headloss, observed during
the first week of testing was 2 psi, when the pressure gauge inaccuracy is accounted for. The
feed water pressure gauge was found to read low by about 2 psi, based on calibration checks
performed on each gauge by Gannett Fleming using a dead-weight pressure tester.
71
-------
70
4/28/04
5/28/04
I * FeedPressure *^~Treated Pressure ^^"Regeneration ^ Headloss I
Figure 4-18. Capacity Test Feed Pressure, Treated Pressure, and Filter Bed Headloss.
4.4.3 Capacity Test On-site Water Quality Analyses
The results of Capacity Test on-site water quality analyses are summarized in Table 4-10. Post-
regeneration on-site water quality data are summarized in Table 4-11. The media regeneration
apparently had no effect on the treated water on-site water quality parameters including pH,
turbidity, alkalinity, fluoride, FAC, calcium, magnesium, and hardness, as shown in Tables 4-10
and 4-11 and Figures 4-19 through 4-25. Copies of the original logbook data sheets and
compiled on-site water quality data are included in Appendix J.
72
-------
Table 4-10. Capacity Test On-site Water Quality Analyses
Number of Mean/
Parameter Units Samples Median'1' Minimum
Raw
pH
Turbidity
Alkalinity
Calcium
Magnesium
Hardness
Fluoride
FAC
Feed(STl)
pH
Turbidity
Alkalinity
Fluoride
FAC
Feed(ST2)
pH
Temperature
Turbidity
Alkalinity
Calcium
Magnesium
Hardness
Fluoride
FAC
Treated
pH
Temperature
Turbidity
Alkalinity
Calcium
Magnesium
Hardness
Fluoride
FAC
units
NTU
mg/L as CaCO}
mg/L as CaCO3
mg/L as CaCO}
mg/L as CaCO}
mg/L
mg/L
units
NTU
mg/L as CaCO3
mg/L
mg/L
units
°C
NTU
mg/L as CaCO}
mg/L as CaCO}
mg/L as CaCO}
mg/L as CaCO}
mg/L
mg/L
units
°C
NTU
mg/L as CaCO}
mg/L as CaCO}
mg/L as CaCO}
mg/L as CaCO3
mg/L
mg/L
uj Median is reported for pH only.
N/A = Standard Deviation and 95%
sets of less than 8.
217
201
96
2
2
2
21
38
107
203
28
21
98
218
201
203
98
29
27
29
43
218
218
201
203
98
29
27
28
42
218
confidence
7.57
0.95
152
221
25
246
0.32
0.06
7.55
0.75
155
0.32
0.94
6.43
13.2
0.70
121
254
22
272
0.30
0.85
6.39
13.2
0.30
117
254
22
271
0.30
0.69
intervals were
6.89
0.10
126
218
24
242
0.13
0.02
7.24
0.25
132
0.13
0.05
5.70
12.0
0.15
62
218
4
238
0.13
0.05
5.51
12.2
0.05
54
228
2
242
0.13
0.04
not calculated
Standard
Maximum Deviation
7.83
7.5
188
224
26
250
0.62
0.31
7.86
2.7
174
0.64
2.80
7.09
14.3
7.6
148
296
40
320
0.65
2.17
7.25
14.4
3.8
162
298
82
322
0.62
2.42
for pH. Statistics not
N/A
1.1
10.2
N/A
N/A
N/A
0.10
0.06
N/A
0.30
7.43
0.11
0.32
N/A
0.22
0.65
14
20.7
11
23.1
0.08
0.24
N/A
0.23
0.30
16
15.7
16
22.6
0.08
0.25
95%
Confidence
Interval
N/A
0.80-1.1
149- 154
N/A
N/A
N/A
0.26-0.37
0.04-0.08
N/A
0.70-0.80
151- 158
0.27-0.38
0.86-1.01
N/A
13.1-13.2
0.60-0.80
117- 124
245-264
17-27
262 - 282
0.27-0.33
0.81-0.89
N/A
13.2-13.3
0.25-0.35
113- 120
247 - 261
15-29
261 -281
0.27-0.33
0.65-0.72
calculated for sample
73
-------
Table 4-11. Post-Regeneration On-site Water Quality Analyses
Number
of Mean/
Parameter Units Samples Median'1' Minimum
95%
Standard Confidence
Maximum Deviation Interval
Raw
PH
Turbidity
Alkalinity
Feed (ST1)
PH
Turbidity
FAC
Feed (ST2)
PH
Temperature
Turbidity
Alkalinity
Calcium
Magnesium
Hardness
Fluoride
FAC
Treated
PH
Temperature
Turbidity
Alkalinity
Calcium
Magnesium
Hardness
Fluoride
FAC
units
NTU
mg/L as CaCO3
units
NTU
mg/L
units
°C
NTU
mg/L as CaCO3
mg/L as CaCO3
mg/L as CaCO3
mg/L as CaCO3
mg/L
mg/L
units
°C
NTU
mg/L as CaCO3
mg/L as CaCO3
mg/L as CaCO3
mg/L as CaCO3
mg/L
mg/L
29
28
11
13
28
11
29
28
28
11
4
4
4
4
28
29
28
28
11
4
4
4
4
28
7.71
0.85
158
7.72
0.75
0.82
6.53
13.3
0.60
123
276
46
322
0.29
0.75
6.51
13.3
0.35
114
262
43
305
0.29
0.69
7.56
0.25
148
7.59
0.40
0.04
6.08
13.2
0.25
102
256
14
290
0.22
0.04
6.05
13.1
0.10
82
242
22
292
0.23
0.02
7.79
3.8
188
7.76
2.0
1.08
6.82
13.4
1.2
134
288
86
374
0.34
1.25
6.71
13.4
0.75
128
278
56
314
0.32
1.18
N/A
0.90
13.4
N/A
0.45
0.38
N/A
0.08
0.25
8.50
N/A
N/A
N/A
N/A
0.45
N/A
0.08
0.15
11.7
N/A
N/A
N/A
N/A
0.43
N/A
0.45- 1.2
147 - 168
N/A
0.55 - 0.95
0.52- 1.12
N/A
13.2- 13.3
0.50 - 0.70
117- 130
N/A
N/A
N/A
N/A
0.54 - 0.95
N/A
13.2- 13.3
0.25 - 0.40
104 - 123
N/A
N/A
N/A
N/A
0.50-0.88
([) Median is reported for pH only.
N/A = Standard Deviation and 95% confidence intervals were not calculated for pH. Statistics not calculated for
sample sets of less than 8.
74
-------
The raw and feed (ST1) pH values were nearly equal with median values of 7.57 and 7.55,
respectively, and were relatively stable throughout the Capacity Test as shown in Figure 4-19.
The feed (ST2) and treated (ST3) water pH values were significantly lower due to the addition of
sulfuric acid to lower pH and improve the arsenic removal capacity of the media. The feed
(ST2) median pH was 6.43 and the treated (ST3) median pH was 6.39. The variability in the
feed (ST2) and treated (ST3) water pH values was due primarily to chemical feed pump
operations. Also, as shown in Figure 4-19, the target pH was reduced in early November, at the
manufacturer's request, to improve arsenic removal. As shown in Figure 4-19, at the request of
the manufacturer, a one-week period of operation with a reduced feed (ST2) pH with a target
range of 6.0 to 6.2 was conducted prior to the media regeneration. In response to the reduction
in feed (ST2) pH, the treated water arsenic concentration was reduced during this period, as
shown later in this section. However, the period of reduced pH operation immediately prior to
regeneration was not included in the calculation of the media capacity. The target pH was
adjusted back to 6.5 after media regeneration.
**y^r>
"Raw ~*~FeedSTl ~*~FeedST2 ~""Treated ST3 ~~Regeneration |
Figure 4-19. Capacity Test pH.
75
-------
Due to the relatively short hydraulic detention time, the feed (ST2) and treated (ST3) water
temperatures were nearly equal throughout the test, both averaging 13.2°C. Capacity Test and
post-regeneration feed (ST2) and treated water temperatures are shown in Figure 4-20. The
water temperatures were relatively stable, with minimum and maximum temperatures separated
by approximately 2°C during the Capacity Test.
16.0
14.0
12.0
g 8.0
8.
4.0
Date of Media Regeneration
1/29/04
Date
2/28/04
"Feed(ST2) ~"~ Treated (ST3) ^~ Regeneration I
Figure 4-20. Capacity Test Temperature.
76
-------
For a groundwater supply, the raw and feed water turbidities were relatively high and variable, as
shown in Figure 4-21. The Capacity Test raw water turbidity averaged 0.95 NTU and was
variable, ranging from 0.10 NTU to 7.5 NTU. The raw water turbidity was apparently impacted
by precipitation events. The feed water turbidity averaged 0.75 NTU (ST1) and 0.70 NTU (ST2)
but at times exceeded the raw water turbidity. This indicates that additional water quality
deterioration, possibly resulting from the oxidation of dissolved iron and manganese and/or the
disturbance of sediment in the chlorine contact tank, occurred prior to the filter unit. In addition,
as discussed with the Integrity Test data in Section 4.3.4, variable demands and the frequency of
well pump operations could result in a lag time between raw and feed water turbidity spikes.
The treated water turbidity averaged 0.30 NTU, with minimum and maximum recorded values of
0.05 NTU and 3.8 NTU, respectively. The reduction in turbidity from the feed to the treated
water indicates that filtration of particles was occurring within the treatment unit.
10/1/03
4/28/04
5/28/04
"Raw ~*~Feed(STl) ~*~Feed (ST2) ~B~Treated (ST3) ^"Regeneration I
Figure 4-21. Capacity Test Turbidity.
77
-------
The Capacity Test alkalinity concentrations are shown in Figure 4-22. The average raw and feed
(ST1) alkalinity concentrations were nearly equal, averaging 152 and 155 mg/L as
respectively, as shown in Table 4-10. The feed (ST2) alkalinity averaged 121 mg/L as
and the treated water alkalinity averaged 117 mg/L as CaCCb. These feed (ST2) and treated
water alkalinities were 20% and 23% less than the raw water alkalinity, respectively. This
alkalinity reduction was a direct result of the addition of sulfuric acid for pH adjustment. In
addition, the treated water alkalinity was approximately 3% less than the feed (ST2) alkalinity,
which indicates that some alkalinity may have been consumed in the filter unit. As shown in
Figure 4-22, at the request of the manufacturer, a one-week period of operation with a reduced
feed (ST2) pH was conducted prior to the media regeneration. As a result of the reduction in
feed (ST2) pH, the feed (ST2) and treated water alkalinity was reduced during the period of
reduced pH.
10/1/03
10/31/03
4/28/04
5/28/04
|~*~Raw ~*~Feed(STl) ~*~Feed (ST2) """Treated (ST3) ^"Regeneration I
Figure 4-22. Capacity Test Alkalinity Concentration.
78
-------
Figure 4-23 shows that raw, feed (ST1 and ST2), and treated water fluoride concentrations were
unaffected by the treatment process, which supports the manufacturer's claim that the media
does not remove fluoride. The fluoride concentration averaged 0.3 mg/L in the raw, feed, and
treated water. The feed (ST2) and treated water fluoride concentrations both averaged 0.30
mg/L. Variable fluoride concentrations at the beginning of the Integrity Test were suspected to
be a result of either malfunction or miscalibration of the fluoride analytical equipment and were
not included in the statistical calculations or in Figure 4-23.
0.70
5/28/04
"Raw ~*~Feed (ST1) ~*~Feed(ST2) ~B~Treated (ST3) ^^Regeneration I
Figure 4-23. Capacity Test Fluoride Concentration.
79
-------
As shown in Figure 4-24, FAC concentrations generally decreased from feed (ST1), to feed
(ST2), to the treated water, likely due to the oxidant demand of the raw water and possibly an
oxidant demand of the media. During the Capacity Test, the raw water FAC, prior to the
hypochlorite feed point, averaged 0.06 mg/L. Low concentrations of FAC detected in the raw
water were likely a result of diffusion of chlorine back from the chlorine detention tank when the
well pump was offline. The feed (ST1) and feed (ST2) FAC concentrations during the Capacity
Test averaged 0.94 mg/L and 0.85 mg/L, respectively, and the treated water had an average FAC
concentration of 0.69 mg/L. The significant variation in FAC that occurred between April 22,
2004 and May 8, 2004 was the result of an HTWSA hypochlorite feed pump malfunction.
"Raw ~*~Feed (STl) ~*~ Feed (ST2) ~"~ Treated (ST3) Regeneration I
Figure 4-24. Capacity Test FAC.
80
-------
Capacity Test water quality analyses indicate that calcium, magnesium, and total hardness
concentrations in the feed water were relatively consistent during the test period and were
apparently unaffected by the treatment process, as shown in Figure 4-25. Feed (ST2) and treated
water calcium concentrations both averaged 254 mg/L as CaCCb. The average feed and treated
water magnesium concentrations were equal at 22 mg/L as CaCOs. The total hardness
concentrations of the feed (ST2) and treated water were 272 mg/L as CaCOs and 271 mg/L as
s, respectively.
" Hardness (Feed)
" Magnesium (Feed)
"Hardness (Treated)
"Magnesium (Treated)
"Calcium (Feed)
"Regeneration
"Calcium (Treated)
Figure 4-25. Capacity Test Calcium, Magnesium, and Hardness.
4.4.4 Capacity Test Laboratory Water Quality Analyses
The results of water quality analyses performed at the PADEP Laboratory are summarized for
the Capacity Test in Table 4-12. Laboratory water quality analyses performed following media
regeneration are summarized in Table 4-13. The media regeneration had no effect on the feed
(ST2) and treated water quality parameters analyzed at the PADEP Laboratory, as shown in
Tables 4-12 and 4-13 and in Figures 4-26 through 4-39. Laboratory water quality data and the
analytical test reports and sample submission forms are included in Appendix K. The raw data
are on file atNSF.
81
-------
Table 4-12. Capacity Test Laboratory Water Quality Analyses
Number
of
Parameter Units Samples Mean Minimum Maximum
Raw
Sodium
Silica
Aluminum
Iron (Total)
Iron (Soluble)
Manganese (Total)
Manganese
(Soluble)
Chloride
Sulfate
Total Phosphorus
Feed (ST1)
Sodium
Silica
Aluminum
Iron
Manganese
Feed (ST2)
Sodium (Total)
Sodium (Soluble)
Silica (Total)
Silica (Soluble)
Aluminum (Total)
Aluminum (Soluble)
Iron (Total)
Iron (Soluble)
Manganese (Total)
Manganese
(Soluble)
Chloride
Sulfate
Total Phosphorus
mg/L
mg/L
Vg/L
Mg/L
Mg/L
Mg/L
Mg/L
mg/L
mg/L
mg/L
mg/L
mg/L
Mg/L
Mg/L
Mg/L
mg/L
mg/L
mg/L
mg/L
Mg/L
Mg/L
Mg/L
Mg/L
Mg/L
Mg/L
mg/L
mg/L
mg/L
28
28
28
90
62
90
62
2
2
2
26
26
26
26
26
41
14
41
14
41
14
96
69
96
69
29
29
29
23.8
28.3
<200
295
153
105
98
35.2
101
0.063
25.0
28.5
<200
268
324
25.1
25.1
28.5
27.6
208(1)
<200
180
38
140
52
36.9
155
0.011
21.4
24.0
<200
81
<20
90
<10
34.2
100
O.010
22.6
25.5
<200
73
93
22.7
23.0
25.7
27.0
<200
<200
47
<20
77
28
36.1
111
O.010
26.9
33.4
<200
2370
733
219
128
36.1
103
0.115
27.5
32.1
<200
2390
2390
29.0
26.8
40.2
28.0
539
<200
1120
202
1070
94
37.6
202
0.016
Standard
Deviation
1.38
1.96
0
316
101
18
17
N/A
N/A
N/A
1.36
1.47
0
443
466
1.41
1.43
2.28
0.27
53
0
158
34
133
9
0.37
17.4
0.002
95%
Confidence
Interval
23.2-24.5
27.5 - 29.2
<200 - <200
220 - 371
123 - 182
101- 110
93 - 103
N/A
N/A
N/A
24.4-25.7
27.9 - 29.2
<200 - <200
61 - 475
106 - 542
24.5 - 25.6
24.1 - 26.0
27.7 - 29.4
27.4-27.7
<200 - 228
<200 - <200
143 - 217
28-47
109- 171
49-54
36.8-37.1
147 - 163
0.011-0.012
82
-------
Table 4-12. Capacity Test Laboratory
Number
of
Parameter Units Samples
Treated
Sodium
Silica
Aluminum
Iron (Total)
Iron (Soluble)
Manganese (Total)
(Soluble)
Chloride
Sulfate
Total
Phosphorus
mg/L
mg/L
Mg/L
Mg/L
Mg/L
Mg/L
Mg/L
mg/L
mg/L
mg/L
41
41
41
96
55
96
55
29
29
29
Water Quality Analyses (Continued)
Standard
Mean Minimum Maximum Deviation
25.3
32.7
<200
68
50
16
<10
37.0
160
0.010
22.8
26.8
<200
<20
<20
<10
<10
36.2
113
<0.010
32.0
47.3
<200
956
556
79
<10
37.7
205
0.011
1.56
5.20
0
117
83
13
0
0.40
19.0
0.000
95%
Confidence
Interval
24.7 - 25.8
30.8-34.6
<200 - <200
41-96
24-76
13-20
<10-<10
36.8-37.1
152 - 169
O.010-
0.010
One feed (ST2) aluminum result of 539 ng/L, which is suspected to be an analytical error, skewed both the
average and upper limit of the 95% confidence interval. If the suspected analytical error is discarded, the average
feed (ST2) aluminum concentration would be <200 |jg/L and the upper limit of the 95% confidence interval would
be <200 ng/L.
N/A = Statistics not calculated for sample sets of less than 8.
Table 4-13. Post-Regeneration Laboratory Water Quality Analyses
Number
of
Parameter Units Samples Mean Minimum Maximum
Raw
Sodium
Silica
Aluminum
Iron (Total)
Iron (Soluble)
Manganese (Total)
Manganese
(Soluble)
Feed (ST1)
Sodium
Silica
Aluminum
Iron
Manganese
mg/L
mg/L
Mg/L
Mg/L
Mg/L
Mg/L
Mg/L
mg/L
mg/L
Mg/L
Mg/L
Mg/L
4
4
4
19
15
19
15
4
4
4
4
4
25.1
29.9
<200
372
135
101
98
27.0
30.3
<200
169
168
24.1
28.5
<200
46
<20
91
77
26.0
29.1
<200
111
118
25.9
31.2
<200
1080
423
110
107
28.7
31.5
<200
302
252
Standard
Deviation
N/A
N/A
N/A
314
120
4
7
N/A
N/A
N/A
N/A
N/A
95%
Confidence
Interval
N/A
N/A
N/A
196 - 548
57-213
99 - 103
93 - 102
N/A
N/A
N/A
N/A
N/A
83
-------
Table 4-13. Post-Regeneration Laboratory Water Quality Analyses (Continued)
Number
of Standard
Parameter Units Samples Mean Minimum Maximum Deviation
Feed (ST2)
Sodium
Silica
Aluminum
Iron (Total)
Iron (Soluble)
Manganese (Total)
Manganese
(Soluble)
Chloride
Sulfate
Total Phosphorus
Treated
Sodium
Silica
Aluminum
Iron (Total)
Iron (Soluble)
Manganese (Total)
Manganese
(Soluble)
Chloride
Sulfate
Total
Phosphorus
mg/L
mg/L
Mg/L
Mg/L
Mg/L
Mg/L
Mg/L
mg/L
mg/L
mg/L
mg/L
mg/L
Mg/L
M^L
Mg/L
M^L
Mg/L
mg/L
mg/L
mg/L
4
4
4
19
15
19
15
4
4
4
4
4
4
19
15
19
16
4
4
4
26.4
30.1
<200
125
36
112
58
37.7
152
0.028
26.8
32.1
<200
74
44
13
10
37.7
161
0.015
26.0
28.9
<200
62
<20
34
<10
36.4
146
O.010
26.0
31.5
<200
<20
<20
<10
<10
36.3
148
O.010
27.1
30.6
<200
306
65
213
99
38.6
162
0.043
27.8
33.0
<200
339
228
39
15
38.6
169
0.025
N/A
N/A
N/A
61
16
41
28
N/A
N/A
N/A
N/A
N/A
N/A
89
54
8
1
N/A
N/A
N/A
95%
Confidence
Interval
N/A
N/A
N/A
91 - 160
26-47
90 - 135
39-76
N/A
N/A
N/A
N/A
N/A
N/A
24 - 124
9-79
9- 18
10- 11
N/A
N/A
N/A
N/A = Statistics not calculated for sample sets of less than 8.
84
-------
Sodium concentrations varied at the four sample points throughout the Capacity Test period, as
shown in Figure 4-26, likely due to rainfall events and minor variations in the sodium
hypochlorite feed rate. Although the total sodium concentration varied from a minimum of 21.4
mg/L in the raw water to a maximum of 32.0 mg/L in the treated water over the Capacity Test
period, the average increase in sodium concentration from the raw water to the treated water was
1 mg/L, which indicates that the sodium hypochlorite contributes a continuous, but small portion
of the treated water sodium concentration. The average raw water sodium concentration was
23.8 mg/L and the feed (ST1), feed (ST2), and treated water sodium concentrations averaged
25.0 mg/L, 25.1 mg/L, and 25.3 mg/L, respectively.
10/1/03
10/31/03 11/30/03 12/30/03 1/29/04
Date
3/29/04
4/28/04
5/28/04
-Raw Feed (ST1) -*-Feed (ST2) Total Feed (ST2) Soluble -""Treated (ST3) Regeration I
Figure 4-26. Capacity Test Sodium Concentration.
85
-------
As discussed with the Integrity Test data, Figure 4-27 shows that silica concentrations in the
treated water were initially much greater than silica concentrations in the raw and feed water,
indicating that the media contributed silica to the treated water. After the initial one to two
months of the test, the treated water silica concentration did not appear to be significantly
affected by the media. During the Capacity Test, the average treated water silica concentration
was 32.7 mg/L, while the raw, feed (ST1), and feed (ST2) silica concentrations averaged 28.5
mg/L.
5/28/04
•Raw Feed(STl) ~**~Feed (ST2) Total Feed (ST2) Soluble ~"~ Treated (ST3) ^"Regeneration I
Figure 4-27. Capacity Test Silica Concentration.
86
-------
Raw and feed water aluminum concentrations were generally less than the method detection
limit (MDL) of 200 |J,g/L and were unaffected by the treatment process, as indicated in Tables 4-
12 and 4-13. Only one feed water sample result was greater than the MDL of 200 ng/L. This
feed (ST2) result of 539 (ig/L is likely erroneous because the raw, feed (ST1), and treated water
samples on that date all had aluminum concentrations of less than the MDL. No aluminum was
detected in the treated water during the test. This data indicates that the media was not releasing
aluminum to the treated water above detectable levels. The feed and treated water aluminum
concentrations are shown in Figure 4-28.
600
500
400
300
100
Date of Media Regeneration
1/29/04
Date
Raw Feed(STl) ~*~Feed(ST2) Total Feed (ST2) Soluble ~B~Treated (ST3) ^"Regeneration I
Figure 4-28. Capacity Test Aluminum Concentration.
87
-------
As shown in Figure 4-29, the raw water total iron concentration was high, averaging 295 [ig/L
during the Capacity Test, which is near the iron SMCL of 300 |J,g/L. Raw water total iron
concentrations were also highly variable, ranging from 81 to 2370 [ig/L. The impact of
precipitation events may have contributed to the variability of the raw water iron concentration.
4/28/04
5/28/04
E
"Raw Total
Raw Soluble '
Regeneration I
Figure 4-29. Capacity Test Raw Water Iron Concentration.
88
-------
As shown in Figure 4-30, similar to the raw water, feed water iron concentrations were highly
variable. However, all feed water iron spikes do not necessarily correspond to raw water iron
concentration spikes. This indicates that the chlorine detention tank may have had significant
effects on the iron concentration, possibly allowing particulate iron to settle when the well pump
was off and having iron resuspended during well pump operation.
The treated water total iron concentration averaged 68 |ig/L, which is much less than the raw
water total iron concentration of 295 |ig/L and the feed (ST2) concentration of 180 |ig/L,
indicating significant removal of iron by the treatment unit. However, the average treated water
soluble iron concentration of 50 jig/L was actually greater than the average feed (ST2) soluble
iron concentration of 38 |ig/L. This is likely a result of the acid addition for pH reduction
resulting in an increase in the metal solubility.
4/28/04
5/28/04
|~*~Feed (ST1) ~*~Feed (ST2) Total ~B~Treated (ST3) Total ^"Regeneration I
Figure 4-30. Capacity Test Feed Water Iron Concentration.
89
-------
The total and soluble iron data, shown in Figure 31 and Figure 32, indicates that iron removal in
the treatment unit occurred primarily by filtration of particulate iron. Figure 4-31 shows that
much of the variability in the feed water iron concentration is a result of variable particulate iron
concentrations. The feed (ST2) total iron concentration averaged 180 |ig/L, as compared to the
feed (ST2) soluble iron average concentration of 38 |ig/L. Feed (ST2) soluble iron
concentrations were more stable throughout the Capacity Test relative to the total iron
concentration.
1200
4/28/04
5/28/04
|~*~Feed (ST2) Total—*~Feed (ST2) Soluble ^"RegenerationI
Figure 4-31. Capacity Test Feed (ST2) Total and Soluble Iron Concentration.
90
-------
The treated water total and soluble iron concentrations are shown in Figure 4-32. The total
treated water iron concentration averaged 68 |ig/L, with 50 |ig/L of soluble iron. Soluble iron
composed 74% of the total treated water iron concentration as compared to the feed (ST2) water
in which the soluble fraction made up 21% of the total iron concentration. It appears that the
treated water iron spikes do not correlate to feed water iron spikes, known weather events, or
operational events (such as backwash or pH adjustment) and can only be speculated to be caused
by breakthrough of iron particles or release of filter material.
Date
["•"Treated (ST3) Total ~~-~Treated (ST3) Soluble ^"Regeneration |
Figure 4-32. Capacity Test Treated Water Iron Concentration.
91
-------
The raw water manganese concentration was generally stable, with the exception of two periods
of paniculate manganese spikes, as shown in Figure 4-33. The raw water manganese
concentration was about double the SMCL of 50 |j,g/L and averaged 105 ng/L, of which 93% (98
was soluble.
1/29/04
Date
I * Raw Total ~"~Raw Soluble Regeneration I
Figure 4-33. Capacity Test Raw Water Manganese Concentration.
92
-------
The feed (ST1) manganese concentration was variable and was consistently greater than the raw
and feed (ST2) concentration, as shown in Figure 4-34. The feed (ST1) average manganese
concentration of 324 (ig/L was skewed by a sample with a concentration of 2,390 ng/L. The
treated water manganese concentration averaged 16 (ig/L during the Capacity Test, indicating
significant manganese removal within the treatment unit.
10/1/03
10/31/03
4/28/04
5/28/04
E
•Raw Total ~*~Feed (ST1) ~*~Feed (ST2) Total ~*~Treated (ST3) Total ^"Regeneration |
Figure 4-34. Capacity Test Manganese Concentration.
93
-------
As shown in Figure 4-35, the variability of the feed (ST2) total manganese concentration is
primarily due to the variability in paniculate manganese concentration. The feed (ST2) soluble
manganese concentration was relatively stable when compared to the feed (ST2) total manganese
concentration. The feed (ST2) total manganese concentration averaged 140 |ig/L, 33% greater
than the raw water, with 52 |ig/L in the soluble form.
10/1/03
|~*~Feed (ST2) Total —*~ Feed (ST2) Soluble ~*~ Treated (ST3) Total ^^Regeneration |
Figure 4-35. Capacity Test Feed (ST2) and Treated Manganese Concentration.
94
-------
The treatment unit removed soluble manganese to less than the detection limit of 10 |ig/L, as
shown in Figure 4-36. However, some particulate manganese did pass through the filter,
resulting in total manganese concentration spikes in the treated water.
I ~B~Treated (ST3) Total ~ Treated (ST3) Soluble Regeneration I
Figure 4-36. Capacity Test Treated Water Manganese Concentration.
95
-------
Chloride concentrations were unaffected by the treatment process, as shown in Figure 4-37. The
feed and treated chloride concentrations averaged 36.9 mg/L and 37.0 mg/L, respectively, during
the Capacity Test and were greater than the chloride concentrations in the two raw water
samples, likely as a result of the sodium hypochlorite addition. A noticeable decrease in chloride
concentration was observed in late April and early May as a result of the previously described
HTWSA hypochlorite feed pump malfunction.
40
35
34
33
31
30
10/1/03
Date of Media Regeneration
10/31/03
11/30/03
12/30/03
1/29/04
Date
2/28/04
3/29/04
4/28/04
5/28/04
* Ra
"Feed(ST2)
Treated (ST3) '
Regeneration I
Figure 4-37. Capacity Test Chloride Concentration.
96
-------
Sulfate concentrations were also unaffected by the treatment process during the early part of the
Capacity Test. However, as shown in Figure 4-38, during the majority of the Capacity Test, the
treated water sulfate concentration was greater than the acidified feed (ST2) sulfate
concentration. A sulfate increase from the raw or feed (ST1) to the feed (ST2) or treated water
was expected as a result of the sulfuric acid addition. However, during the Capacity Test, the
feed (ST2) sulfate concentration averaged 155 mg/L and the treated water sulfate concentration
averaged 160 mg/L, which is within the 95% confidence interval calculated for the feed (ST2)
water.
Raw ~*~Feed (ST2) ~B~ Treated ^"Regeneration I
Figure 4-38. Capacity Test Sulfate Concentration.
97
-------
As shown in Figure 4-39, feed (ST2) phosphorous concentrations were low, only slightly greater
than the MDL for some samples. However, phosphorus was removed from the feed water to
below the MDL of 0.010 mg/L in the treated water by the media for much of the Capacity Test.
During the last several weeks of the test, following the media regeneration, feed (ST2)
phosphorus concentrations increased, as did the treated water concentrations of phosphorus.
Some removal of phosphorus continued to occur within the treatment unit during the time of
increased feed (ST2) phosphorus concentrations.
0.14
0.12
ep o os
Date of Media Regeneration
0.02
*=-*=4^-*=i
Date
I * Raw * Feed (ST2) • Treated (ST3) Regeneration I
Figure 4-39. Capacity Test Phosphorus Concentration.
98
-------
4.4.5 Capacity Test Laboratory Arsenic Analyses
The results of Capacity Test arsenic analyses performed by the NSF laboratory are summarized
in Table 4-14. NSF Laboratory arsenic analyses for the period following media regeneration are
summarized in Table 4-15.
For calculation of the media's capacity to remove arsenic from the feed water, 434,107 gallons
were treated from October 8, 2003 through April 22, 2004 during the Capacity Test. The treated
water volume represents 25,231 media bed volumes, based on the calculated bed volume of 2.3
ft3. At the request of the manufacturer, a one-week period of operation with a reduced feed
(ST2) pH was conducted prior to the media regeneration. In response to the reduction in feed
(ST2) pH, the treated water arsenic concentration was reduced during this period, as shown later
in this section. However, the period of reduced pH operation was not included in the calculation
of the media capacity. Based on the feed and treated water total arsenic concentrations during
the Capacity Test, the capacity of the media for this system was 465.3 |j,g of arsenic per gram of
media. As shown in Table 4-14, the feed water total arsenic concentration of 21 (ig/L included 8
Hg/L of arsenic in the particulate form. The mechanisms for removal of this particulate arsenic
are not clear based on the test data, and could include adsorption and/or physical filtration of the
particulate arsenic. The treated water arsenic was nearly all in the soluble form. As discussed in
Section 4.5, the backwash water arsenic concentration averaged 539 mg/L. Based on 14
backwashes of 103 gallons each during the Capacity Text, the arsenic removed by each
backwash represents only a fraction of the total arsenic removed from the feed water. If the
arsenic in the backwash water is removed from the capacity calculation, the capacity of the
media for arsenic in this system is reduced to 406 mg of arsenic per gram of media.
The media regeneration effectively returned the arsenic adsorption capacity of the media to
approximately that of the new media. Within one day following media regeneration, the treated
water total arsenic concentration was 5 ng/L. However, as shown in Tables 4-14 and 4-15 two
post-regeneration samples, taken within several hours of returning the unit to service, had very
high arsenic concentrations and skewed the post-regeneration arsenic results. The effect of
media regeneration on treated water arsenic concentrations and the high arsenic concentrations
observed immediately following the regeneration are shown and discussed in more detail later in
this section.
99
-------
Table 4-14. Capacity Test Laboratory Arsenic Analyses
Raw
Number of
Samples
Mean
Minimum
Maximum
Standard
Deviation
95%
Confidence
Interval
Total Soluble
Arsenic Arsenic
(ug/L) (ug/L)
87 72
16 16
8 7
24 20
3 3
15-17 15-16
Number of
Samples
Mean
Minimum
Maximum
Standard
Deviation
95%
Confidence
Interval
Arsenic
III
(Mg/L)
13
7
<2
14
4
4- 10
Total
Arsenic
(ug/L)
121
7
2
20
N/A
N/A
Calculated
Arsenic V
(ug/L)
13
6
2
12
3
4-8
Total
Arsenic
(ug/L)
121
21
12
63
8
19-23
Treated
Soluble Arsenic
Arsenic III
(ug/L) (ug/L)
65 13
8 <2
3 <2
19 <2
N/A N/A
N/A N/A
Feed (ST2)
Soluble Arsenic Calculated
Arsenic III Arsenic V
(ug/L) (ug/L) (ug/L)
65 13 13
13 <2 10
8 <2 6
18 <2 16
203
12-13 <2 - <2 8 - 12
Calculated
Arsenic V
(ug/L)
13
4
1
10
N/A
N/A
N/A = Statistics not appropriate for treated water arsenic concentrations.
100
-------
Table 4-15.
Post-Regeneration
Laboratory Arsenic Analyses
Raw
Number of
Samples
Mean
Minimum
Maximum
Standard
Deviation
95%
Confidence
Interval
Total Soluble
Arsenic Arsenic
(ug/L) (ug/L)
15 16
15 15
8 9
18 18
3 3
13-17 13-17
Number of
Samples
Mean
Minimum
Maximum
Standard
Deviation
95%
Confidence
Interval
Arsenic
III
(Mg/L)
1
13
13
13
N/A
N/A
Total
Arsenic
(ug/L)
15/13(1)
18/4(1)
2
200/6(1)
N/A
N/A
Calculated
Arsenic V
(ug/L)
1
3
3
3
N/A
N/A
Total
Arsenic
(ug/L)
15
16
13
20
2
15- 17
Treated
Feed
Soluble
Arsenic
(ug/L)
16
14
10
19
2
13- 15
(ST2)
Arsenic Calculated
III Arsenic V
(ug/L) (ug/L)
1 1
<2 11
<2 11
<2 11
N/A N/A
N/A N/A
Soluble Arsenic Calculated
Arsenic III
(ug/L) (ug/L)
16/14(1) 1
16/4(1) <2
2 <2
180/5(1) <2
N/A N/A
N/A N/A
Arsenic V
(ug/L)
1
2
2
2
N/A
N/A
A brief treated water arsenic spike occurred within 6 hours following regeneration. The
statistics are presented both with and without the two treated water samples that had elevated
arsenic concentrations.
N/A = Statistics not calculated for sample sets of less than 8. Statistics not appropriate for
treated water arsenic concentrations.
101
-------
During the Capacity Test, the raw water total arsenic concentration averaged 16 jig/L and was
primarily soluble arsenic, which also averaged 16 |ig/L. Based on the 13 samples speciated for
arsenic, the soluble arsenic was comprised of variable fractions of arsenic III and arsenic V.
Capacity Test and post-regeneration raw water arsenic concentrations are shown in Figure 4-40.
1/29/04
Date
2/28/04
3/29/04
5/28/04
"Total P SolubleArsenic IIIArsenic V
Figure 4-40. Capacity Test Raw Arsenic Concentration.
102
-------
The feed (ST2) total arsenic concentration averaged 21 jig/L and was highly variable, especially
during the first third of the Capacity Test, as shown in Figure 4-41. The feed (ST2) total arsenic
concentration was 24% greater than the raw water total arsenic concentration. However, the feed
(ST2) average soluble arsenic concentration of 13 jig/L was approximately 19% less than the
average raw water soluble arsenic concentration. The increase and variability in feed (ST2) total
arsenic concentration may have been a result of resuspension of particulate arsenic in the
chlorine contact tank by the well pump operation. As shown in Table 4-14, Table 4-15, and
Figure 4-41, the feed (ST2) soluble arsenic was completely converted to the arsenic V species
prior to entering the treatment unit.
10/1/03
10/31/03
11/30/03
12/30/03
1/29/04
Date
3/29/04
4/28/04
Total H Soluble Arsenic III Arsenic V
Figure 4-41. Capacity Test Feed (ST2) Arsenic Concentration.
103
-------
The treated water arsenic concentration during the Capacity Test averaged 7 |ig/L, all of which
was soluble arsenic. The post-regeneration treated water arsenic concentration averaged 18
Hg/L. As shown in Figure 4-42 and 4-43, the mean post-regeneration treated water total arsenic
concentration was skewed by two samples with very high arsenic concentrations, which occurred
immediately following the media regeneration. Two treated water arsenic samples collected
within eight hours of the media regeneration had arsenic concentrations of 200 and 17 |ig/L,
respectively. The sample collected the day after media regeneration had an arsenic concentration
of 4 |ig/L. The post-regeneration treated water arsenic concentration ranged from 2 to 6 |ig/L,
with the exception of the two samples. Arsenic in the treated water was primarily soluble arsenic
in the arsenic V species. Following the regeneration process, the arsenic removal capacity of the
media appeared to return to a level approximately equal to that of the new media, the elevated
treated water arsenic concentrations for several hours following the media regeneration could be
problematic in a full-scale operation and may need to be addressed in the manufacturer's
regeneration and operating procedures.
250
200
< 100
Date of Media Regeneration
-V
10/1/03 10/31/03 11/30/03 12/30/03
1/29/04
Date
4/28/04
I * Total B Soluble Arsenic III Arsenic V |
Figure 4-42 Capacity Test Treated Water Arsenic Concentration.
104
-------
5/28/04
|
Date of Media Regeneration
Bed Volumes Treated
I ~*~ Total~*~ Soluble Arsenic III Arsenic V I
Figure 4-43. Capacity Test Treated Water Arsenic Concentration (0-25 ug/L Scale).
105
-------
Raw, feed, and treated water total arsenic concentrations are shown for the Capacity Test in
Figure 4-44.
Date of Media Regeneration
"Raw~*~Feed (ST2) """"Treated (ST3) |
Figure 4-44. Capacity Test Total Arsenic Concentration.
106
-------
The manufacturer indicated that maintaining the specified feed (ST2) and treated water pH is
critical to achieving efficient arsenic removal. As shown in Figure 4-45, feed (ST2) and treated
water pH does appear to have a significant impact on the arsenic removal capacity of the media.
For example, near December 30, 2003, when optimal pH was not maintained due to acid feed
pump operational problems, the treated water arsenic concentration increased noticeably in
conjunction with the increase in treated water pH.
100
Following Media Regeneration.
7.5
10/1/03 10/31/03 11/30/03 12/30/03 1/29/04 2/28/04
Date
3/29/04 4/28/04
Feed (ST2) Total Arsenic ~»~Treated (ST3) Total Arsenic ~*~ Feed (ST2) pH ~*~Treated (ST3) pH I
Figure 4-45. Capacity Test Arsenic Concentration and pH.
107
-------
Figure 4-46 shows feed (ST2) and treated water arsenic concentration as well as feed (ST2) and
treated water FAC concentration. FAC oxidizes arsenic III to the arsenic V species, which the
manufacturer indicates can be removed by MEDIA G2
®
As shown in Figure 4-46, the feed (ST2) FAC concentration was generally maintained above
0.5 mg/L, and averaged 0.75 mg/L, which was adequate to convert the arsenic to the arsenic V
species. HTWSA experienced problems with the sodium hypochlorite feed pump in late April
2004, and the FAC concentration was reduced to near zero. However, the media was already
nearing "exhaustion", with treated water concentrations near 10 |ig/L. The media was
regenerated during the period of low FAC concentration. Following the regeneration, the media
produced treated water with an arsenic concentration of 5 |ig/L through May 9, 2004, when a
continuous FAC concentration was reestablished in the feed water. Although the treatment unit
arsenic removal efficiency did not appear to be affected by the low FAC concentration, because
the low FAC concentration occurred at a time when the media was nearing exhaustion and then
newly regenerated, conclusions on the long-term effect of low or no FAC concentration in the
feed water cannot be made.
100
Following Media Regeneratio
|~*~Feed(ST2) Total Arsenic ~B~Treated (ST3) Total Arsenic Feed (ST2) FAC ~*~Treated (ST3) FAC I
Figure 4-46. Capacity Test Arsenic and FAC.
108
-------
4.4.6 Capacity Test Equipment Operation
During the Capacity Test, minimal time and/or attention were required to operate the pilot test
equipment, although significant time was spent conducting on-site analyses. The time required
for daily operation of the treatment unit included approximately ten minutes to monitor the flow
rate, acid tank level, totalizer and loss of head readings, and verification that there were no leaks
in the system. Periodically refilling the acid batch tank required additional time. Operational
problems with the acid feed pump required significant operational attention during those events,
as pH adjustment is a critical parameter to maintain the treatment system performance.
Permanent installation of the equipment would also require daily pH and FAC analyses or online
monitoring equipment, as well as periodic on-site arsenic analyses and/or collection of samples
for laboratory analyses. FAC and pH analyses require 15 to 20 minutes. On those days in which
on-site arsenic analyses with a field test kit are also performed, the total analytical time is about
45 minutes.
The pilot adsorption media filter is manually operated and uses electricity only for powering the
feed water solenoid valve and the chemical feed pumps for metering sulfuric acid, sodium
hypochlorite, and caustic soda. Chemical, electricity, and media consumption are described in
Section 2.3.2. A total of 590.5 gallons of 0.5% sulfuric acid solution were used during the test
for pH adjustment. This corresponds to approximately 1.16 gallons of sulfuric acid per 1,000
gallons treated. Fifty gallons of 1% caustic soda were used during the media regeneration,
corresponding to 0.11 gallons per 1,000 gallons treated. Sodium hypochlorite was used by
HTWSA for disinfection of the well supply and was not fed as part of the est. Therefore,
sodium hypochlorite usage was not measured. The electricity used by the chemical feed pumps
and solenoid valve was not measured as part of the test.
One to 1.5 hours of operator time was required for each manual backwash of the filter. The
adsorption media is regenerated in place and requires the operator to backwash the filter, to
prepare and feed a caustic soda solution and a sulfuric acid solution, and to monitor the effluent
pH before returning the unit to service. Including a pre-regeneration backwash, media
regeneration required about five hours of operator time. Media regeneration is discussed in
Section 4.6.
4.5 Capacity Test Backwash Water Quality, Quantity, and Flow Rate
Fifteen manual filter backwashes were performed during the Capacity Test. Filter backwashes
were performed twice per month, based on the filter approaching the maximum allowable
pressure drop of 10 psi, as indicated by the manufacturer. The first filter backwash was
performed during the second week of the Integrity Test. This backwash was performed prior to
reaching the time or pressure drop criteria for initiating a backwash in order to fulfill the
requirements of the Integrity Test plan. The sampling protocol and flow rate calibration
procedures detailed in Chapter 3 were followed for both the backwash and rinse wastewater.
The results of the analyses of composite samples for these wastewaters are presented on
Table 4-16.
The backwash water was turbid and dark brown in color, likely due to the very high
concentration of iron and manganese detected in the laboratory sample. The high concentrations
109
-------
of iron and manganese in the backwash water indicate iron and manganese were physically
filtered from the water. This was anticipated given that much of the feed water iron and
manganese was in the particulate form. The aluminum and silica concentrations in the
wastewater were much greater than the average treated water concentrations, indicating that
silica and aluminum were displaced from the media during backwash. In addition, phosphorus
concentrations in the backwash water samples were much greater than the feed water
concentrations.
The backwash water arsenic concentration averaged 539 ng/L. However, the soluble arsenic
concentration in the backwash water averaged 12 |J,g/L, which is equal to the feed (ST2) soluble
arsenic concentration. This indicates that the elevated arsenic concentration in the backwash
water is a result of the removal of particulate arsenic accumulation from the filter, rather than
desorption of arsenic from the media.
The manual filter backwash procedure described in the manufacturer's operating instructions
was generally followed during backwash and rinse. The manufacturer's Operations Manual
indicates that the filter should be backwashed for 15 minutes at a flow rate of 3.2 gpm initially
during startup, but should be backwashed for 15 minutes at a rate of 3.9 gpm after initial startup.
However, a backwash rate of 3.2 gpm was inadvertently used for the first two backwashes after
initial startup.
The manufacturer's operating instructions state that the filter should be rinsed until the rinse
water is clear (approximately 15 minutes) at the normal operating flow rate of 1.7 gpm. At the
specified rinse rate, variable rinse durations were required for the rinse water to clear. Therefore,
the volume of rinse water used varied from 48 to 110 gallons per backwash.
Wastewater from each filter backwash and rinse was discharged to a sanitary sewer adjacent to
the well station. The total water usage for each backwash and rinse was 200 gallons, for a total
backwash and rinse water usage of 2,800 gallons. The backwash and rinse water usage
represents 0.5% of the total throughput of 519,400 gallons during the test, including the Integrity,
Capacity, and Post-Regeneration phases.
110
-------
Table 4-16.
Volume
Flow Rate
pH
Turbidity
Arsenic
Iron
Manganese
Aluminum
Silica
Sodium
Chloride
Sulfate
Phosphorus
Alkalinity
FAC
Calcium
Hardness
Fluoride
Capacity
Units
gallons
gpm
unit
NTU
H8/L
H8/L
Hg/L
H8/L
mg/L
mg/L
mg/L
mg/L
mg/L
mg/L
mg/L
mg/L
mg/L
mg/L
Test Backwash and Rinse ^
Number
of Backwash
Samples Average
15
15
2
1
13
4
4
4
4
4
3
3
3
2
1
1
1
1
103
3.6
7.3
16.6
539
49,348
22,775
4,290
74.3
25.4
36.8
152
1.07
166
1.84
274
306
0.49
Water Characteristics
Number
of Rinse
Samples Average
14
14
2
1
13
4
4
4
4
4
3
3
3
2
1
1
1
1
90
1.7
7.1
0.68
16
1,180
370
259
34.4
25.3
36.4
126
0.020
165
0.72
316
316
0.50
4.6 Media Regeneration
Spent MEDIA G2® media must be regenerated by the operator by feeding an alkaline caustic
soda solution to the media. The increase in pH above 10 with the addition of the alkaline solution
causes the previously adsorbed arsenic to solubilize and release from the media. Following this
step, the operator is required to feed sulfuric acid to the filter to neutralize the high pH caused by
the caustic solution. Media regeneration was performed once during the Capacity Test. The
combination of filter backwash and regeneration required about five hours of operator time,
including setup and sample collection. Regeneration wastewater quality parameters are
summarized in Table 4-17. The arsenic concentration in the regeneration and regeneration rinse
waters are summarized in Table 4-18. Concentrations of sodium, silica, iron, manganese, and
aluminum in the backwash and rinse waters far exceeded feedwater concentrations, indicating
that the regeneration process had resulted in their removal from the filter media.
Regeneration produced approximately 50 gallons of regenerant wastewater. A portion of mixed
regenerant wastewater was collected for TCLP analysis. The TCLP sample was filtered and
analyzed according to the TCLP protocol. The results of TCLP analyses are summarized in
Table 4-19. It is important to note that the manufacturer's standard procedure for treating spent
regenerant water before discharge was not followed. The standard procedure is to adjust pH to
6.0, causing the arsenic and iron to form insoluble ferric arsenate, which will then settle out of
solution. The manufacturer reports that, if the standard procedure is followed, the settled sludge
and the supernatant will pass the TCLP test.
Ill
-------
The arsenic concentration in the media regeneration wastewater TCLP analysis is greater than
the TCLP regulatory limit. However, this does not imply that the spent media will not pass the
TCLP test for disposal of the spent media. The ability to discharge regeneration wastewater with
a relatively high arsenic concentration and potential impacts on the receiving wastewater
treatment plant should be evaluated on an individual basis. Laboratory data qualifications for the
arsenic results are included with the Wastewater TCLP Analytical Reports in Appendix M. In
addition, wastewater was generated following regeneration due to rinsing of the filter bed until
the treated water is within one pH unit of the feed water. Both of these wastewaters were
discharged to the sanitary sewer during the media regeneration performed during the Capacity
Test.
The media regeneration procedure is intended to remove arsenic from the media. Therefore,
since the pilot system was operated for only one month following media regeneration, which
would not have allowed for the accumulation of a significant quantity of arsenic, performing
TCLP and California Waste Extraction Tests (CA WET) analyses on the media would not have
been representative of the leaching characteristics of fully spent media. Therefore, TCLP and
CA WET analyses were not performed, as planned, following the end of the post-regeneration
testing.
112
-------
Table 4-17. Regeneration Wastewater Quality
Sodium Silica Iron Manganese Aluminum
Date Time (mg/L) (mg/L) (Hg/L) (ng/L) (ng/L)
4/30/2004
4/30/2004
4/30/2004
4/30/2004
4/30/2004
4/30/2004
4/30/2004
11:31
11:41
11:51
12:25
13:00
13:10
13:30
24.8
35.9
230
1,110
5,160
3,650
4,320
30.8
71.0
385
454
1,686
392
907
26
22
68
86
147
190,000
23,300
<10 <200
<10 <200
<10 2,900
<10 5,370
<10 16,300
2,440 168,000
329 32,900
Description
Regeneration Waste
Regeneration Waste
Regeneration Waste
Regeneration Waste
Regeneration Rinse
Regeneration Rinse
Regeneration Rinse
Table 4-18. Regeneration Wastewater Arsenic Concentration
Arsenic
Date Time (^lg/L) Description
4/30/2004 13:00
4/30/2004 13:10
4/30/2004 13:20
4/30/2004 13:30
5,000 Regenerate Rinse
1,800 Regenerate Rinse
11,000 Regenerate Rinse Mix
5,900 Regenerate Rinse/Composite
Table 4-19. Media Regeneration Wastewater Characterization
Parameter
Arsenic
Barium
Cadmium
Chromium
Copper
Lead
Mercury
Selenium
Silver
Zinc
TCLP Result
Result
(mg/L)
10.3
ND
0.068
ND
ND
ND
ND
ND
ND
ND
Reporting Limit
(mg/L)
0.40
0.40
0.020
0.160
0.040
0.20
0.0004
0.40
0.020
0.40
TCLP(1)
Regulatory Limit
(mg/L)
5.0
100.0
1.0
5.0
N/A
5.0
0.2
N/A
5.0
N/A
^^ 40 CFR 261.24 Toxicity Characteristics.
ND = Non-Detect.
Media gradation analyses indicate that following the test the media contained significantly less
fine material than new media. The loss of fine-grained material could have occurred during
normal operation or more likely during media backwashes. Media gradation reports are included
in Appendix N.
4.7 Task 3: Documentation of Operating Conditions and Treatment Equipment
4.7.1 Introduction
During each day of verification testing, the arsenic adsorption media filter operating conditions
were documented. The volumetric flow rate and feed water pH through the MEDIA G2® are
113
-------
both critical parameters, and were monitored and documented. MEDIA G2® performance is
affected by the EBCT, which varies directly with the volumetric flow rate through the vessel.
The MEDIA G2® performance was also shown to be sensitive to feed water pH, which varied
directly with the quantity of acid that was metered to the feed water.
4.7.2 Experimental Objectives
The objective of this task was to accurately and fully document the operating conditions and
performance of the equipment, as stated in Section 3.11. This task was performed in conjunction
with both the system Integrity Test and the Capacity Test, as presented in Sections 4.3.3 and
4.4.2, respectively.
4.8 Task 4: Data Management
The data management plan was executed as presented in Task 4 (Data Management), located in
Section 3.12. Data were entered into computer spreadsheets and submitted in electronic and hard
copies. QA/QC forms, field notebooks, and photographs are included in the appendices of this
report.
4.9 Task 5: Quality Assurance/Quality Control
4.9.1 Introduction
Appropriate quality assurance and quality control measures were performed to ensure the quality
and integrity of all measurements of operational and water quality parameters during the ETV
testing. QA/QC procedures for the operation of the arsenic adsorption media filter and the
measured water quality parameters were maintained during the verification testing program, as
specified in the test plan and described in Section 3.13.
On-site QA/QC activities were recorded in the logbooks, included as Appendix J. QA/QC
efforts included review of laboratory raw data (run logs and bench sheets); calibration of on-site
analytical instrumentation; calibration of totalizer meters; calibration of the flow meter; analyses
of split samples to verify Hach Test Kit analyses for alkalinity, calcium, and hardness; pressure
gauge calibration; collection of duplicate samples for on-site and laboratory analyses; and spiked
sample analyses. Performance evaluation analyses were also performed by Gannett Fleming to
demonstrate proficiency and accuracy of the analytical equipment and laboratory techniques
required for all on-site water quality analyses. All data entry performed by the field engineer
was checked by a second person.
An on-site system inspection and audit for sampling activities and field operations was
conducted by NSF. The Gannett Fleming QA officer also conducted an on-site inspection during
the first two weeks of operation.
114
-------
4.9.2 Data Quality Indicators
Data quality indicators include the following:
• Representativeness;
• Accuracy;
• Precision;
• Statistical Uncertainly; and
• Completeness.
4.9.2.1 Representativeness. Representativeness refers to the degree to which the data accurately
and precisely reflects the conditions or characteristics of the parameter, as measured by the data.
Representativeness was ensured by executing consistent sample collection protocol, by using
each method to its optimum capability to achieve a high level of accuracy and precision, and
collecting sufficient data to be able to detect a change in operations.
4.9.2.2 Accuracy. Accuracy refers to the difference between a sample result and the true or
reference value. Accuracy was optimized through equipment calibrations, collection of split
samples, analysis of performance evaluation (PE) samples, and analysis of spiked samples, as
specified in the PSTP.
4.9.2.2.1 Field Equipment Calibrations. Periodic calibration of field test equipment
included calibration of the pressure gauges, flow meter, totalizer meter, portable
turbidimeter, pH meter, portable colorimeter, and fluoride meter/electrode, as specified in
Table 4-20.
115
-------
Table 4-20. Field Instrument Calibration Schedule
Instrument Calibration Method
Acceptable
Frequency Accuracy
Pressure Gauges
Flow Meter
Totalizer Meter
Portable Turbidimeter
Portable pH/ISE Meter with
Combination pH/Temperature
Electrode
Portable Colorimeter
dead weight calibration tester biannual ± 10%
volumetric "bucket and stop
watch"
volumetric "bucket and stop
watch"
secondary turbidity standards
primary turbidity standards
three-point calibration using
4.0, 7.0 and 10.0 buffers
approximate 4.0 mg/L chlorine
standard
weekly
± 10%
weekly ± 1.5%
daily PE sample
weekly
Thermometer (NIST-traceable) calibration not required
Portable pH/ISE Meter with
Fluoride ISE
0.2 mg/L fluoride standard and
2.0 mg/L fluoride standard
daily
daily
N/A
daily
±5%
±25%
±2%
Calibration tests were performed on the electronic flow meter and totalizer meter before
the initiation of the Integrity Test. These calibration tests indicated accuracy within the
stated ranges of the instrument manufacturers. The calibration data for the electronic flow
meter indicated that a flow rate of 1.7 gpm was produced when the meter indicated 1.7
gpm. The calibration data for the totalizer meter indicated a production of approximately
10 gallons when the meter had an incremental reading of 10 gallons.
4.9.2.2.2 Split Samples. Split samples for alkalinity, calcium, and hardness were
analyzed both on-site by field personnel and by the PADEP Laboratory staff to verify the
accuracy of the Hach methods for on-site analyses of these parameters. The results of
split sample analyses are included in Table 4-21.
Alkalinity analyses were not performed in the field on the day that split samples were
collected for PADEP Laboratory analyses. Therefore, true split sample results for
alkalinity were not available. However, field measured alkalinity was relatively stable
during the Capacity Test. From May 19, 2004, through May 27, 2004, five raw water
alkalinity sample results ranged from 148 mg/L to 152 mg/L. During the same time
period five feed (ST2) results ranged from 124 mg/L to 128 mg/L and five treated water
samples ranged from 112 mg/L to 120 mg/L. Although it does not provide a true split
sample comparison, the average of the five sample results at each of the three locations
was calculated for comparison to PADEP Laboratory results.
116
-------
Split sample results for the raw and treated (ST3) were within the acceptable limits of
accuracy of ±30% established by NSF. However, split sample results for the feed (ST2)
water were generally outside the acceptable limits of accuracy. Feed (ST2) alkalinity
measured on-site was greater than the alkalinity measured in the laboratory and was
slightly outside the limits of acceptable accuracy. The on-site feed (ST2) calcium result
was within acceptable accuracy limits when compared to one PADEP Laboratory result
and outside the acceptable limits of accuracy when compared to a second PADEP
Laboratory result. On-site feed (ST2) hardness results were not within the acceptable
limits of accuracy as established by NSF.
Table 4-21. Split-Samples (May 25, 2004)
Parameter
Alkalinity1^ (mg/L as
CaCO3)
Calcium (mg/L as CaCO3)
Hardness (mg/L as
CaCO3)
Parameter
Alkalinitylu (mg/L as
CaCO3)
Calcium (mg/L as CaCO3)
Hardness (mg/L as
CaCO3)
Parameter
Alkalinitylu (mg/L as
CaCO3)
Calcium (mg/L as CaCO3)
Hardness (mg/L as
CaCO3)
Raw Water
Field
150
PADEP
Lab
164
228
262
%
Difference
-8.5%
PADEP
Lab
162
229
264
%
Difference
-7.4%
Feed (ST2)
Field
126
288
374
PADEP
Lab
94.4
206
241
%
Difference
33.5%
39.8%
55.2%
PADEP
Lab
96.4
229
263
%
Difference
30.7%
25.8%
42.2%
Treated (ST3)
Field
116
278
300
PADEP
Lab
91.8
229
263
%
Difference
26.4%
21.4%
14.1%
PADEP
Lab
92.2
225
259
%
Difference
25.8%
23.6%
15.8%
^ Field alkalinity data is an average of five sample results from May 19, 2004, through May 27, 2004,
because field alkalinity analyses were not performed May 25, 2004.
4.9.2.2.3 Performance Evaluation Samples for Water Quality Testing. PE samples are
samples of known concentration prepared by an independent performance evaluation
laboratory and provided as unknowns to an analyst to evaluate his or her analytical
performance. Analyses of laboratory PE samples were conducted before the initiation of
verification testing. The control limits for the PE samples were used to evaluate the field
analytical method performance.
117
-------
A PE sample comes with statistics that have been derived from the analysis of the sample
by a number of laboratories using EPA-approved methods. These statistics include a true
value of the PE sample, a mean of the laboratory results obtained from the analysis of the
PE sample, and an acceptance range for sample values. The field laboratory and the
PADEP Laboratory provided results from the analysis of the PE samples that meet the
performance objectives of the verification testing. PE sample results for the PADEP
Laboratory and the results of PE checks for on-site water quality parameters are included
in Appendix O.
4.9.2.2.4 Spike Sample Analyses. Analyses of matrix spikes were performed by
Gannett Fleming for on-site water quality parameters during the Capacity Test. Spike
sample analyses results for alkalinity, calcium, hardness, fluoride, and FAC are included
in Appendix J. Calcium spike sample percent recoveries were within the acceptable
accuracy of 70 to 130% recovery, with the exception of one spike sample, which had
only a 55% recovery. It is likely that the poor recovery percentage for this sample is a
result of an error in entering data in the logbook rather than analytical error. The volume
of titrant (1.1 mL) recorded in the logbook for the 40 mg/L spike was the same as the
volume of titrant recorded for the 20 mg/L spike sample. Considered with the fact that
all other calcium spike samples were within the acceptable range of accuracy, recording
error is most likely the cause of the single unacceptable result.
Hardness spike sample percent recoveries were within the acceptable range of accuracy
of 70 to 130% recovery established by NSF. Alkalinity spike sample percent recoveries
were within the acceptable range of accuracy of 70 to 130% recovery, with the exception
of one spike sample, which had a percent recovery of 138%. This sample represents less
than 1% of the alkalinity spike samples performed. However, more than 30% of the FAC
spike samples and nearly 50% of the fluoride spike samples analyzed were not within the
acceptable accuracy range of 70 to 130% recovery.
The results of spike sample analyses performed by the PADEP Laboratory are included
in the laboratory analysis summary tables included in Appendix K. Spike sample
analyses were performed by the PADEP Laboratory at a frequency of 10%. Spike sample
analysis percent recoveries for iron, manganese, aluminum, sodium, and silica were
within the acceptable accuracy range of 70 to 130% recovery. Spike sample results for
chloride and sulfate were within the acceptable accuracy range of 80 to 120% recovery
and total phosphorus was within the acceptable accuracy range of 90 to 110% recovery.
The results of NSF laboratory spike sample analyses for arsenic are included in the
laboratory QA/QC data in Appendix L. Spike sample analyses were performed by the
NSF laboratory at a frequency of 10%. Percent recoveries for arsenic were within the
acceptable accuracy range of 70 to 130% recovery.
4.9.2.3 Precision. Precision refers to the degree of mutual agreement among individual
measurements and provides an estimate of random error and can be measured by replication of
analyses. The precision levels for all duplicate analyses were calculated.
118
-------
On-site water quality relative percent deviation calculations are included with the On-site Water
Quality Data in Appendix J. Relative percent deviations calculated for pH, FAC, temperature,
alkalinity, calcium, hardness, and fluoride were all within the acceptable precision level of 30%.
Approximately 10% of duplicate turbidity samples were not within the acceptable precision
level.
Relative percent deviation calculations for PADEP Laboratory duplicates are included in
Appendix K. The PADEP Laboratory performed duplicate analyses at a 10% minimum
frequency. All PADEP Laboratory duplicate analyses were within the acceptable levels of
precision of 30% for iron, manganese, aluminum, sodium, and silica; 20% for chloride and
sulfate; and 10% for total phosphorus. No duplicates of the parameters analyzed by the PADEP
Laboratory were collected in the field.
NSF relative percent deviation calculations for laboratory arsenic duplicates are included in
Appendix L. All NSF laboratory arsenic duplicate analyses were within the acceptable precision
level of 30%. All field duplicates of arsenic samples were within the acceptable precision level
of 30%.
4.9.2.4 Statistical Uncertainty. Statistical uncertainty of water quality parameters (for data sets
of eight or more parameters) was evaluated through the calculation of the 95% confidence
interval around the sample mean.
4.9.2.5 Completeness. Completeness refers to the amount of valid, acceptable data collected
from a measurement process compared to the amount expected to be obtained. The completeness
objective for data generated during this verification test was based on the number of samples
collected and analyzed for each parameter and/or method, as defined below.
Number of Samples Per
Parameter and/or Method
0-10
11-50
>50
Percent Completeness
80%
90%
95%
Completeness was defined as the following for all measurements:
%C = (V/T) X 100
where: %C = percent completeness
V = number of measurements judged valid
T = total number of measurements
Calculation of data completeness was made for on-site water quality measurements, PADEP
Laboratory water quality measurements, and arsenic measurements. These calculations are
presented in Appendix J, K, and L of this report, respectively.
119
-------
During the Integrity Test, the completeness percentages for on-site water quality analyses were
within the acceptable completeness levels; however, duplicates of some parameters were
collected at a rate of 50% or 0% of that proposed. During the Capacity Test, the completeness
percentages for on-site water quality analyses were within the acceptable completeness levels.
Duplicates were also performed within the acceptable completeness levels, with the exception of
feed (ST1) pH and turbidity, which were performed with a completeness of 88% and 50%,
respectively. Additional on-site water quality samples, not specified in the test plan, were
analyzed to better characterize the feed (ST1) and raw waters. Completeness for on-site water
quality analyses during the post-regeneration period were within the acceptable completeness
levels, with the exception of feed (ST1) pH (87%), and feed (ST2) and treated calcium,
magnesium, hardness, and fluoride (33%). Duplicate analyses during the post-regeneration
period were all within the acceptable completeness levels.
During the Integrity Test, samples for PADEP analyses were collected at 100% or greater
completeness of the proposed amount specified in the test plan. However, field duplicates were
not collected. During the Capacity Test, samples for PADEP analyses were within the
acceptable completeness levels; however duplicates were not collected for any of the applicable
parameters. Completeness ranged from 67% to greater than 100% of the proposed sample
frequency specified in the test plan for PADEP parameters during the post-regeneration phase of
the test. Those parameters that were not collected within the acceptable completeness level
include sodium, silica, aluminum, chloride, sulfate, and phosphorus at the feed (ST2) and treated
water locations. Additional samples (that were not specified in the test plan) were collected for
the raw and feed (ST1) water.
Although duplicate arsenic samples were not collected during the Integrity Test, 100% or greater
completeness was achieved for samples and duplicates during the Capacity Test, including the
post-regeneration phase.
For the TCLP analysis of the regeneration wastewater, nickel was also supposed to be analyzed;
however, this parameter was inadvertently missed. All other proposed parameters for TCLP
analysis were analyzed for.
120
-------
Chapter 5
References
The following references were used in the preparation of this report:
ETV Protocol for Equipment Verification Testing for Arsenic Removal. U.S. EPA/NSF
International. April 2002.
Hack Water Analysis Handbook. Hach Company, Loveland, Colorado. 1992.
Standard Methods for Examination of Water and Wastewater. 20th ed., Washington, D.C.,
APHA. 1998.
U.S. EPA Drinking Water Methods for Chemical Parameters. EPA Office of Ground Water and
Drinking Water. EPA Methods are available from the National Technical Information Service
(NTIS).
121
-------
Chapter 6
Vendor Comments
ADI International Inc. submitted the following comments concerning the ETV test and report.
These statements were not validated in the verification test and are the opinion of ADI
International Inc.:
"Results were as expected. Despite the relatively poor quality of the feed water (i.e., relatively
high and variable turbidity, high manganese concentration, and the possibility of carry-over of
accumulated iron particulate from the chlorine contact tank to the filter), consistent reduction of
arsenic to below the 10 jig/L maximum contaminant level was obtained.
Feed water manganese and iron concentrations both consistently exceeded secondary maximum
contaminant levels (manganese by as much as a factor of twenty); the filter reduced both to well
below their MCLs. While other iron-based media may be adversely affected by high iron and/or
manganese, MEDIA G2® is not (and in fact is an excellent filter for their removal, as was shown
in this study and in numerous full-scale installations).
Throughout the majority of the capacity test the pH of the feed water was held in the desired 6.3
- 6.5 range. One significant variation occurred during the final week of 2003 and first week of
2004, when the pH increased to over 7.0. A corresponding increase in treated water arsenic
concentration was seen at this time, but the concentration immediately declined again when pH
was brought back under control. In a full-scale plant, automated pH control is used and such
fluctuations should not occur.
During this test, raw water was used for backwashing. While acceptable, this is not ideal. Raw
water of course contains arsenic, and its introduction into the gravel layer at the filter bottom
during backwash means there is potential for an elevated arsenic output when the filter is put
back into normal (downflow) service. In a full-scale plant, if treated water is used, this would
not be the case, and less rinsing would be required to ensure steady low values of arsenic in the
treated water.
As shown in the testing of backwash water, the media holds the arsenic tightly, resulting in a
non-hazardous residual suitable for sewerage. Manganese and iron were readily backwashed
from the media and meaning little accumulation within the filter bed. If no sewer is available,
the backwash can be collected in a holding tank, where the solids will settle, and the supernatant
can be slowly blended back into the raw water entering the plant.
The media was regenerated after seven months of operation. This was premature, as the media
had not yet reached arsenic saturation (treated water was still well below 10 |ig/L). However,
due to time constraints, a regeneration was carried out anyway so that a sample of regenerant
waste could be obtained. Therefore, the true adsorption capacity of the media was not
determined. The outlet arsenic concentration was not trending upward (in fact it was trending
slightly downward), so the filter could have remained in operation without need of regeneration
for an indefinite period. One important observation is that the performance of the regenerated
media returned to that of new media, as expected.
122
-------
A brief spike in treated water arsenic concentration following regeneration was probably due to
insufficient rinsing. In a full-scale plant this is addressed by rinsing the filters with slightly
acidified water to neutralize traces of residual sodium hydroxide within the filter bed, followed
by rinse-to-waste until on-site testing shows the filter pH has returned to neutral and arsenic
concentration is acceptable. With the acid-dosed water rinse, the time elapsed before returning
to service without fear of arsenic spikes is greatly reduced. The volume of wastewater
production is also greatly reduced.
In reviewing the data on regeneration wastewater characteristics it is extremely important to note
that the proper treatment of this waste was not carried out before TCLP testing. Proper
procedure for dealing with this waste is to lower the pH to 6.0 and allow solids to settle. At this
pH, the arsenic combines with the iron in the waste to form insoluble ferric arsenate, which
settles out of solution. Both the settled sludge and the supernatant pass the TCLP test in all cases
ADI has looked at.
One of the major advantages of this technology is its low operating cost. The consumables used
were sulfuric acid (for pH correction of raw water), chlorine (for oxidation of arsenic and
disinfection), and sodium hydroxide (for media regeneration). The calculated dosage of sulfuric
acid was 73 mg/L. The need for pH correction depends on the particular water chemistry at a
given site; many MEDIA G2® plants require no acid at all. Chlorine was added, at a dosage of
1.2 mg/L; this oxidizes the arsenic and provides disinfection for the water entering the
distribution system. Many plants chlorinate their water anyway, in which case this would not be
considered an additional operating expense. For those that do not already chlorinate, the need
for chlorine for acceptable performance is dependent upon the particular water chemistry at the
site (although the cost of a low dosage is relatively small anyway). Regeneration requires three
bed volumes of 1% sodium hydroxide be passed through the filter to desorb the accumulated
arsenic. The procedure involves injecting 50% NaOH into the raw water upstream of the filters
at a rate equal to l/50th of the raw water flow rate. The amount of waste created is minimal; the
waste volume from this pilot plant was only 50 gallons after treating 434,107 gallons of raw
water (and operation should have continued longer - the plant was regenerated prematurely).
Typical waste volume from full-scale MEDIA G2 plants (including regeneration and monthly
backwashing) is about 0.1% of treated water volume.
Electricity costs during the study were negligible - a few watts to run the metering pumps;
actually, similar sized pumps could be used for full-scale plants.
Based on experience at over two dozen installed MEDIA G2 plants, typical operating cost is
$0.10 - $0.25 per 1000 gallons, including chemicals, electricity, waste disposal, and media
replacement. The lower end of the range would be for cases without pH correction and relatively
low raw water arsenic concentration.
Overall, this testing proved the media's ability to provide excellent removal of arsenic,
manganese and iron, with little operator attention and production of a very small volume of non-
hazardous residuals."
123
------- |