Comprehensive Performance Evaluation Protocol to Address Harmful Algal Blooms and Associated Cyanotoxins


oEPA

United States
Environmental Protection
Agency

Comprehensive Performance Evaluation Protocol to
Address Harmful Algal Blooms and Associated
Cyanotoxins

Prepared By:

U.S. EPA Office of Water

Office of Ground Water and DrinkingWater

Standards and Risk Management Division

Technical Support Center

26 West Martin Luther King Drive

Cincinnati, Ohio 45268

Office of Water (MS-140)

EPA 815-B-22-005

June 2022

-------
TABLE OF CONTENTS

Disclaimer	3

1. Area-Wide Optimization Program (AWOP) and Harmful Algal Bloom/CYANOTOXIN ("HAB")
Treatment Optimization Background	4

2.	Harmful Algal Bloom CPEProtocol	6

2.1	Off-Site Pre-CPE Activities	 8

2.1.1	Site Prioritization and Selection	 8

2.1.2	CPE Coordination	 8

2.2	CPE Activities	10

2.2.1	Pre-Event (Day 1)	10

2.2.2	Entrance Meeting (Day 2)	10

2.2.3	WaterTreatment PlantTour (Day 2)	11

2.2.4	Performance Assessment (Days 2-4)	11

2.2.5	Identification of Performance Limiting Factors (Day 4)	18

2.2.6	Exit Meeting (Day 5)	18

2.3	Post-CPE Activities	19

2.3.1	Final Report	19

2.3.2	PWS Follow-up	19

3.	Implementing HAB Treatment Optimization	20

3.1	Effective Leadership and Management	20

3.2	Adopt Water Quality Goals	20

3.3	Establish a Consistent Sampling Approach	21

3.4	Monitoring for HAB Treatment Optimization	21

Appendices

AppendixA: Pre-CPE Preparation
Appendix B: On-Site Materials
AppendixC: Exit Meetingand Final Report

2

-------
DISCLAIMER

As used in this document, the term "optimization" refers to voluntary efforts on the part of
primacy agencies (typically states) and public water systems (PWSs) to optimize PWS
operations, without significantcapital improvements, often leadingto performance above and
beyond the U.S. Environmental Protection Agency's (EPA's) regulatory requirements. As such,
the contents of this guidance document do not have the force and effect of law and the agency
does not bind the public in any way. The use of the term "should" in this document refers to
recommended actions for those who choose to apply the described, optional optimization
approach. Forthose PWSs that may not be in a position to optimizetheiroperationsin the
near-term, but are simply seeking - as a first step - to improve their operations, many of the
same concepts apply. In the lattercases, the state or PWS may wish to establish alternate,
interim goals that differ from traditional AWOP program goals.

3

-------
1. AREA-WIDE OPTIMIZATION PROGRAM (AWOP) AND HARMFUL ALGAL
BLOOM/CYANOTOXIN ("HAB") TREATMENT OPTIMIZATION BACKGROUND

In the late 1980s the U.S. Environmental Protection Agency (EPA) began its development of a
voluntary national program to optimize surface water treatment plant performance for
protection against drinking water microbial contaminants, such as Giardia and Cryptosporidium.
This program is now known as the Area-Wide Optimization Program (AWOP) and is coordinated
by EPA's Technical Support Center (TSC) in Cincinnati, Ohio. The AWOP approach includes the
Comprehensive Performance Evaluation (CPE), or evaluation phase, and can include a
Comprehensive Technical Assistance (CTA), or performance improvement phase. The
philosophy of the program is to optimize existing public water system (PWS) facilities to achieve
desired performance goals without majorcapital improvements.TheCPE was originally
designed to assess plant performance, administration, and operationsand maintenance
practices to identify factors that may adversely impact the plant's ability to achieve microbial
performance goals (United States Environmental Protection Agency, 2004). For more
information, referto the EPA's handbookon "Optimizing Water Treatment Plant Performance
Using the Composite Correction ProgramThe CPE approach has since been applied to
distribution system optimization and this protocol describes yet a not her application for it.

More frequent occurrence and detection of cyanobacteria, along with the cyanotoxins they
produce, in drinking watersources has become an increasingly importantconcern for some
PWSs. In June 2015, EPA released Health Advisoriesfortwo cyanotoxins: microcystins and
cylindrospermopsin (U.S. EPA 2015a, 2015b), as well as Health Effects Support Documents
(HESDs) for three cyanotoxins: microcystins, cylindrospermopsin, and anatoxin-a (U.S. EPA
2015c, 2015d, 2015e). The Health Advisories include information on health effects, analytical
methods and water treatment. The HESDs provide a comprehensive review of published
literature on physical and chemical properties, environ mental fate, known occurrence
information, and health effects. Additionally, EPA released a set of documents gnd tools
intended to gssist drinking wgter utilities in prepgring for gnd responding to HABs in their source
wgter. Specifically pertinenttothe content of this document is the "Wgter Tregtment
Optimizgtion for Cygnotoxins" document, which is among the tools linked above.

To help address HAB and related cyanotoxin concerns and based on its experience developing
and implementingoptimization tools, EPA partnered with the Ohio Environmental Protection
Agency to develop a CPE approach, hereinafter referred to as a "HAB-based CPE" for simplicity,
to evaluate drinking water treatment plants. The focus of the approach was on plant capability
to treat for cyanotoxins and to identify factorsthat may limittreatment plant performance
duringa source water HAB.This HAB CPE protocol was developed overthe course of four pilot
CPE field events conducted between August 2016 and March 2018 at Ohio water treatment
plants whose source water is impacted by HABs. The purpose of this document is to describe
the process of conducting a HAB CPE, as developed during EPA's pilot project with Ohio EPA for
State drinking water staff or drinking water treatment plant operators.

As shown in Figure 1, the "capable plant optimization model" appliesto HABtreatment
optimization, where the objective is to achieve optimized performance. This is initiated through
process control and utilizing data to help optimize plant operations. Sustaining optimization

-------
requires a capable PWS that has a strong foundation of administration-, design-, and
maintenance support.

* Figure 1: Capable plant optimization model

The HABCPE utilizes theframework of the microbial CPE dueto its applicability to particulate-
removal, includingcyanobacterial cells (and associated "intracellular" cyanotoxins). On-site
studies during the HAB CPE also address the removal of dissolved ("extracellular") cyanotoxins,
with focus on adsorption and oxidationcapabilities.

Additional resources that may be useful for PWSs seeking to improve their operations can be
found at EPA's website, "Building the Capacity of Drinking Water Systems."

5

-------
2. HARMFUL ALGAL BLOOM CPE PROTOCOL

The components of the HAB CPE are shown in Figure 2 and are explained further in the
followingsection.

The main activities of the HAB CPE are typically conducted overa five-day period.Typically, the
CPE is conducted by an evaluation team with a minimum of three members, one of whom
should be designated asthe CPE coordinator.

This document also includes four append ices that provide sup porting documentation for the
CPE:

•	Appendix A: Pre-HAB CPE Activities and Preparation

•	AppendixB:On-Site Materials

•	Appendix C: Exit Meeting and Final Report

6

-------
Pre-CPE activities
(led by CPE Coordinator; CPE team provides input as needed)

1.	Site selection (with CPE team input)

2.	CPE coordination

•Contact system and begin historical data analysis
•Identify CPE teams

•Prepare for on-site studies (source water sampling, jar testing, oxidation hold study,
filter evaluation safety protocol)

•Determine sampling plan and laboratory that will perform sample analyses

I

On-site HAB CPE activities
(conducted by all teams unless noted)

\ >

Performance-limiting factors meeting



>

f





Exit meeting



Figure 2: HAB CPE Protocol Framework

7

-------
2.1 Off-Site Pre-CPE Activities

Activities that occur prior to the CPE include the following:

• Site prioritizationand selection,

• Contactingthe PWS abouttheevent,

• Identifying,organizingCPEteam

• Compiling and assessing historical water quality and PWS data, and

• Preparing for studies that may require a longer duration or more in-depth analysis than
can be completed on-site duringthe week of the CPE.

The first of these activities should be initiated approximately two to three months in advance of
the CPE if possible. A detailed description of the pre-CPE activities is provided below.

2.1.1 Site Prioritization and Selection

States can use their HAB monitoring results, knowledge of source waters impacted by HABs,
and knowledge of disadvantaged communitiesto help guide PWS prioritization and selection.
Examples of other selection criteria are provided in Appendix A. The site selection process
should include a discussion with the candidate PWS to explain the objective and activities of the
CPE and confirm theiravailability.

2.1.2 CPE Coordination

Once a PWS has been selected, the CPE coordinatorand the PWS should addressthe activities
described in the followingsections beforethe CPE.

2.1.2.1 Team Identification

A key component of preparingfor a CPE is identifyingthe CPE evaluationteam coordinatorand
members of the three sub-teams that address administration/financial, design, and
operation/maintenance. Each sub-team will have a leader. Generally, the CPE coordinator also
acts as one of the sub-team leaders. The sub-team leader is responsible for understanding the
CPE protocol, organizing and leading the team's activities, and obtaining the necessary
materials and supplies. Each sub-team should also establish a tentative schedule of activities for
the week.

2.1.2.2 Site Visit and CPE Overview Letter

The CPE coordinator should consider a site visit to the PWS to provide more details about the
CPE and discuss the PWS's role duringthe CPE. Discussion topics may include accessing
historical data, set-up for jar testing or hold study, information needed to support the major
unit process evaluation, and safety considerations associated with the filter entry and
inspection.

-------
The CPE coordinator should send a formal letter to the PWS prior to the CPE. See Appendix A
for an example. In this letter, the PWS should be asked to provide particular data prior to the
CPE and to identify key personnel to assist duringthe CPE, participate in interviews, and
support data collection activities. The critical components of this CPE overview letter are
described below:

•	Agent/cr; The tentative agenda should provide the PWS with a general idea of what to
expect and when.

•	Identification of Key PWS Personnel: Both operators and managers, includingthose
responsible for ma king financial decisions at the treatment plant, will typically be asked
to participate in interviewsand to supportthe historical data collectionactivities.

•	Data, PWS information and other resources: Information that is helpful forthe CPE
team to have priorto arrivingon-site includesthefollowing:

•	Raw, individualsettled, and/ortop-of-filter, individual filtereffluent (IFE), and
combined filtereffluent (CFE) turbidity data

•	Any historical cyanobacteria orcyanotoxin-related monitoringdata (e.g.,
chlorophyll-o, phycocyanin, microcystins, etc.)

•	Chlorine dose and residual, pH, and temperature data, or equivalent information
used to calculate CT

•	Administrative data, such as budget information, capital improvement plans, and
organizational charts

•	Design data, such as as-built plantdrawings

•	Recent sanitary survey report, which may include water quality monitoring methods
and locations, and chemical dosing information

•	HABtreatmentcontingencyoroptimization plans

2.1.2.3	CPE Equipment List

A recommended equipment list is provided in Append ix A. Each of the four pilot HAB CPEs
included source water sampling, jar testing, and other studies that may require particular
equipment. The supplies and approach for sampling may vary by site, so this list should be
updated accordingly.

2.1.2.4	Initiate Historical Data Analysis

Any data received from the PWS priorto the CPE should be compiled electronically, such as in
spreadsheets, for ease of use. For turbidity data, we recommend using EPA's Optimization

9

-------
Assessment Spreadsheet. This spreadsheet helps organize and assess plant settled and filtered
waterturbidity performance and can be found on EPA's website here.

The CPE coordinator should also compile other relevant information that the PWS provides
such as a plant schematic, organizational chart, PWS financial data, cyanotoxin sampling data,
chemical dosing data, or recent sanitary surveys. The coordinator typically reviews this
information with the CPE evaluationteam duringthe pre-CPE meeting, described below.

2.1.2.5 Initiate Pre-CPE Studies

Studies anticipated to run longer than the duration of the CPE, or those that will provide data to
support the exit meeting, such as a PAC jar test study or oxidant kinetic study, should be
initiated off-site in advance to provide results in time for the exit meeting. Dependingon the
contact times and dosing concentrations that the team would like to evaluate, these types of
studies may need to be conducted for longer durations, orthrough several iterations. AWWA
has published a PAC jartesting protocol that is helpful forthis purpose. An EPA cyanotoxin
oxidation hold study protocol is included in Append ixB.

2.2 CPE Activities

Based on the pilot HAB CPE project, as well as experience with microbial CPEs, three to four
days will typically be needed to conduct a HAB CPE. The following discussion ad dresses the
activities that will generally take place over that period.

2.2.1 Pre-Event(Day 1)

The first day will often involve team travel to the CPE location and a pre-CPE team meeting.
Equipment can be dropped off at the plant, and on-site studiescan be set up and initiated as
needed. A study organizational template to assist in planning on-site studies is included in
Appendix B. On this first day we recommend that the CPE team briefly meet to review team
assignments, plant information, and water quality performance data; confirm that each team
understands their respective tasks; and discuss any outstanding logistics.

2.2.2 Entrance Meeting (Day 2)

On Day 2, an Entrance Meeting is conducted with PWS staffto introducethe CPE team, explain
the importance of HAB water treatment optimization and the purpose of the CPE (see example
bulleted ideas for including in a "Why Optimize?" presentation included in Appendix C), and
develop a schedule for the team's activities in cooperation with the PWS staff. Additional
examples of entrance meeting materials are provided in Appendix B.

-------
2.2.3 Water Treatment PlantTour (Day 2)

After the Entrance Meeting, the operators will typically lead the CPE team on a tour of the
water treatment plant. The objective of this tour is to familiarize the team with the plant
facilities, including physical layout, chemical dosing, and water-quality monitoring locations.
Plant managers and staff who are familiar with the plant's design, operation and maintenance
should lead the discussions. A list of suggested information to obtain during the plant tour is
included in the CPE data collection forms in Appendix B. After the plant tour is complete, the
CPE team, particularly the O&M team, should identify areas of interest for samp ling and on-site
studies.

2.2.4 Performance Assessment (Days 2-4)

During the performance assessment, three sub-teams will assess water quality performance
(current and historical), plantdesign, administration, and operationsand maintenance
practices. Refer to Figure 2 for sub-team roles. The information gathered prior to the CPE and
during the performance assessment will support the identification of performance-limiting
factors (PLFs) - factorsthat could compromise the PWS's ability to optimize. Performance
assessment activitiesgenerally begin afterthe planttourand are generally completed no later
than the morningof Day 4. A description ofthe performance assessment activities is provided
in thefollowingsections.

2.2.4.1 On-Site Studies

On-site studies a re conducted during the CPE to support the identification of PLFs. The data
integrity checks, filterassessment, source water sampling, plant profile, and jartestingare the
primary studies, but otherstudies may be conducted iftime permits.The on-site studies
generally begin on the afternoon of Day 2 and are completed no later than the morningof Day
4. However, we recommend that any studies that require cyanotoxin analysis be initiated early
(before the start ofthe CPE if need be) so that sample results can be included with the other
exit meeting materials being prepared for Day 4. Arrangements may need to be made with the
supporting lab to ensure that they can accommodate a quick turn-around time. On-site studies
are discussed in more detail below.

2.2.4.1.1 Jar Testing Study

Jar testing is a valuable tool for assessing and optimizing water treatment plant chemical
dosing, especially during a HAB event. When source water quality changes, it is important to
determine if corresponding treatment changes need to be made to remove cyanotoxins,
keeping in mind the importance of othertreatment objectives, such as turbidity and TOC
removal. Operators are encouraged to conduct jar testing in advance of, or during, the initial
stages of HAB occurrences. Utilizing concentrated raw water samples or spiking commercially-
available stock cyanotoxins can help a plant prepare for the changes in chemical dosing that
may be necessary to addressthe HAB.

Jar testing is designed to simulate the water treatment plant's processes and evaluate the
impact of treatment changes by adjusting mixing speeds or times.

-------
Jar tests during a HAB-focused CPE typically evaluate PACor pre-oxidant addition, or traditional
coagulation, flocculation, sedimentation processes where cyanobacteria cells are a component
ofthe raw water particulate loading. Note that jar test samples that contain PACshould be
filtered using0.6 |im glass fiberfilters (e.g., Whatman, grade GF/F) to remove the PAC prior to
cyanotoxin analysis. See A WWA's PAC Jar Testing Protocol.

2.2.4.1.2 Data Integrity Study

A series of studies may be conducted to assess the accuracy, precision and representativeness
(the "integrity") of the data by the plant. These studies focus on all aspects of measurement,
including the samp ling configuration and instrument operation and settings, since each can
impactdata quality.The team conductingthe study checks the following:

• Onlineturbidimeterflowratesusinga graduated cylinderor measuringcup,and
stopwatch. When the flow is above the maximum recommended rate, the potentialfor
turbulence and non-representative turbidity spikes increases.

• Sample detention time, which is the sample travel time between the filter effluent tap
and the turbidimeter, based on the length and diameter of sample tubing. We
recommend one minute or less so that water quality changes through filters are rapidly
observed.

• OnlinelFEand CFE turbidimeter settings are also checked, including signal averaging,
output span, bubble reject, and data logging settings. The turbidimeter out put span
should be at least 0 to 5.1 NTU to identify the magnitude of turbidity spikes and avoid
"capping" data.

• The online turbidimeter readings are also compared with grab samples analyzed on a
portable turbidimeter. Grab samples for this comparison study are ideally obtained from
a sample tap off the turbidimeterfeed line, however samples can also be obtained from
the drain lines ofthe continuousturbidimeters. Readings fromthe continuous
turbidimeterare generally considered to be the most accurate, due to the instrument
design and continuoussamplestream (i.e., no sample handling), presumingthatthe
instrumentsare well maintained and routinely calibrated. By collectinga representative
sample and using good testing techniques, high-quality readings may also be obtained
from the portableand benchtopturbidimeters.Those readingsareexpected to be
within 0.05 NTU of the continuous readings from IFE and CFE turbidimeters. Deviations
greaterthan 0.05 NTU should be investigated and maysuggest a continuous meterin
need of calibration or maintenance.

• Chemical feeder calibration. Determiningaccurate chemical feed rates is an important
part of plant process control and ensuring data integrity. Depending on the chemicals
dosed at the plant, the team may wish to check coagulant, PAC, softening chemicals, or
oxidantfeedsversus reported dosages.

-------
For more information on assuringthe integrity of turbidity data, referto EPA's "Generating
High-Quality Turbidity Data in Drinking Water Treatment Plants to Support System Optimization
and Monitoring" document.

2.2.4.1.3 Plgnt Profile Study

Process control sampling thro ugh the water treatment plant provides information on how each
unit process is performingand contributingto meeting water quality goals. Developinga plant
profile is a useful way of trendingthese process control sampling results. Duringa HAB, it is
importantforwater utilities to understand howeach water treatment unit process is
performing at removing cya no bacteria cells and cyanotoxins, while maintaining other
treatment objectives, such as turbidity and TOC removal and disinfection. Plant profile trending
can provide operators with warning of a cyanotoxins propagating thro ugh the treatment plant
and assist in identifying incremental process control changes that can be made to avoid passage
of cyanotoxins to the finished water.

Plant profilescan be developed fromsamplingresultsobtainedfromgrab samplesordata
sonde readings, or a combination of the two. Typically, data sondes can be fitted with sensors
to provide chlorophyll-o, phycocyanin, temperature, pH and turbidity measurements. The
pigmentschlorophyll-oand phycocyanin can be used as indicatorsoftotal algal biomass
(chlorophyll-o) and cyanobacteria biomass (phycocyanin, or"blue-green algae"). Grab samples
can be analyzed for cyanotoxins. Conducting this study duringa CPE, even if there is not a HAB
at the time, is a valuable way to demonstrate the study approach to operators.

2.2.4.1.4 Cygnotoxin Oxidgtion Kinetic Hold Study

A cyanotoxin oxidation hold study approach wasdeveloped to simulate waterquality dynamics
relative to cyanotoxin oxidation in the clearwell of a water treatment plant. Duringthisstudy,
water is collected from a location between the filters and clearwell (e.g., combined filter
effluent tap), dosed with known concentrations of a concentrated cyanotoxin solution and
chlorine (if not previously added in the treatment process) and held in a container to simulate
clearwell conditions. Waterquality samplesare periodically collected and used to estimatethe
oxidation rate of cyanotoxins in the water. Appendix B includes a protocol for conducting the
hold study. See EPA Method 127, Appendix A for a protocol for making and standardizing a
chlorine stock solution.

2.2.4.1.5 Source Wgter Sgmpling/Profiling Study

Many factors contribute to the concentration and vertical distribution of cyanobacteria in the
source water column, including depth, temperature, turbulence (e.g., wind-induced mixing or
currents), and cyanobacteria composition. For example, some bloom-formingcyanobacteria
genera, such as Microcystis, have gas vesicles to regulate their buoyancy and can form scums
on thesurface ofthe water. However, othergenera, such as fWamentous Plgnktothrix, are
typically distributed throughout the water column. Given the suite of influential variables,
assessing the water quality and susceptibility to cyanobacteria blooms and cyanotoxins is an
important step to inform avoidance strategies.

-------
The source water cyanobacteria and cyanotoxin assessment includes a reviewof historical raw
water data and on-site sampling at the intake structure or wet well. Parameters that can inform
this study include chlorophyll-a, phycocyanin, cyanobacteria cell identification and counts,
cyanotoxin concentrations, cyanotoxin-producinggene counts (i.e., through PCR), or other
indicator parameters that could be correlated tothe proliferation of cyanobacteria and
cyanotoxin production.

Watercolumn vertical profiling can also be conducted usinga data sonde equipped with
cyanobacteria-related sensors, such as chlorophyll-o and phycocyanin, as well as temperature,
DO, and pH. Vertical profiling conducted at or near the raw water intake structure can inform
decision-making related to water quality at the intake depth.

2.2.4.1.6 Filter Assessment Study

This study involves an evaluation ofthefilter media by probingand excavatinga small amount
of media for assessment (see below), as well as an evaluationofa filter backwash, including
bed expansion measurement, developinga backwash waste turbidity profile, filter backwash
recovery profile, and filter-to-waste data review. This assessment istypically conducted on the
plant's worst-performing filter as judged by effluent turbidity, filter run times, head loss, etc.

Media assessment and safety considerations

Filter entry involves some safety precautions that require training, setting up and securing
ladders, monitoring air quality, and lock-out/tag-out (LOTO) of the filter valves after the filter is
drained and while personnel are on the filter media surface (engulfment hazard). Many filters
require entry by ladder and have limited means of egress, which can classify them as confined
spaces. This may warrant specialized training for personnel performing the inspection, including
confined space and fall protection training.

The filter media assessment begins by d raining the filter. LOTO devices are then attached to the
appropriate valves or instrumentation to eliminate the engulfment hazard. Air monitoring is
conducted throughout the space (for example, with an air meter with the sensor hanging down
into the filterspace) to determine if an atmospheric hazard is present. After LOTO precautions
are made and it has been confirmed that no atmospheric hazard is present, the "Permit-
Required Confined Space" may be temporarily reclassified through a certification as a "Non-
Permit-Required Confined Space". At this point, the filter is deemed safe to enter and
personnel can descend onto the filter media. Due to the risk of a fall, a harness with a
retractable fall arrester, or some othertype of fall protection equipment, should be used for
filter entry. Throughout the evaluation, continuous atmospheric monitoring is conducted. The
filter media assessment involves one ortwo personnel descendingonto the filter media surface
and probing the media to determine the overall depth of media in the filter. This is
accomplished by probingthe filter at approximately equally-spaced distances in a grid-like
pattern across the surface of the filter. Media depth measurementsare plotted and used to
determine if areas of the filter bed are uneven and where media loss may have occurred. Test
pits are also excavated, either by hand or usinga shovel or trowel, to examine the extracted
media for mud balls and to determine if media is still stratified as-designed. Once the filter
media inspection is completed and personnel are off the filter, atmospheric monitoring is

-------
discontinued. LOTO devices are then removed from the plant equipment and the remaining
portion ofthe filterassessment can be conducted (i.e., the filter backwash assessment).

Backwash assessment

After the filter media assessment, the CPE team typically requests that the plant operators
conduct a filter backwash on the same filter in which the media was assessed. During the
backwash, bed expansion is measured and grab samplesof the backwash wastewaterare
analyzed for turbidity to create a backwash waste turbidity p rofile. This can provide information
about the efficacy ofthe backwash. When the filter is placed back online (either filtering to
waste or regular filtration), a backwash recovery turbidity profile is developed by observing the
turbidity readings on the SCADA screen. The IFE turbidimetersample tap should be located
upstream of the filter-to-waste valve to support a proper assessment.

Filterbed expansion is measured duringa typical backwash.The percent bed expansion can be
calculated from the measurement of media expansion. This measure helps operators
understand the effectiveness ofthe backwash in cleaning the media and the ability for media to
re-stratify following a backwash. A minimum of 20 percent filter bed expansion is desirable;
however, filtersthat use airscour can achieve satisfactory backwashingat lower bed expansion
levels(e.g., 15 percent). Bed expansion is typically measured duringa backwash usinga Secchi
disk attached to a pole, although other types of bed expansion measurement tools have been
developed. The team marks the pole when the disk is sitting on top ofthe media prior to the
filter backwash, and again at the high backwash flow rate when the Secchi disk is observed to
disappear below the fluidized media. The measurement between these two markings
represents the depth of media expansion.

Concurrent with the bed expansion measurement, the team also obtains grab samples ofthe
backwash wastewater and analyzes them for turbidity. The purpose of this samp ling is to
determine the amount of time necessary for effective media cleaning. The equipment used to
perform this partof the study includesa collection device forgrab samplesand a portable
turbidimeter. The CPE team collects turbidity grab samples at varied times from the discharge
ofthe backwash waste trough, usinga long pole with a sample cup attached atthe end. Grab
samples are analyzed on a portable turbidimeter and then plotted in a spreadsheet. To support
filter backwash optimization, a waste backwash water turbidity target can be established to
determine when to end the backwash (e.g., 5 to 20 NTU). This target, along with other
backwash-related parameters such as post-backwash recovery turbidity, can be used to
optimize filter backwash.

Followingthe inspection and backwashing ofthe filter, the filter ripening period is then
monitored, typically by observing the SCADA HMI screen. The IFE turbidimetersample tap
should be located upstream ofthe filter-to-waste valve to support proper monitoring. The filter
ripening period is the time from when filter is placed back in operation, inclusive of the filter-to-
waste period, until the time the turbidity meets the optimization goal. For plants with filter-to-
waste capability, the optimization goal is to return the filter to service at < 0.10 NTU. Achieving
0.10 NTU turbidity before placinga filter back in service after a backwash cycle can reduce the
number of particles, including pathogens and cyanobacteria cells, that pass to the clearwell.

-------
The study team can also retrieve historical backwash data to obtain a better understandingof
typical operational practices and water quality performance during backwash ing and returning
filters to service. This type of data analysis can be used to identify areas where studies can be
conducted to improve plant performance and potentially save operatortime and reduce water
usage associated with backwashing.

2.2.4.2 Information Collection

Information iscompiled to assess historical watertreatment performance, assesstreatment
plant operations based on operational data, and evaluate administration, operations and
maintenance practices through discussions with PWS personnel.This information mayalso
supportthe on-site studies.

2.2.4.2.1 Water Quality Performance Assessment

The PWS's historical waterquality data should beassessed relative to the microbial (turbidity)
optimization goals listed in Table 1 and the EPA Health Advisory values for microcystins and
cylindrospermopsin (U.S. EPA, 2015a, 2015b). Ideally, these data will have been compiled using
the Optimization Assessment Spreadsheet (OAS) priorto the CPE (see Section 2.1.2.4). If these
data were not already collected, or if data were missing, this should be completed atthistime.

* Table 1: AWOP's Microbial Optimization Goalsfor Rapid Rate Filtration Plants
Raw Water

Minimum Data Monitoring Goal

> Record maximum daily raw water turbidity.

Individual Sedimentation Basin

Performance Goals

^ Settled water turbidity < 2.0 NTU in 95% of readings when the annual average raw water turbidity is > 10

NTU. Optimization is based on the daily maximum values recorded from all readings.

^ Settled water turbidity is < 1.0 NTU in 95% of readings when the annual average raw water turbidity is <
10 NTU. Optimization is based on the daily maximum values recorded from all readings.

Monitoring Goals

^ Record individual sedimentation basin effluent turbidity readings at intervals of 4-hours or less if taking
grab samples, or 15 minutes or less for continuous monitoring.

Individual and Combined Filters

Performance Goals

> Combined filter effluent turbidity < 0.10 NTU in 95% of readings. Optimization is based on the daily
maximum values recorded from all readings.

> Individual filter effluent turbidity < 0.10 NTU in 95% of readings (excluding 15-minute period following
filter backwash). Optimization is based on the daily maximum values recorded from all readings.

> Post-backwash individual filter effluent turbidity for filters without filter-to-waste capability: Maximum

-------
individual filter effluent turbidity following backwash < 0.30 NTU and achieve < 0.10 NTU within 15
minutes.

> Post-backwash individual filter effluent turbidity for filters with filter-to-waste capability: Minimize
individual filter effluent turbidity during filter-to-waste period and record maximum value. Return the
filter to service at < 0.10 NTU.

Monitoring Goals

> Record individual and combined filter effluent turbidity readings at intervals of 1-minute or less for
continuous monitoring.

Disinfection

Performance Goals

> Meet CT requirements to achieve inactivation of Giardia and viruses plus a system-specific factor of
safety.

Monitoring Goals

> Record disinfectant residual, temperature, and pH at maximum daily flow for CT calculations.

2.2.4.2.2 Administration, Operations, and Maintenance Assessment

The purpose of the administration, operations and maintenance assessment isto collect
information to help identify PLFs related to the management, physical integrity, operation and
maintenance of the water treatment plant. A data collection form is used to summarize
discussions with plant staff and administrators. A copy of the Administration, Operations and
Maintenance Assessment form is provided in Append ixB. The Administration and O&M Teams
should conduct interviews with plant staff and key administrators, potentially including board
members and the utility director, as appropriate.

Assessment activities and interviewtopics include:

• Administration/Financial (Administrative/Financial Sub-Team): Review organizational
structure, staff roles and responsibilities, communication, management policies
including water quality goals, financial support, long-term planning, and financial
information. Collect information on administrative policiesand procedures through
interviews with PWS administrators and review pertinent financial records.

• Operations (O&M Sub-Team): Review water qualitygoals, policies, and practices, and
decision-making related to areas such as unit process control, chemical dosing, and
instrument calibration and maintenance. Review data management, problem solving
skills, and laboratory capability. A critical element of this assessment is to understand
the plant staff's approach to a voiding complacency and ensuring plant reliability and
water quality.

• Maintenance (O&MSub-Team): Review preventative and corrective maintenance
practices, aswell as resources availableformaintenanceactivitiessuch asequipment
repairand parts, maintenance expertise, availability oftools, and maintenancetracking.

-------
2.2.5 Identification of Performance Limiting Factors (Day 4)

Once each sub-team completes their performance assessment activities, they should conduct a
"pre-factors" meeting. The purpose of this meeting is for each sub-team to individually review
the list of potential PLFs (see Appendix B) and identify any that may be relevant based on the
team's assessment. Once all sub-teams are ready, the collective CPE team meets to identify,
rank, and prioritize the PLFs for the PWS. The Administrative Sub-Team typically leads this
meeting, utilizingthe information collected duringtheirassessment, interviews, and notes to
guide the team through the factor identification process.

The summary of the prioritized PLFs provides a guide for the PWS's future optimization efforts.
It is helpful to organize the factors based on relative priority so that the PWS can appropriately
target their initial efforts to address them. This is particularly true when many PLFs have been
identified and there is potential forthe PWS to otherwise be overwhelmed.

2.2.6 Exit Meeting (Day 5)

The exit meeting, conducted atthe end of the CPE, is an opportunity to presenttheteam's
findings to the PWS personnel and to establish priorities for pursuing the optimization
performance goals. The evaluation team may have additional points to discuss after the
weeklong evaluation, but the exit meeting is the team's opportunity to provide an initial
summary of the PWS's performance. A presentation of results generally includes the following:

• Recap of microbial/turbidityand HABtreatmentoptimization

• Performance goals

• Performance assessment findings

o Historical performance data assessmentfindings
o On-site studiessummary

• Summary of PLFs

• Suggestions for further study as an initial approach for pursuing optimization

Example exit meeting files are included in Appendix C. The exit meeting should focus on
presenting the data that supports the PLFs that were identified during the CPE. If CPE activities
do not result in significant findings, the team typically does not present about those studies
duringthe exit meeting.

Typically, all information that will be included in the final report is summarized atthe exit
meeting. However, if results from any on-site samplingare not available until afterthe CPE, the
PWS staff should be informed that this supplemental information will be made available in the
final report.

-------
2.3 Post-CPE Activities

The two main activities afterthe CPE include preparation of the final reportand support forany
follow-up optimization activities that the PWS wishes to pursue.

2.3.1	Final Report

The purpose of the final report is to document the findings of the CPE for the PWS and to help
establish priorities for pursuing optimized plant performance. The final report should generally
reflect the exit meeting summary.

The members of the evaluation team are responsible for developing the sections of the report
that correspond with their focus areas during the CPE. The evaluation team should strive to
complete the report as quickly as possible following the CPE.

2.3.2	PWS Follow-up

Follow-up with the PWS after the CPE is very system specific but is typically initiated by delivery
of the final report to PWS staff members, and further review of the report with the PWS as
needed. Some PWSs may wantto initiate furthersamplingorstudies, while others may not
currently have the resources. Interested PWSs should be encouraged, and supported if
possible, bythe state to beginthe steps of implementing HABtreatmentoptimization. Ideally
the PWS personnel will ad dress the PLFs, but often this will require support from the state
personnel. Comprehensive technical assistance or performance-based training a re two
approaches utilized by state personnel that have proven to be successful in achievingand
sustainingoptimized performance at water treatment plants.

19

-------
3. IMPLEMENTING HAB TREATMENT OPTIMIZATION

This section summarizes information on how a PWS could begin pursuing water treatment
optimization, or even simply improving their ability to avoid/manage HAB impacts, whether it
be in conjunction with a CPE or on theirown initiative (i.e., independentof a CPE). Similar
information is also provided in the exit meeting handout, "Possible Further Studies for Plant
Staff to Conduct to Support Plant Optimization" Template that is included in AppendixC. Often,
there are both technical and non-technicalchallengesthat PWSsface in implementing
treatment optimization.

3.1 Effective Leadership and Management

Supportand leadershipfrom plant management isan important first step in achieving water
treatment optimization and protecting public health. This might include:

• Establishing optimization goals that have the buy-in of utility staff and management.

• Creating accountability by defining expectations through clear roles and responsibilities,
documentation of meeting outcomes, and assignment of tasks.

• Making decisions based on data to gain support from utility staff and management for
making treatment process changes. Applying problem-solving skills, such as conducting
studies and trending and interpreting data.

• Developing operational policies and procedures to enhance communication among
utility staff and managementon critical activities.This could include establishing
sampling schedules for cyanotoxins and developing monitoring protocols.

• Establishing routine communicationthrough regular meetings, data distribution, or
memorandums to continuously assess PWS performance and provide a feedback loop.

3.2 Adopt Water Quality Goals

Adoptingand communicatingwaterqualitygoalsthroughoutthe utility isa critical step for
optimizing treatment and helps to ensure that all staff is committed to maintaining water
qualitythroughoutthe PWS.

The process of pursuing optimization to increase public health protection will likely require
changes in operations and daily activities, which is why it is critical that all parties a re
committed. Duringthe CPE, optimizationgoalsare referenced duringthe historical data
performance assessment and summarized duringthe exit meeting.

-------
3.3 Establish a Consistent Sampling Approach

It is importantto evaluate treatment efficacy throughout a PWS through regular monitoring
and data trending. Establishing, documenting, and communicating a regular samp ling approach
for collecting water quality samples in the treatment plant ensures that everyone samples
consistently. Monitoring at multiple locations in the process train can help water treatment
plant operators evaluate the effectiveness of each unit process. The selection of process control
sample locations will dependont he plant configuration and chemical feed locations.

3.4 Monitoring for HAB Treatment Optimization

The foundation of a water treatment optimization program is monitoring data. The historical
data review and OAS analysis conducted as part of the CPE could be the start of a longer-term
process to trend and analyze SCADAand grab samp ling data. A sufficient quantity and
frequency of representative monitoring from each unit process in the treatment plant is
recommended to establish a water quality data base from which process control decisions can
be made.

There are three basic steps to establishing a monitoring plan:

1. Identify grab sample and continuous monitoring locations. These locations might
include source water, raw water, recycled waterfeed, after chemical addition in the
rapid mix (i.e., afterchemicals are completely mixed), settled water, individual and
combined filtereffluent, and finished water.

2. Collect and analyze samples and record data. Individual and combined filtereffluent
turbidity data can be compiled from SCADA and trended in a spreadsheet (such as the
Optimization Assessment Spread sheet). This turbidity dataset should be updated
regularly. Grab samples or sonde data that indicate cyanobacteria orcyanotoxinssuch
as chlorophyll-oand phycocyaninare helpful for establishing baseline raw water quality
and for assessing cyanotoxin treatment performance if regular cyanotoxin sampling is
prohibitive. Grab samplesforcyanotoxin analysis can be collected based on the results
of indicator parameters, as appropriate. This HAB data should be compiled into a
spreadsheet. This database can also incorporate data obtained from the sampling
conducted during the CPE. These sampling results, along with established compliance
monitoring locations, can be the basis of a long-term water quality monitoring plan.

3. Trend and analyze data. Trend the data that has been compiled on a regular basis,
observingany relevanttrendsthat may indicate the need forchanges in process control.
The approach to makingtreatment adjustmentsforcyanotoxins dependson the
monitoring results and type of cyanotoxins present. These tools help operators use data
to make better process control decisions and can provide an early warning system for
water quality problems throughoutthe treatment plant.

-------
REFERENCES

Hegg, B. A., DeMers, L. D., Bender, J. H., Bissonette, E. M., & Lieberman, R. J. (2004). Optimizing
Water Treatment Plant Performance Using the Composite Correction Program. Cincinnati:
United States Environmental Protection Agency.

U.S. EPA. 2015a. Drinking Water Health Advisory for the CyanobacterialToxin
Cylindrospermopsin. EPA 820R15101. Available online at

https://www.epa.gov/sites/production/files/2017-06/documents/cylindrospermopsin-report-
2015.pdf

U.S. EPA. 2015b. Drinking Water Health Advisory forthe Cyanobacterial Microcystin Toxins. EPA
820R15100. Available online at https://www.epa.gov/sites/production/files/2017-
06/documents/micro cyst ins-report-2015.pdf

U.S. EPA. 2015c. Health Effects Support Document for the Cyanobacterial Toxin Anatoxin-a. EPA
820R15104. Available online at https://www.epa.gov/sites/production/files/2017-
06/documents/a natoxin-a-report-2015.pdf

U.S. EPA. 2015d. Health Effects Support Documentforthe CyanobacterialToxin
Cylindrospermopsin. EPA 820R15103. Available online at

https://www.epa.gov/sites/production/files/2017-06/documents/cylindrospermopsin-support-
report-2015.pdf

U.S. EPA. 2015e. Health Effects Support DocumentfortheCyanobacterialToxin Microcystins.
EPA 82015102. Available online at https://www.epa.gov/sites/production/files/2017-
06/documents/micro cyst ins-support-re port-2015.pdf

22

-------
Appendix A:

Pre-HAB CPE Activities and Preparation

Contents

Example Drinking Water Treatment Plant Selection Considerations	2

Example HABCPE Overview Letter and Information Request Template	3

Example HABCPE Itinerary	5

Example Water System Data Request	7

Example Equipment and File List	8

Office of Water (MS-140)

EPA 815-B-22-005

June 2022

-------
Example Drinking Water Treatment Plant Selection Considerations

1. Plant Selection Considerations

• WTP staffing-The plant should have sufficient staffingto host a HAB-related CPE
development team of ~15 people. At a minimum, a plant superintendent and two
operators are typically needed to support a CPE team of this size.

• WTP construction-Theteam should avoid selecting plants that would be undergoing
majorconstruction projects during the targeted dates for a CPE.

• WTP performance - The CPE tool is most appropriate for surface water treatment plants
that are currently not achieving the optimization turbidity goals. Plants with individual-
filteror combined-filterturbidity performance of > 0.15 NTU would be good CPE
candidates.

• Turbidity data availability — Ideally, the host plant will have readily accessible electronic
turbidity records for raw, settled, individual filter, and combined filter turbidity available
for the past year. Access to turbidity data on plant SCADA is sometimes challenging,
depending on the age and capability of the system. The CPE team can identify the data
of interest to the WTP staff in advance to see if the team can readily access it, and if not,
to see if the plant can provide access (likely by modifyingthe SCADA programming)
before the team arrives on site.

• Ability to feed PAC - the team will learn more about optimizing this process at a water
system that is alreadyfeeding PAC, or has the ability to feed PAC and has done so in the
past.

2. Other Considerations

• Willingness to host - Some utilities that would otherwise be good CPE candidates may
be hesitant to host an evaluation of their plant for different reasons. The team should
consider whetherthe concerns of the utility can be addressed priorto schedulingthe
CPE.

• Other performance issues-The primary focus of a traditional, microbial-focused CPE
(which serves as the foundation fora HAB CPE) is turbidity removal and disinfection
practices in the plant. If the utility performance issues are not related to these
parameters (e.g., utility whose primary issue is elevated DBPs), they may not be the best
candidate.

• Accommodations-The proximity of lodging for the CPE team should be considered
during plant selection. Theteamwill be on-site atthe plantfor3-l/2 daysand ideally
lodging will be close to the plant to minimize travel time.

-------
Example State-to-PWS HABCPE Overview Letter and Information Request

Date

Jane A. Doe
City ofXXXX
123 Main Street
City, State 12345

RE: Microbial / Harmful Algal Bloom (HAB) Comprehensive Performance Evaluation (CPE) at the City
ofXXXX Water Treatment Plant on «Date»

Dear Ms. Doe:

John Smith of the State Environmental Protection Agency recently contacted you regarding an
upcoming evaluation of your water treatment plant (WTP) and visited to collect some preliminary
information. To further prepare you, this letter provides some additional information about the
evaluation and describes the activities that will be conducted.

This CPE process was developed through EPA's national drinking water optimization program,
which is coordinated out of the Technical Support Center (TSC) in Cincinnati, Ohio with contractor support
from Process Applications, Inc. (PAI) of Fort Collins, Colorado. EPA's program develops compliance
assistance tools and approaches that can be used by State drinking water programs and water systems to
improve drinking water quality - either to compliance levels, or beyond - to enhance public health
protection.

The philosophy of the program is to optimize existing facilities and staff to achieve the desired water
quality performance goals. The program was originally developed to optimize surface water treatment
plant performance for protection against microbial contaminants such as Giardia and Cryptosporidium.
Recently, that approach was adapted for plants challenged by harmful algal blooms (HABs) and
cyanotoxins. During the CPE, all aspects of your water system's administration, design,
operation/treatment, and maintenance will be evaluated to assess their impact on achieving optimized
performance.

Attachment 1 provides the tentative schedule of activities during the CPE and some additional
details are provided below. If this schedule needs to change as the week progresses, we will work with all
involved to avoid interfering with the water plant staff's responsibilities.

The CPE will begin with a brief entrance meeting on «Date» at 8:00 a.m. with plant staff and
administrators. During this meeting, a brief overview of the optimization approach will be provided and
the planned activities and schedule over the next three days will be discussed. Any questions or concerns
regarding the evaluation can also be raised atthis time. We recommend that the plant administrators and
those responsible for plant budgeting and planning be present because this evaluation will include an
assessment of these aspects of the water system.

At various times during the week, the CPE team will need the assistance from one or more water
system staff. For example,

-------
• On Day 1, following the entrance meeting, the team will need someone that is knowledgeable
about water treatment plant design, operation, maintenance, and water quality to lead the water
treatment plant tour.

• Starting on the afternoon of Day 1 through the morning of Day 3, the team will review questions
with plant staff about performance, operations and maintenance practices. The team may also
conduct particular studies to investigate, or simulate, the performance capabilities of the various
unit treatment processes, including filtration. Requests to inspect filter media and monitor filter
backwashes will be coordinated with plant staff to minimize the impact on plant operation.

• Members of the evaluation team may meet with system administrators on Day 1 or Day 2 to
review the administrative policies, procedures, and financial records. Additionally, plant staff and
administrator interviews will be conducted the morning of either Day 2 or Day 3. This can be
scheduled around their availability.

The evaluation will close with an exit meeting and discussion of the CPE results on Day 4 at 8:00
a.m. In addition to the plant staff and administrators that participated in the CPE, any City of XXXX
managers (decision makers) are invited to attend. An assessment of the performance capabilities of the
treatment processes will be presented and any factors appearing to limit the performance of the plant
will be discussed. The evaluation team will also answer questions regarding the results of the evaluation.
The results presented during the exit meeting will form the basis of the final report, which will be likely
completed within two to three months after the event.

Attachment 2 contains a list of information that the team will need during our site visit. Having this
information available when the team arrives will allow us to make the best use of time during the week.

We look forward to conducting this evaluation at your facility and thank you for your collaboration.
I will soon follow-up with a phone call to discuss this upcoming event, and to review the availability of
information requested in Attachment 2. In the meantime, if want to contact me, please do so at (123)-
456-7890 or «email@email.com».

Sincerely,

Name

Organization

-------
Example HAB CPE Itinerary

PWS Name, City, State
Date

Day 1, «DATE»

8:00 A.M.: Entrance Meeting at the WTP building or other designated conference room facility (all
facility personnel involved with supporting the CPE are encouraged to attend)

•	Introductions

•	Background and purpose of the field event

•	Review activities planned for the week

•	Discuss involvement of plant staff and management (sampling, data collection, and interviews)

•	Answer any questions and discuss any concerns

8:45 A.M. - Noon: Plant discussion and tour for CPE team by the plant manager/staff:

•	Explanation of overall treatment process - including design and basic information on the water
treatment plant.

•	Tour of the water treatment plant - note the type and locations of water quality monitors (e.g.,
raw water monitoring, turbidimeters, streaming current monitor, etc.) and chemical addition
points (e.g., preoxidant, coagulant, disinfection, etc.).

•	Historical water quality and plant performance - discuss any potential concerns.

Afternoon: CPE team activities, with assistance from WTP staff as needed and (limited) management
support; anticipate two to three staff needed to assist with:

•	Additional data collection/review activities:

¦	Historical data: turbidity, cyanotoxins or HAB related monitoring data (e.g., chlorophyll-a,
phycocyanin)

¦	Grab samples from the treatment process train, as necessary

¦	Recent sanitary survey report

¦	HAB treatment optimization plan

•	Historical data review to support identification of potential studies (e.g., treatment train sampling
study, jar testing) with operator input. Begin studies if time allows.

•	Review of plant design and assessment of plant processes to judge turbidity removal; consider
impact on HABs/HAB removal

• Begin review of administrative policies and financial records for the plant.

Wrap-up at the Water Plant by 3:30 p.m. (or end of shift)

Day 2, «DATE» (anticipate 8 a.m. to 3:30 p.m.)

5

-------
Teams will continue with activities started on Day 1. Some from the CPE team will sample and conduct
studies for some/all of the day, likely with assistance of plant operators. Other team members will focus
on data gathering and interviews with water plant management and staff. These interviews will be
scheduled with individuals as appropriate.

Day 3, «DATE»

Morning:

Collect additional samples and conduct studies, as needed (plant staff may be needed to assist)
Conduct additional interviews with plant representatives, if needed

Afternoon:

Compile the data and prepare for the Day 4 exit meeting (no assistance needed from water system)
Day 4, «DATE»

8:00 A.M.: Exit Meeting at the WTP building or other designated conference room facility (all facility
personnel involved with supporting the CPE are encouraged to attend)

•	Preliminary presentation of findings from the CPE

•	Discuss the team's observations and answer questions

•	Discuss next steps and final report schedule

9:30 A.M.: CPE Team will depart

6

-------
Example Water System Data Request
PWS Name, City, State
Date

•	Historical water quality data: -Any data (mentioned below) that are available in electronic format. If
only paper records exist, the most important (and available) data can be identified. The team would
like to review one year of historical data if available.

¦	Raw, Settled, Top of Filter, Individual Filter Effluent, and Combined Filter Effluent (and/or
Finished) water turbidity values.

¦	Any historical cyanotoxin/HAB-related monitoring data (e.g., chlorophyll-a, phycocyanin)

¦	Chlorine dose and residual, pH, and temperature data (or equivalent information used to
calculate CT)

•	Administrative data

¦	Budget information, including revenues and expenses, rate structure and debt service

¦	Capital Improvement Plan(s)

¦	Organizational Chart/Staff positions/certifications

•	Design data: as-built drawings ofthe plant

•	Recent sanitary survey report, which may include:

¦	Type and locations of water quality monitors (e.g., turbidimeters, streaming current
monitor, raw water monitoring, etc.)

¦	Chemical dosing information (locations and chemicals used)

•	HAB treatment plan (if developed)

7

-------
Example Equipment and Document List (for use by CPE Team)

Equipment:

•	Colorimeters or SL1000 (2 or 3?)

o Associated reagents for:

¦	Free ammonia

¦	Total ammonia

¦	Free chlorine

¦	Total chlorine

¦	Color-Hach Method 8025 (TSC doesn't have this...check with Nick...order some
for next CPE)

¦	Nitrate (8039)

¦	Nitrite (8507)

¦	Alkalinity

•	ADDA-ELISA sampling kit?

o Amber glassware
o Coolers with ice?

o Ensure that lab is informed of anticipated sampling load and to bring
sampling/preservation stuff

•	pH meter (there is also a pH meter on the sonde)

•	YSI EXO sonde w/ probes: phycocyanin, chlorophyll, turbidity, DO (?), conductivity (?), pH,
temperature

o Associated calibration standards (rhodamine solution for pigments), etc.

•	AquaFluor fluorometer for chlorophyll-o

•	Thermometers?

•	Turbidimeter? Hach 2100Q portable turbidimeter

•	Spectrophotometer to run UV254? Need sampling cells and filtration apparatus, miliQ water

•	Jar testing equipment

o	Phipps & Bird apparatus

o	Cylindrical glass jars + square jars

o	Syringes and filters for PAC jar testing

o	Sample cups

o	Filtration method to remove PAC (assume glass is needed)

o	Determine PAC, coagulant/polymer types, WTP chemical dosing at sampling point in
order to accurately replicate in jar test

8

-------
¦ Microsyringes, slide covers, plastic dosing syringes and cups for dosing

•	MC-LR toxins to spike in jars? Use concentrated raw water sample?

•	Carboy to bring raw water back

•	Sample pump in case needed to collect raw water

•	Climbing gear: harness, strap and retractable cord, helmets, other, for filter inspection if
needed.

•	Sample dipper (2 - one long, one short). Make sure long enough to reach backwash troughs!

•	Bed expansion test measurement tool again?

•	Filter probe with measurement increments?

•	Kimmerer sampler (depths) or well pump?

•	Folding table

•	Tape measures (recommended 200 ft)

•	Flashlight

•	Tool bag

•	Plastic sample cups

•	Dl water - verify that they have this at the plant, otherwise bring our own

•	Kim wipes

•	2 coolers

•	Bag of post-it notes, pencils, pens, markers, tape

•	Safety equipment (nitrile gloves, safety glasses, work gloves,

•	Projector for entrance and exit meeting presentations

•	Clip boards

•	M57 and Walker texts, other supporting papers (PAC, oxidation, etc.)

•	Portable table

9

-------
Electronic Files:

•	HAB CPE forms

•	"Why Optimize" presentation slides

•	Blank data log sheets

•	Plant historical data

•	AWWA's CyanoTOX, PAC Calculator spreadsheets and protocols

•	OAS

•	Latest Optimization document

•	Supporting papers (PAC, oxidation, etc.)

10

-------
Appendix B:

On-Site Materials

Contents

Data Collection Forms and HAB CPE Agenda 	2

Study Format. Elements, and Template	81

Cyanotoxin Oxidation Hold Study Protocol	83

HAB CPE Performance-Limiting Factors (PLFs)	88

Office of Water (MS-140)

EPA 815-B-22-005

June 2022

-------
Data Collection Forms and HAB CPE Agenda

A. KICK-OFF MEETING AGENDA

1.	Purpose of the CPE

•	Background on CCP process development and application

•	Basis for conducting the CPE at the utility

•	Assess ability of plant to meet optimized performance goals
Optimized performance criteria description

Multiple barrier concept for microbial protection

•	Identify factors limiting plant performance

•	Describefollow-up activities

2.	Schedule CPE events	Utility Staff Involved	Date/Time

•	Plant tour		 	

•	On-site data collection

Performance				

Design				

Operations				

Maintenance				

Administration				

•	Studies

•	Interviews

2

-------
Exit meeting

-------
Information Resources

•	Performance monitoring records

•	Plant operating records

•	As-built construction drawings

•	Plant flow schematic

•	As-built construction drawings

•	O & M manuals

•	Equipment manuals

•	Previous and current year budgets

•	Organizational structure

•	Water rate structure

-------
B. ATTENDANCE LIST

Utility Name		Date

Name

Title/Position

Telephone No.













































































































5

-------
A. NAME AND LOCATION

1.	Name of Facility		

2.	Utility Name		

3.	Current Date		

4.	Contact Information:



Administration

Plant



Contact Name







Title







Mailing Address



















Phone







Fax















B. ORGANIZATION

1. Governing Body (name and scheduled meetings)

2. Utility structure (attach organizational chart if available)

6

-------
B. ORGANIZATION (CONT.)

3. Plant Organizational Structure(includeoperations, maintenance, laboratory personnel; attach chart if
available)

C. WATER QUALITY

1. Utility Vison / Mission Statement

7

-------
2. Water Quality Goals (turbidity, disinfections, DBPs, cyanotoxins)

3. Water Quality Reporting (type of reports, data reviewed by administrators)

C. WATER QUALITY (CONT.)

4. Management Style and Impact on Plant Operations and Performance (i.e., decision making process,
chain-of-command, level of involvement)

8

-------
5. Complacency and Reliability (approaches / activities used to prepare for unexpected events)

Topic

Description/Information

1. Complacency



• How does utility respond to



unexpected or infrequent water



quality events (e.g., harmful algal



bloom, seasonal changes)?



• Does utility have an emergency
response plan? How does staff
trainfor unusual conditions or
events?











2. Reliability

• Does staff capability to make
process control decisions







9

-------
Topic

Description/Information

exist at more than one level?

• Have process or equipment
limitations/deficiencies been
identified and corrections plans
been developed?







D. COMMUNICATIONS

Type

Description

~ Staff Meetings









~ Administrator/Board
Visits to Plant









~ Reports (plant staffto
manager; manager to
governing board)







10

-------




~ Public Relations/
Education











E. PLANNING

1. Short-Term Needs

2. Long-Term Needs

11

-------
F. PERSONNEL

Title/Name

No.

Certification

Pay Scale

% Time
at Plant





































































































Comments (e.g., vacant positions, adequacy of current sta

Ffing):







G. PLANT COVERAGE

1. Shift Description (e.g., length, number per shift, weekend/holiday coverage)

12

-------
2. Unstaffed Operation Safeguards (e.g., alarm/shutdown capability, dialer)

13

-------
H.	FINANCIAL INFORMATION

I.	Budget (basis for budget: total utility ~ plant only ~)



Last Year Actual

Current Year Budget

Enter Year





1. Beginning Cash on Hand





2. Cash Receipts





a. Water Sales Revenue





b. Other Revenue (connection fees, interest)





c. Total Water Revenue (2a +2b)





d. Number of Customer Accounts





e. Average Charge per Account (2a h- 2d)





3. Total Cash Available (1 + 2c)





4. Operating Expenses





a. Total O&M Expenses *





b. Replacement Expenses





c. Total 0,M&R Expenses (4a + 4b)





d. Total Loan Payments (interest + principal)





e. Capital Purchases





f. Total Cash Paid Out (4c + 4d + 4e)





g. Ending Cash Position (3 - 4f)





5. Operating Ratio (2a h- 4c) ±





14

-------
6. Coverage Ratio (2c - 4c) h- (4d) t





7. Year End Reserves (debt, capital improvements)





8. End of Year Operating Cash (4g - 7)





Source: USEPA Region 8 Financial Analysis Document (1997)

* Includes employee compensation, chemicals, utilities, supplies, training, transportation, insurance, etc.

± Measure of whether operating revenues are sufficient to cover 0,M&R expenses. An operating ratio of
1.0 is considered minimum for a self-supporting utility.

t Measure of the sufficiency of net operating profit to cover debt service requirements of the utility.
Bonding requirements may require a minimum ratio (e.g., 1.25).

15

-------
2. Supporting Financial Information

Category

Information

~ Rate Structure

•	User fees

•	Connection fees

•	Planned rate changes













~ Debt Service

•	Long-term debt

•	Reserve account











~ Capital

Improvements

•	Planning

•	Reserve account











~ Budget Process

• Staff involvement











16

-------
~ Spending Authorization

•	Administrator

•	Plant staff





















17

-------
A.

PLANT SCHEMATIC AND CAPACITY INFORMATION

1. Attach or draw plant flow schematic; include the following details:

•	Source water type/location	• Chemical injection locations

•	Major unit processes	• Piping flexibility

•	Flow measurement locations	• On-line monitoring type/location

18

-------
2. Flow Conditions:

Parameter

Flow



Design Capacity





Average Annual Flow





Peak Instantaneous Flow





19

-------
B. MAJOR UNIT PROCESS INFORMATION

1. Intake:

Topic

Description

Information

1. Description

Locations



List intake depths



Description of source mixing or
aeration facilities



2. HAB control /
impacts

Ability to add preoxidant or
other chemicals



Ability to adjust intake depth





Ability to change source water
or use other intakes



Does design promote algae
growth (Uncovered/long
detention time)



3. Other design
Information or
limitations observed.















20

-------
B. MAJOR UNIT PROCESS INFORMATION (CONT.)

2. Rapid Mix:

Topic

Description

Information

1. Description

Type (reel, hydraulic,turbine)



Describe Flow Splitting



Control (variable/constant
speed)



2. Unit Process
Evaluation

Mixing Energy (G)



3. Other design
information or
limitations observed















Calculation of mixing energy as expressed by the mean velocity gradient (G) for mechanical mixing:

r \1/2

( P ^

J

G = Velocity gradient, sec 1
|j = viscosity, Ib-sec/ft2
v = volume, ft3

Viscosity of Water Versus Temperature

21

-------
P = energy dissipated, ft-lb/sec
= hp x 550 ft-lb/sec/hp

Calculation of G for hydraulic mixing:

( u V/2
' P hL ^

p = water density, 62.4 lb/ft3
hL = head loss, ft
t = detention time, sec

Temp. (°F)

Temp. (°C)

Viscosity

x 10 5
(lb-sec/ft2)

32

0

3.746

40

4

3.229

50

10

2.735

60

16

2.359

70

21

2.050

80

27

1.799

90

32

1.595

100

38

1.424

B. MAJOR UNIT PROCESS INFORMATION (CONT.)

3. Adsorption (PAC for cyanotoxin reduction, TOC removal):

Topic

Description

Information

1. Description

Type (wood, coal, other)



Feed location



Feed capacity (lb/day)



Available contact volume (gal)



2. Target contaminant
removal

Influent concentration



Target effluent concentration



3. Unit Process
Evaluation

Required PAC dose (mg/L)



Assigned process capacity *



4. Other design
information or
limitations observed
(mixing to keep carbon
suspended)





22

-------
* Assigned process capacity (use historical data and site specific studies to determine expected reduction in
contaminants by PAC)

23

-------
B. MAJOR UNIT PROCESS INFORMATION (CONT.)

4. Flocculation:

Topic

Description

Information

1. Description

Type (reel,turbine, hydraulic)





Number trains/stages per train





Control (constant/variable
speed)



2. Dimensions

Length per stage:





Width per stage:





Depth per stage:





Total volume:



3. Major Unit Process
Evaluation

Detention Time (min)



4. Other Design
Information or
limitations observed (G
values*)











* See mixing energy calculation in Rapid mix section.

24

-------
B. MAJOR UNIT PROCESS INFORMATION (CONT.)

5. Sedimentation:

Topic

Description

Information

1. Description

Type (Conventional/tube
settlers)



Number of trains



Weir location



Sludge collection



2. Dimensions

Length or Diameter



Width:



Depth:



Total Surface Area:



3. Unit Process
Evaluation

Surface Loading Rate



Assigned process capacity



4. Other design
information or
limitations observed
(sludge removal
capability, ability to
handle carbon
removal)















25

-------
B. MAJOR UNIT PROCESS INFORMATION (CONT.)

6. Filtration:

Topic

Description

Information

1. Description

Type (Mono/dual/mixed)



Description (sand, anthracite,
GAC)



Number of filters



Filter control (Constant/declining
rate)



Surface wash or air scour?



2. Dimensions

Length or diameter:



Width:



Total surface area



3. Media design conditions (depth/effective size/uniformity coefficient)





4. Backwash

Backwash initiation (Time, headloss,
turbidity)



Sequence (surfacewash/air
scour/ramping up/down/filterto
waste)







Backwash storage/ disposal



26

-------
B. MAJOR UNIT PROCESS INFORMATION (CONT.)

6. Filtration (cont.):

Topic

Description

Information

5. Unit Process
Evaluation

Surface Loading rate (gpm/sf)



Assigned process capacity



6. Other design
information or
limitations observed
(abilityto add a filter
aid polymer, abilityto
remove carbon fines)





























27

-------
B. MAJOR UNIT PROCESS INFORMATION (CONT.)

7. Disinfection/oxidation:

Topic

Description

Information

1. Description

Contact Type (Clearwell/storage
tank)



Tio/Tfactor (See Table 4-4 or use
tracer study results)



2. Dimensions

Length or Diameter:



Width:



Minimum Operating Depth:





Total Volume





Volume Adjusted for Tio/T



3. Unit Process

Evaluation

(disinfection)

Disinfectant (free
chlorine/chloramines)





Max. disinfectant residual (mg/L)





Maximum pH





Minimum temperature (°C)





Required Giardia inactivation





Required virus inactivation





Assigned process capacity



3. Unit Process
Evaluation
(cyanotoxin
oxidation)

Toxin reduction (|jg/L)



Required CT



Assigned process capacity



28

-------
5. Other design
information or
limitations









29

-------
C. MISCELLANEOUS EQUIPMENT INFORMATION

1. Miscellaneous Equipment/Unit Processes:

Topic

Description

Information

1. Presedi mentation

Detention Time



Flexibility to by-pass



Chemical feed capacity
(Preoxidant, etc.)



Design limitations







2. Backwash decant
treatment

Description









Recycle practices



Design limitations











3. Sludge Handling

On-sitestorage volume



Long-term disposal



Design limitations



30

-------
C. MISCELLANEOUS EQUIPMENT INFORMATION (CONT.)

2. Chemical Feed Equipment:

Chemical Feed System

•	Chemical name/characteristics
(e.g., product density, strength)

•	Purpose (e.g., coagulant, filter
aid, cyanotoxin removal,
disinfection)

•	Number/type feed pumps or dry
feeders

Capacity (ML/min
or mg/min)

•	Design

•	Operating range

Comments

•	Dose control (e.g., flow paced)

•	Manufacturer's information

•	Calibration method

•	Design issues

1.























2.























3.





























4.

















31

-------
C. MISCELLANEOUS EQUIPMENT INFORMATION (CONT.)

3. Instrumentation:

On-Line Instrumentation

•	Type (e.g., turbidimeter,flow
meter, particle counter, pH
monitor, chlorine monitor,

fluorescence sensor)

•	Manufacturer

Location

• Process
stream

Comments

•	Process purpose

•	Calibration

•	Alarm/shutdown capability

•	Design issues

1.























2.























3.























4.























5.





32

-------
C. MISCELLANEOUS EQUIPMENT INFORMATION (CONT.)

4. Pumping:

Flow Stream Pumped

•	Location

•	Number of pumps

•	Rated capacity

Pump Type

•	Turbine

•	Centrifugal

Comments

•	Flow control method

•	Design issues

•	Source of rated capacity (name plate,
specifications, flow meter)

1.























2.























3.























4.























5.





34

-------


















CD























7.























A. PROCESS CONTROL STATEGY AND COMMUNICATION

Describethe process control strategy used by the staff and associated communication mechanisms.

Topic

Description/Information

1. Process Control Strategy

•	Does the staff set specific

performance targets/goals?
Are they posted?

•	Who sets process control
strategies and decisions?















35

-------
• Are appropriate staff members
involved in process control
and optimization activities?









2. Communication Methods

•	Does the staff have routine
plant/shift meetings?

•	How is communication
conducted among operations,
maintenance, and lab?

•	Does the staff develop and
follow operational procedures?



































36

-------
B. PROCESS CONTROL PROCEDURES

Describe specific process control procedures for the following available processes.

Process

Description/Information

1. Source Water/Intake



Structure



• Source water description/



quality







• Monitoring (turbidity, pH, TOC,



algae identification, chlorophyll,



phycocyanin, cyanotoxins - intra



& extracellular)







• Flexi bi I ity to draw water from



different locations & depths







• Operational problems (e.g.,



capacity limitations, prone to



seasonal algal blooms)











2. Pumping/Flow Control



• Flow measurement and control



• Proportioning to multiple units



37

-------
Process

Description/Information

• Operational problems



3. Presedi mentation

•	Chemicals used / purpose

•	Dose control

•	Monitoring (turbidity, pH, TOC,
algae identification, chlorophyll,
phycocyanin, cyanotoxins - intra
& extracellular)

















•	Sludge removal

•	Operational problems











4. Preoxidation

• Chemicals used / purpose/
location





38

-------
Process

Description/Information

• Dose control

•	Monitoring (oxidant residual,
chlorophyll, phycocyanin,
cyanotoxins - intra &
extracellular)

•	Operational problems

5. Powdered Activated Carbon

• Chemicals used / purpose/
location

• Dose control

• Monitoring (TOC, chlorophyll,
phycocyanin, cyanotoxins - intra
& extracellular)

• Operational problems
(inadequate mixing, insufficient
feed rate)

39

-------
Process

Description/Information





6. Rapid Mix/ Coagulation
• Chemicals used / feed location



•	Dose control (adjustment for
flowchanges; adjustment for
water qual ity - jar testi ng,
streaming current, pilot filter)

•	Monitoring (streaming current)



• Operational problems



7. Flocculation
• Mixing energy adjustment



• Use of flocculant aid



• Monitoring



40

-------
Process

Description/Information

• Operational problems



8. Sedimentation



• Performance goals/



monitoring (turbidity,
chlorophyll, phycocyanin,
cyanotoxins - intra &
extracellular)



• Sludge removal (control,



adjustment)



• Operational problems (e.g.,
turbidity/ carbon carryover,
inadequate sludge removal,
release of cyanotoxins)



9. Filtration



• Performance goals / monitoring
(turbidity, particles, chlorophyll,
phycocyanin, cyanotoxins - intra
& extracellular)



41

-------
Process

Description/Information

• Rate control (constant,
declining)



• Use of filter aid polymer



• Basis for backwash initiation



(turbidity, particles, headloss,



time)



• Backwash procedures (wash



sequence, duration and rates,



basis for returning filterto



service)



• Filter/media inspections



(frequency and type)



• Operational problems (e.g.,
turbidity/ carbon breakthrough,
post backwash turbidity spikes,
short filter runs, insufficient
backwash supply or waste
storage)



42

-------
Process

Description/Information

10. Disinfection



• Performance goals/



monitoring (residual, CT,
cyanotoxins)



• CT factors (pH, minimum depth



of contactor, Tio/T, maximum



residual)



• Operational problems



11. Stabilization



• Chemical used / purpose



• Feed location



• Performance goals/



monitoring (pH, corrosion
index, corrosion inhibitor)



• Operational problems



43

-------
Process

Description/Information

12. Decant Recycle

•	Duration, % of plant flow

•	Type of treatment (settling,
chemical addition)

•	Operational problems (e.g.,
recycle of cyanotoxins)

















13. Sludge Treatment (on-site,
off-site disposal / reuse)



44

-------
C. DATA MANAGEMENT

Describe data collection and management approaches and tools used by plant staff.

Topic

Description/Information

1. Data collection

•	Type of forms used (water
quality testing, shift rounds,
plant log)

•	Computer (SCADA, database)













2. Data application

•	Use of daily, monthly
reports (request examples)

•	Use of trend charts (request
examples)









D. PROBLEM SOLVING AND OPTIMIZATION ACTIVITIES

Describe specific approaches and tools used to solve problems or optimize plant processes.

Topic

Description/Information

1. Problem solving/optimization



45

-------
•	Use of studies (request
example documentation)

•	Pilot plant

•	List recent and ongoing
problem solving/optimization
activities

• Available resources (technical
assistance providers, training,
manuals of practice)

E. COMPLACENCY AND RELIABILITY

Describe specific approaches used to address complacencyand reliability issues inthe plant.

Topic

Description/Information

1. Complacency



• How does utility respond to
unexpected or infrequent water
quality events (e.g., harmful algal
blooms, seasonal changes)?







• Does utility have an emergency



46

-------
Topic

Description/Information

response plan? How does staff



trainfor unusual conditions or



events?



2. Reliability



• Does staff capability to make



process control decisions



exist at more than one level?



• Have process or equipment
limitations/deficiencies been
identified and corrections plans
been developed?







47

-------
F. LABORATORY CAPABILITY

1. Describe available analytical testing capability.

Analytical Capability

Capability V

Description/Comments

• Color





• Jar test





• Particlecounting





• pH





• Solids (dissolved)





• Taste and odor





• Temperature





• Turbidity











• Aluminum





• Calcium





• Fluoride





• Hardness





• Iron





• Magnesium





• Manganese





• Sodium











• Alkalinity





48

-------
• Ammonia Nitrogen





• Nitrite/Nitrate





• Phosphate





• Sulfate











• Chlorine residual





• Bacteriological





• Chlorophyll





• Phycocyanin





• Algae/cyanobacteria cell
identification





• Microcystins (ELISA,
LC/MS/MS, other screening
assays)





• Disinfection byproducts









49

-------
2. Describe laboratory space/equipment and procedures.

Process

Description/Information

Lab Space and Equipment

•	Does adequate lab space exist?

•	Do adequate equipment and
facilities exist?













Lab Procedures

•	Is testing conducted following
standard procedures?

•	Where is lab data recorded?

•	Describe quality control
procedures.





















Equipment Calibration
• Describe procedure for
calibrating turbidimeters











50

-------
• Describe procedures for
calibrating other equipment
(continuouschlorineand pH
monitors)

51

-------
A. MAINTENANCE PROGRAM

Describethe plant maintenance program.

Topic

Description/Information

1. Preventive Maintenance

•	Describe equipment inventory
method (cards, computer).

•	Describe maintenance scheduling
method (daily, weekly, monthly,
annual).



















2. Corrective Maintenance

•	Describethe work order system
(issuing orders/documentation).

•	Describe priority setting
(relationshipto process control
and plant performance needs).

•	List major equipment out of
service within last 6 months.























52

-------




3. Predictive Maintenance
• Describe methods used to
predict maintenance needs
(vibration, infrared analysis).









4. Housekeeping

• Does poor housekeeping detract
from plant performance/image?











53

-------
B. MAINTENANCE RESOURCES

Describethe available maintenance resources at the plant.

Topic

Description/Information

1. Equipment Repair and Parts

•	Are critical spare parts stored at
the plant?

•	Can vendors provide quick
responseto spare parts needs?

•	What is the policy on parts
procurement by staff?





















2. Maintenance expertise

•	Describe staff expertise
(mechanical, electrical,
instrumentation).

•	Does the staff use any contract
maintenance services? How
responsive are they to needs?

•	Do staff develop and use
maintenance procedures?



















54

-------
3. Work Space and Tools
• Does the plant have adequate
work space and tools to perform
maintenance tasks?











4. Performance Monitoring
• How is maintenance performance
measured (timeto complete
task, work order backlog)?











55

-------
A. HISTORICAL WATER PRODUCTION DATA

1. Use the following table to determine the peak instantaneous operating flow for the plant.

Month/Year

Maximum
Daily Flow

Operating
Time per Day

Flow during
Operation (1>

Instantaneous
Peak Flow ,2)





















































































































































































56

-------
(1) If a plant operates less than 24 hr/day, flow during operation can be determined from the equation
below:

Qt 24 hr

Qa= —x ,

T day

Qa = Average flow during operation
Qt = Total flow in 24-hour period
T = Time of plant operation, hours

,2) Peak instantaneous flow through a plant is often different than the average flow due to changing
water demands that the plant must meet. The peak instantaneous flow during a day can
sometimes be obtained from plant logs (e.g., raw pump operation, rate change timeand flow).

57

-------
B. WATER USAGE

1. Determine the water usage per capita based on water production records and population served.
Water usage statistics forthe United States are shown in the table below.

Qc=ir

Qc = Usage per capita per day
Qt = Total flow in 24-hour period
P = Population served

Population 	

Qc Avg.		

Qc Peak		

58

-------
State

Use (gpcpd)

State

Use (gpcpd)

Alabama

191

Nebraska

174

Alaska

134

Nevada

306

Arizona

191

New Hampshire

85

Arkansas

154

New Jersey

131

California

175

New Mexico

184

Colorado

188

New York

166

Connecticut

120

North Carolina

107

Delaware

124

North Dakota

114

Florida

146

Ohio

127

Georgia

160

Oklahoma

173

Hawaii

180

Oregon

164

Idaho

163

Pennsylvania

128

Illinois

154

Rhode Island

115

Indiana

115

South Carolina

148

Iowa

131

South Dakota

121

Kansas

144

Tennessee

148

Kentucky

128

Texas

176

Louisiana

147

Utah

255

Maine

81

Vermont

80

Maryland

165

Virginia

119

Massachusetts

119

Washington

217

Michigan

136

West Virginia

96

Minnesota

105

Wisconsin

118

Mississippi

127

Wyoming

188

Missouri

131

Puerto Rico

115

Montana

164

Virgin Islands

63

Source: Solley, W. B., Preliminary Estimates of Water Use in the United States, 1995,
U.S. Geological Survey (1997)

59

-------
2.

Determine unaccounted for water based on monthly or annual water production and meter
records. Unaccounted for water typically varies from 10to 12 percent for new systems and 15to
30 percent for older systems (Metcalf and Eddy, Inc. 1991).

Q°-= x10°

Vt

Q% =% unaccounted

Qt = Total plant water production for month or year
Qm = Total metered water for month or year

Qt		

Qm

Q%		

3. Determine backwash water percent based on volume of water filtered and volume of water used
for backwash. Typically, the amount of water used for backwash ranges for 2 to 6 percent for
conventional plants. Higher percentages can occurfor direct filtration plants.

BW =(VF VBw)x1QQ

VF

BW% =% backwash water

Vf = Volume of water filtered

Vbw = Volume of water used for backwash

Vf		

Vbw

BW%

60

-------
C. IN-PLANT STUDIES

Describe results of in-plant studies conducted during the CPE.

Topic

Description/Information/Findings

1. Filter media evaluation

•	Check media depth and type

•	Check media condition (presence
of chemicals/debris, mudballs,
worn media)

•	Check support gravel level
(variation of less than 2 inches
acceptable)































2. Backwash evaluation
• Check backwash rate (measure
rise rate inthe filter versus time
and convert to backwash rate;
> 15 gpm/ft2 acceptable)











61

-------
Check bed expansion
> 20 percent acceptable)

-------
C. IN-PLANT STUDIES (CONT.)

Describe results of in-plant studies conducted during the CPE.

Topic

Description/Information/Findings

2. Backwash evaluation (cont.)

• Observe backwash procedure

(flow distribution, ramping of flow

rate, turbidity of water at end of
backwash)





















3. Coagulant dosage evaluation
• Verify reported dose with actual;
measure liquid or dry feed rate
(Ib/min, mL/min) and convert to
dose (mg/L)



















4. Turbidity meter evaluation



63

-------
Check meter calibration or
compare with calibrated meter

-------
C. IN-PLANT STUDIES (CONT.)

Data Integrity Study 1: Sample Line Detention Calculation

Use the table below to calculate the detention time in the sample line from each sample tap
to the turbidimeter. Is the detention time excessive or are there other findings?

Sample Tap
Description

Line Volume,
gal

Line Flow Rate,
gal/min

Line Detention Time,
min

































































65

-------
C. IN-PLANT STUDIES (CONT.)

Data Integrity Study 2: Turbidimeter Settings

Review the turbidimeter settings and record findings in the table below.

Turbidimeter Location

















Turbidimeter Model

















Controller Model and Data
Logging Setting (1)

















Signal Averaging (2)

















Bubble Reject (3)

















Error Hold Mode (4)

















Output Span (5)

















Other

















(1)	Check to see if current data and time are correct. Check frequency of data logging. Default is 15 minutes for Hach models.

(2)	Default for Hach models is 30 seconds. This is acceptable in most cases.

(3)	Default is Yes for Hach models. This is acceptable in most cases.

(4)	Specific to Hach 1720E and FilterTrak 660 models. Default is to Hold Outputs (HO) and send last known value to SCADA when turbidimeter loses
communication with controller. Better option is Transfer Ouputs (TO) to send an operator-selected value to SCADA (e.g., 0, 10) to make operator aware
of problem.

(5)	To avoid "capping" of data to SCADA, the output span should be at least 0 to 5.1 NTU (applicable to analog signals).

Accessing output span for Hach SC200 controller: Menu/SC200 setup/Output setup (select 1 or 2; select Source to see which turbidimeter is highlighted
and then Back button)/Activation (low value; high value).

66

-------
C. IN-PLANT STUDIES (CONT.)

Data Integrity Study 2 (cont.)

Findings:

-------
C. IN-PLANT STUDIES (CONT.)

Data Integrity Study 3: Turbidity Data Signal Verification

Simultaneous readings of the signal output monitor at a turbidimeter, as well as remote
locations such as the HMI monitor or a PLC readout.

Turbidimeter

Instrument Values

Remote Location No.1
and Values

Remote Location No.2
and Values

























































Findings:

68

-------
C. IN-PLANT STUDIES (CONT.)

Data Integrity Study 4: Identification and Verification of Log Removal Value (LRV) Calculation Input Parameters

Identify the location of the Log Removal Value (LRV) calculation input variables shown in the table below. Provide as much
specific information as possible (e.g., instrument tag number, physical location).

LRV Input

Plant Location Input Measured
or Obtained

Data Assess (e.g., SCADA
screen, historian report)

Comments on Data Integrity
(e.g., frequency of meter
calibration)

Pressure Decay Test Start and End
Pressures (PTest Start, PTest End, psig)

Enter test values below:







Pressure Decay Test Hold Time
(minutes)

Enter test value below:







Back Pressure (BP, psig)(1)
Enter test value below:







Atmospheric Pressure (Patm, psi)(2)
Enter test value below:







69

-------
LRV Input

Plant Location Input Measured
or Obtained

Data Assess (e.g., SCADA
screen, historian report)

Comments on Data Integrity
(e.g., frequency of meter
calibration)

Water Flow Rate (Qfiow, gpm),3>
Enter test value below:







Water Temperature (F)
Enter test value below:







Transmembrane Pressure (TMP,
psig)(3)

Enter test value below:







Vendor Provided Parameters:
System volume (L)

Volume concentration factor
(VCF)

Resistance coefficient (K)

Net expansion factor (Y)

(1)	Water head above the DIT sensor (feet converted to psi).

(2)	Sea level pressure adjusted to site elevation.

(3)	Value or average value if variable prior to the DIT.

70

-------
C. IN-PLANT STUDIES (CONT.)

Data Integrity Study 4 (cont.)

Findings:

-------
c.

IN-PLANT STUDIES (CONT.)

Data Integrity Study 5: Portable and Bench Turbidimeter Monitoring and Comparison to
Online Turbidimeters

Develop and implement a sampling plan to compare individual filter(s) effluent turbidity
using a portable or bench top turbidimeter and compare the results with the continuous
reading turbidimeter(s). Follow quality control procedures to assure comparable results
between grab samples and continuous meters (i.e., use indexed sample cells and use
clean, scratch-free sample cells, oil cells, and de-gas samples). Take three measurements
for each turbidimeter and average the results. Lab and portable turbidimeter readings for
filtered water should be within 0.05 NTU of online turbidimeter readings.

Sample Location

Continuous
Turbidimeter
Readings, NTU

Lab Turbidimeter
Readings, NTU

Portable
Turbidimeter
Readings, NTU









C. IN-PLANT STUDIES (CONT.)

Data Integrity Study 5 (Cont.)

Findings:

72

-------
73

-------
C. IN-PLANT STUDIES (CONT.)

Describe results of in-plant profiles conducted during the CPE (e.g., manganese, cyanotoxins).

Process	Parameter	Parameter	Parameter	Parameter

Location

74

-------
A. INTERVIEW GUIDELINES

The following interview guidelines are provided to assist CPE providers with the interview process.

1. Conduct interviews with one staff person at a time in a private location.

¦ It is important to create a comfortable environment for the interview process to take place.
Confidentiality of the interview should be explained

2. Keep the interview team size small.

¦ The number of people included on each interviewteam should be kept to a minimum (e.g., 1 to 3) to
avoid overwhelming the person being interviewed. If more than one person is included on the
team, one person should be assigned as the lead interviewer.

3. Allow 30 to 45 minutes for each interview.

¦ Interview times will vary depending on the personality of the individual being interviewed and the
number and type of issues involved. It is the responsibility of the interviewer to maintain the focus
on performance-related issues. Interviews can easily be detracted by individuals who find an "open
ear" for presenting grievances.

4. Explain the purpose of the interview and use of the information.

¦ It is important for the people being interviewed to understand that any information obtained from
this process is only used to support identification of factors limiting performance (i.e., areas
impacting performance). The interview information is not used to place blame on specific
individuals or departments.

5. Conduct interviews after sufficient information has been gathered from CPE activities.

¦ Utilize results and observations gained from the plant tour, performance assessment, major unit
process evaluation, and data collection activitiesto identify areas of emphasis during the interviews.

6. Progress through the interview in a logical order.

¦ For example, if an administrator is being interviewed, focus questions on administrative support,
then on design issues,followed by operation and maintenance capabilities.

7. Ask relevant questions with respect to staff area of involvement.

¦ For example, when interviewing maintenance personnel, ask questions related to relevant topics
such as maintenance responsibilities, communication with supervisors, and administrative support
for equipment.

8. Ask open-ended questions.

-------
¦ For example, a question such as "Are you aware of any design deficiencies with the current plant? "
would provide better information than a question like "Do you think that the flocculation basin
provides sufficient detention timefor flocculation?".

9. Ask the questions: don't give the answers.

¦ The purpose of the interviewis to gain the perspective of the person being interviewed. Askthe
question, and wait for the response (i.e., don't answer your own question based on information you
may have received from previous activities). Rephrasingthe question may sometimes be necessary
to provide clarity.

10. Repeat a response to a question for clarification or confirmation.

¦ For example, the interviewer can confirm a response by stating, "If I understand you correctly, you
believe that the reason for poor plant performance during April was due to excessive algae growth
in the source water."

11. Avoid accusatory statements.

¦ Accusatory statements will likely lead to defensiveness by the person being interviewed. Rather, if
an area of concern is suspected, ask questions that can confirm or clarify the situation.

12. Use the interview to clarify or confirm field information.

¦ For example, if performance problems occurred during one month of the past year, ask questions to
clarifythe perceived reasons forthese problems.

13. Note specific responses that supports factor identification.

¦ During or followingthe interview, the interviewer may want to note or underline specific responses
that support the identification of possible factors limiting performance. This summary can then be
used during team debriefing and factor identification meetings.

-------
B. PERSONNEL INTERVIEW FORM

Name	 Title 	

Time at plant	 Years of experience

Education/training/certification 	

Interview notes (concerns, recommendations in administration, design, operation, and maintenance):

77

-------
A. ATTENDANCE LIST

Utility Name		Date

Name

Title/Position

Telephone No.













































































































78

-------
B. MUTIPLE BARRIER CONCEPT FOR MICROBIAL CONTAMINANT PROTECTION

Coagulant
Addition

Variable
Quality
Source
Water

jU

era
»»»

* * » » »

czH-czi
»»|i| »»

». y *»

Flocculation/Sedimentation
Barrier

' Turbidity ^

Goal
V< 0.1 NTU/

mmmmm

Disinfectant
Addition

/ Achieve \
/ Inactivation\
\ Goal /

Filtration
Barrier

Finished
Water

Disinfection
Barrier

•	Given a variable quality source water, the treatment objective is to produce a consistent, high
quality finished water.

•	Protozoan parasites, such as Giardia and Cryptosporidium, are found in most source waters;
however, it is difficultto quantify their presence and assess their viability.

•	Microbial pathogens in the source water, such as protozoan parasites, bacteria, and viruses, can
be physically removed as particles intreatment processes and inactivated through disinfection.

•	Multiple barriers are provided in a treatment plant to remove or inactivate microbial pathogens.

•	Key treatment barriers includeflocculation/sedimentation, filtration, and disinfection.

•	Since measurement of protozoan parasites is difficult, surrogate parameters, such as turbidity,
particle counting, and pathogen inactivation, are used to assess the performance of each barrier.

79

-------
C. OPTIMIZATION PERFORMANCE CRITERIA

A summary of performance criteria for surface water treatment plants to provide protection

against microbial contaminants is presented below:

I.	Minimum Data Monitoring Requirements

¦	Daily raw water turbidity

¦	Settled water turbidity at 4-hour time increments from each sedimentation basin

¦	On-line (continuous) turbidityfrom each filter

¦	One filter backwash profile each month from each filter

II.	Individual Sedimentation Basin Performance Criteria

¦	Settled water turbidity less than 1 NTU 95 percent of the time when annual average raw
water turbidity is less than or equal to 10 NTU

¦	Settled water turbidity less than 2 NTU 95 percent of the time when annual average raw
water turbidity is greater than 10 NTU

III.	Individual Filter Performance Criteria

¦	Filtered water turbidity less than 0.10 NTU 95 percent of the time (excluding 15-minute
period following backwashes) based on the maximum values recorded during 4-hour time
increments

¦	Maximum filtered water measurement of 0.3 NTU

¦	Initiate filter backwash immediately after turbidity breakthrough has been observed and
before effluent turbidity exceeds 0.10 NTU.

¦	Maximum filtered water turbidity following backwash of less than 0.3 NTU

¦	Maximum backwash recovery period of 15 minutes (i.e., return to less than 0.10 NTU)

¦	Maximum filtered water measurement of less than 10 particles (inthe 3 to 18 |jm range)
per milliliter (if particle counters are available)

IV.	Disinfection Performance Criteria

¦	CT values to achieve required log inactivation of Giardiaanti virus

80

-------
Study Format, Elements, and Template

Study Topic: Identify the name of the study and briefly describe why the study is being conducted
(i.e., one to two sentences).

Hypothesis:

Describe what is to be proved by completing the study (show cause/effect relationship).

Focus study on a specific activity.

Approach and Resources:

Describe how the study will be conducted (i.e., processes and equipment involved).

Describe resources required (i.e., staff, sampling, and testing).

Involve plant staff in development (operations, maintenance, and laboratory).

Determine whether any background data is needed before initiating the study.

Duration of Study:

Define the time estimated to complete the study (important to clarify for staff).

Expected Results:

Describe expected results from the study.

Describe how the data will be presented to support the hypothesis.

Define measures of success for the study.

Summary & Conclusions:

To be completed at the end of the study.

Document results of the study (brief written summary with charts).

Present findings to utility staff and management (use as training tool for all utility staff).

Implementation:

To be completed at the end of the study.

Document changes to current plant procedures based on study results.

81

-------
Study Topic:

Hypothesis:

Approach & Resources:

Duration of Study:

Expected Results:

Summary and Conclusions:

Implementation:

82

-------
Cyanotoxin Oxidation Hold Study Protocol

Overview:

The objective of the Cyanotoxin Oxidation Hold Study is to simulate water quality dynamics relative to
cyanotoxin oxidation within the clearwell of a water treatment plant. During this study, water is
collected from a location between the filters and clearwell prior to chlorination (e.g., combined filter
effluent tap), dosed with known concentrations of a concentrated cyanotoxin solution and chlorine (if
not previously added in the treatment process) and held in a container to simulate clearwell conditions.
Water quality samples are periodically collected and are used to estimate the oxidation rate of
cyanotoxins in the water.

Hypothesis:

The hypothesis of this study will be system-specific, depending on the desired objective of the study (see
the Overview section, above).

Resources:

•	Required Personnel:

o One to two (1-2) investigators

•	Required Equipment:

o Large Erlenmeyer flask (e.g., 6 liter) prepared chlorine demand-free, and wrapped in
aluminum foil to minimize UV light penetration

Note: All glassware should be pretreated to be chlorine demand-free using the
following, or similar, procedure:

¦	Completely fill each glass container with a 10 - 20 mg/L chlorine solution, by
adding 0.3 mL of household bleach1 (typically 5.25% w/v), or stock chlorine
solution of comparable strength, per liter of water2. Assuming a household
bleach of 5% chlorine (SDS states 5-10%) and a target chlorine concentration of
15 mg/L, it would take 0.32 mL of bleach per liter of water.

¦	Allow the chlorine solution to soak in the containers for at least 24 hours.

¦	Thoroughly rinse each bottle three times with water2.

o One (1) portable colorimeter with necessary instructions and reagents fortotal chlorine,
free chlorine (DPD and indophenol method reagents), monochloramine, and free
ammonia residual analysis

o Magnetic stir plate and large stir bar

1	Confirm that product contains only sodium hypochlorite and does not include other chemicals or fragrances.

2	Water used to prepare glassware chlorine demand-free should be of the highest quality available. If laboratory
clean water (RO/IX/GAC, distilled, or deionized) is not available, treatment plant effluent water may be used.

83

-------
o Sample bottles, preservatives, and quenching agents for cyanotoxin analysis

o Sample bottles, filtration apparatus forTOC and DOC. UV spectrophotometer and quartz
cuvette for UV254 analysis

o 50-100 mL glass Griffen beakers for sample collection, prepared chlorine demand-free
as noted above

o 250 mL amber glass Packer bottles with caps, prepared chlorine demand-free as noted
above fordemand study.

o Glass pipets prepared chlorine demand-free with rubber bulb

o Pipettes and disposable pipette tips (pipette volumes dictated by necessary dosing)

o One (1) pH meter with calibration standards

o One (1) digital thermometer

o Deionized (Dl) water

o One (1) water bath to incubate samples at the clearwell temperature if unable to
conduct hold study in a temperature-controlled environment. Options include:

¦ Laboratory water bath or incubator that can maintain a specific temperature
(preferred)

¦ Container designed (or modified) for continuous flow-through of study water
(i.e., plant effluent, sink's cold tap)

¦ Cooler filled with water and changed periodically

Procedure

1. Make a chlorine stock solution and standardize its concentration, according to the protocol
found in Appendix A of EPA Method 127 (p. 24-27).

2. To determine the appropriate chlorine dose for the hold study, an oxidant demand study may
need to be conducted. This is especially beneficial when using a concentrated cyanotoxin spike
from an ambient water body as the challenge forthe hold study, as there could be additional
chlorine demanding constituents, such as ammonia, organics, iron or manganese, that exert
oxidant demand concurrent with cyanotoxins. The objective of this study is to determine the
appropriate chlorine dose to achieve breakpoint chlorination such that free chlorine residual is
available for cyanotoxin oxidation.

a) The demand study is conducted using 250 mL amber glass bottles with caps, all prepared to
be chlorine demand-free (see above procedure). The desired free chlorine residual can be
used as a benchmark, and a range of doses selected based on that target residual.

-------
b) The challenge water is prepared using a plant water sample (post-filter, pre-chlorination)
and the concentrated cyanotoxin solution to achieve the desired cyanotoxin concentration.
Bottles are then filled with this challenge water.

c) Each bottle is then dosed with chlorine, at varying doses within the range selected
previously. To dose the chlorine, the appropriate amount of challenge water is pipetted out
and replaced by chlorine stock (see Free Chlorine Stock UV-VIS (SOP-V4).docx for a protocol
on making and standardizing the chlorine stock solution; see Toxin Oxidation Hold Study.xlsx
spreadsheet for calculating doses based on the chlorine stock solution concentration).

d) The bottles are then held for a sufficient amount of time such that breakpoint chlorination
can be observed. Intermediate samples can be taken to better understand the dose and
time where breakpoint occurs and better inform sampling later in the study. Typically,
breakpoint chlorination will occur relatively quickly («15 min.), but under certain conditions
additional time may be needed3 for the breakpoint reactions to take place.

e) For each sample, total and free chlorine are measured using the colorimetric DPD method,
and paired with free chlorine analysis by the indophenol method due to the potential for
interferences. The indophenol method is less prone to positive interference from
chloramines and manganese than the DPD free chlorine method.

f) At the end of the study, the dose that resulted in the desired free chlorine residual is the
dose that should be used for the hold study.

3. Calculate the appropriate dosing volume for the volume of challenge water. Determine the
necessary volume of challenge water needed based on the number and volume of samples to be
taken during the hold study.

4. Set up a sampling plan. It is important to take frequent samples in the initial moments of the
hold study once chlorine is dosed, as the initial chlorine and cyanotoxin decay can often occur
quickly. For example, immediately after dosing, a 30-second sample should be taken, and
thereafter at approximately 5-minute intervals forthe first 30 minutes, or as frequently as
sample analysis will allow. Samples can be collected less frequently after the first hour of the
study, such as 15-minute or30-minute sampling intervals. Typically, free and total chlorine and
cyanotoxins samples are analyzed. TOC/DOC/UV254 can also be measured as deemed
appropriate. Other chlorine demanding constituents such as free ammonia, iron, manganese,
and TOC should also be measured prior to the start of the hold study. If free ammonia is present
in the challenge solution, it is recommended that a chlorine demand study should first be
conducted, so the appropriate chlorine dose maybe determined. Monochloramine, free
ammonia, and free chlorine by indophenol method should also be analyzed to ensure that
breakpoint chlorination has occurred. The initial chlorine dose may need to be increased to
achieve breakpoint chlorination and the desired free chlorine residual formicrocystins oxidation
analysis.

3 The U.S. EPA Office of Research & Development has created a web-based application

dittps://usepaord sliinyapps.io/Breakpoint-Curve/) that may be used to estimate the time needed for breakpoint

reactions to take place under specific conditions (e.g., pH, temperature).

-------
5. Make the challenge solution (see the figure below) by combining a sample from the water
treatment plant process just prior to chlorine addition before water enters the clearwell with
the concentrated cyanotoxins solution. The latter can be a laboratory grade cyanotoxin
standard, or be concentrated from an ambient water body with a phytoplankton net. If opting
for the latter, the sample will need to undergo freeze/thaw and filtration through a 0.45 fim
glass fiber filter to ensure that the cyanotoxins are extracellular. If cyanotoxins break through to
the clearwell at a water treatment plant, they will likely be in extracellular form, as the
coagulation/flocculation/sedimentation and filtration processes would likely remove the
intracellular cyanotoxins by removing the cyanobacteria cells.

6. Dose the chlorine to the challenge solution and mix at a slow rate to mimic clearwell conditions.
Collect samples according to the sampling plan and analytical methods. Sampling and analysis
vessels should be rinsed with Dl water after each sample is analyzed to prevent cross-
contamination of subsequent samples.

86

-------
7. Plot the sample results with time on the x-axis and chlorine residual and cyanotoxin

concentrations on the y-axis. This is helpful to visualize the decay curves and make informed
process control decisions.

Considerations:

•	Headspace vs. headspace-free?

•	Temperature control during the study (water bath, controlled temperature room, etc.)

•	Limit UV penetration (such as wrapping hold study vessel in aluminum foil, or using amber glass)

•	How to introduce the cyanotoxins? Laboratory-grade cyanotoxin solution from a vendor vs.
concentrating a cyanotoxin solution from ambient water using a phytoplankton net.

87

-------
HABCPE Performance-Limiting Factors (PLFs)

CPE Factor Summary Sheet Terms

Plant Type

Brief but specific description of plant type (e.g., conventional
with flash mix, flocculation, sedimentation, filtration and
chlorine disinfection; or direct filtration with flash mix,
flocculation and chlorine disinfection).

Source Water

Brief description of source water (e.g., surface water including
name of water body).

Performance Summary

Brief description of plant performance based on performance
assessment component of the CPE (i.e., ability of plant to meet
optimized performance goals).

Ranking Table

A listing of identified performance limiting factors that directly
impact plant performance and reliability.

Rank

Relative ranking of factor based on prioritization of all "A" and
"B" rated factors identified during the CPE.

Rating

Rating of factor based on impact on plant performance and
reliability:

A — Major effect on a long-term repetitive basis

B — Moderate effect on a routine basis or major effect on a
periodic basis

C — Mi nor effect

Performance Limiting
Factor (Category)

Factor identified from Checklist of Performance Limiting
Factors, including factor category (e.g., administration, design,
operation, and maintenance).

Notes

Brief listing of reasons each factor was identified (e.g., lack of
process control testing, no defined performance goals).

88

-------
CPE Performance Limiting Factors Summary

Plant Name/Location:

CPE Performed By:

CPE Date:

Plant Type:

Source Water:

Performance Summary:

Ranking Table

Rank

Rating

Performance Limiting Factor (Category)

Rating Description

A — Maj or effect on long-term repetitive basis.

B — Moderate effect on a routine basis or maj or effect on a periodic basis.
C — Minor effect.

89

-------
Performance Limiting Factors Notes

Factor

Notes



•



•



•



•

90

-------
Checklist of Performance Limiting Factors

A. ADMINISTRATION

1.	Plant Administrators

1.	~ Policies		

2.	~ Familiarity With Plant Needs		

3.	~ Supervision		

4.	~ Planning		

5.	~ Complacency		

6.	~ Reliability		

7.	~ Source Water Protection		

2.	Plant Staff

1.	~ Number		

2.	~ Plant Coverage		

3.	~ Personnel Turnover		

4.	~ Compensation		

5.	~ Work Envi ronment		

6.	~ Certification		

3.	Financial

1.	~ Operating Ratio		

2.	~ Coverage Ratio		

3.	~ Reserves		

2. DESIGN

1.	Source Water Quality

1. ~ Microbial Contamination

2.	Unit Process Adequacy

1.	~ Intake Structure

2.	~ Presedimentation Basin

3.	~ Raw Water Pumping

4.	~ Flow Measurement

5.	~ Chemical Storage and Feed

Facilities

6.	~ Flash Mix

7.	~ Flocculation

8.	~ Sedimentation

91

-------
9.	~ Filtration

10.	~ Disinfection

11.	~ Sludae/Backwash Water

Treatment and Disposal

3. Plant Operability

1.	~ Process Flexibility

2.	~ Process Controllability

3.	~ Process Instrumentation/

Automation

4.	~ Standby Units for Kev

Equipment

5.	~ Flow Proportioning

6.	~ Alarm Systems

7.	~ Alternate Power Source

8.	~ Laboratory Space and Equipment

9.	~ Sample Taps

OPERATION

1.	Testing

1.	~ Process Control Testing

2.	~ Representative Sampling

2.	Process Control

1.	~ Time on the Job

2.	~ Water Treatment Understanding

3.	~ Application of Concepts and

Testing to Process Control

3.	Operational Resources

1.	~ Training Program

2.	~ Technical Guidance

3.	~ Operational Guidelines/Procedures

MAINTENANCE

1. Maintenance Program
1. ~ Preventive

92

-------
2. ~ Corrective

3. ~ Housekeeping

2. Maintenance Resources

1.	~ Materials and Equipment

2.	~ Skills or Contract Services

Definitions for Assessing Performance Limiting Factors

NOTE: The following list of defined factors is provided to assist the evaluator with identifying
performance limitations associated with protection against microbial contaminants in water
treatment systems. Performance limiting factors aredescribed below using the following format.

A. CATEGORY

1. Subcategory

a. Factor Name

~ Factor description

> Example of factor applied to specific plant or utility

A. Administration

1. Plant Administrators
1. Policies

~ Do existing policies or the lack of policies discourage staff members from making
required operation, maintenance, and management decisions to support plant
performance and reliability?

>	Utility administration has not communicated a dear policy to optimize plant
performance for public health protection.

>	Multiple management levels within a utility contribute to unclear
communication and lack of responsibility for plant operation and
performance.

>	Cost savings is emphasized by management at the expense of plant
performance or at the expense of HAB preparedness.

93

-------
>	Utility managers do not support reasonable training and certification
requests by plant staff.

>	Administration continues to allow connections to the distribution system
without consideration for the capacity of the plant.

2.	Familiarity With Plant Needs

~	Do administrators lack first-hand knowledge of plant needs?

>	The utility administrators do not make plant visits or otherwise communicate
with plant staff.

>	Utility administrators do not request input from plant staff during budget
development.

>	Administrators are not familiar with HAB preparedness needs at the plant
(e.g., on site PAC, chemical supplier reliability, critical equipment O&M
status).

3.	Supervision

~	Do management styles, organizational capabilities, budgeting skills, or
communication practices at any management level adversely impact the plant to
the extent that performance is affected?

>	A controlling supervision style does not alio w the plant staff to contribute to
operational decisions.

>	A plant supervisor's inability to set priorities for staff results in insufficient
time allocated for process control.

4.	Planning

~	Does the lack of long range planning for facility replacement or alternative source
water quantity or quality adversely impact performance?

>	A utility has approved the connection of new customers to the water system
without considering the water demand impacts on plant capacity.

>	An inadequate capital replacement program results in utilization of outdated
equipment that cannot support optimization goals.

~	The utility does not have sufficient capability to handle additional sedimentation
and backwash sludge/decant treatment or disposal.

>	A HAB event results in additional sludge production and backwash waste
overloading existing facilities.

>	A HAB event results in the need to stop waste decant recycle in the plant and
an alternative disposal option is not available (e.g., discharge to sanitary
sewer, discharge to receiving stream with NPDES permit).

5.	Complacency

~	Does the presence of consistent, high quality source water result in complacency
within the water utility?

>	Due to the existence of consistent, high quality source water, plant staff are
not prepared to address unusual water quality conditions.

94

-------
>	A utility does not have an emergency response plan in place to respond to
unusual water quality conditions or events.

>	Utility does not have a contingency plan to prepare for a HAB event including
an alternate raw or finished water source and considerations for
simultaneous treatment objectives (e.g., DBPs, corrosion control).

>	Utility has perception that a HAB event is not likely at their utility, and this
position has deterred them from being prepared (e.g., monitoring, providing
treatment).

6.	Reliability

~	Do inadequate facilities or equipment, or the depth of staff capability, present a
potential weak link within the water utility to achieve and sustain optimized
performance?

>	Outdated filter control valves result in turbidity spikes in the filtered water
entering the plant clearwell.

>	Plant staff capability to respond to unusual water quality conditions exists
with only the laboratory supervisor.

7.	Source water management and planning

~	Does the utility have the ability to access multiple water sources; does the plant
have the ability to draw water from multiple intake locations or water levels?

>	The utility is limited to one intake location during a HAB event.

8.	Source Water Protection

~	Does the water utility lack an active source water protection program?

>	The absence of a source water protection program has resulted in the failure
to identify and eliminate the discharge of failed septic tanks into the utility's
source water lake.

>	Utility management has not evaluated the impact of potential contamination
sources on water quality within their existing watershed including HABs.

2. Plant Staff

1.	Number

~	Does a limited number of people employed have a detrimental effect on plant
operations or maintenance?

>	Plant staff are responsible for operation and maintenance of the plant as well
as distribution system and meter reading, limiting the time available for
process control testing and process adjustments.

2.	Plant Coverage

~	Does the lack of plant coverage result in inadequate time to complete necessary
operational activities? (Note: This factor could have significant impact if no
alarm/shutdown capability exists - see design factors).

>	Staff are not present at the plant during evenings, weekends, or holidays to
make appropriate plant and process control adjustments.

95

-------
>	Staff are not available to respond to changing source water quality
characteristics.

Personnel Turnover

~	Does high personnel turnover cause operation and maintenance problems that
affect process performance or reliability?

>	The lack of support for plant needs results in high operator turnover and,
subsequently, inconsistent operating procedures and low staff morale.

Compensation

~	Does a low pay scale or benefit package discourage more highly qualified persons
from applying for operator positions or cause operators to leave after they are
trained?

>	The current pay scale does not attract personnel with sufficient qualifications
to support plant process control and testing needs.

Work Envi ronment

~	Does a poor work environment create a condition for "sloppy work habits" and
lower operator morale?

>	A small, noisy work space is not conducive for the recording and
development of plant data.

Certification

~	Does the lack of certified personnel result in poor 0 & M decisions?

>	The lack of certification hinders the staff's ability to make proper process
control adjustments.

3. Financial

1.	Operating Ratio

~	Does the utility have inadequate revenues to cover operation, maintenance, and
replacement of necessary equipment (i.e., operating ratio less than 1.0)?

>	The current utility rate structure does not provide adequate funding and
limits expenditures necessary to pursue optimized performance (e.g.,
equipment replacement, chemical purchases, spare parts).

2.	Coverage Ratio

~	Does the utility have inadequate net operating profit to cover debt service
requirements (i.e., coverage ratio less than 1.25)?

>	The magnitude of a utility's debt service has severely impacted expenditures
on necessary plant equipment and supplies.

3.	Reserves

~	Does the utility have inadequate reserves to cover unexpected expenses or future
facility replacement?

>	A utility has a 40-year-old water treatment plant requiring significant
modifications; however, no reserve account has been established to fund
these needed capital expenditures.

96

4.

5.

-------
2. Design

1.	Source Water Quality

1. Microbial Contamination

~	Does the presence of microbial contamination sources in close proximity to the
water treatment plant intake impact the plant's ability to provide an adequate
treatment barrier?

>	A water treatment plant intake is located downstream of a major wastewater
treatment plant discharge and is subject to a high percentage of this flow
during drought periods.

2.	Unit Process Adequacy

1.	Source Treatment

~	Does the lack of source water treatment facilities result in degraded water quality?

>	Inadequate mixing or aeration of the source water results in stagnant water
that supports HABs.

2.	Intake Structure

~	Does the design of the intake structure result in excessive clogging of screens,
excessive detention time, build-up of silt, or passage of material that affects plant
equipment?

>	The location of an intake structure on the outside bank of the river causes
excessive collection of debris, resulting in plugging of the plant flowmeter
and static mixer.

>	High detention time in uncovered intake structure results in excessive algae
growth.

>	The design of a reservoir intake structure does not include flexibility to draw
water at varying levels to minimizealgae concentration.

3.	Presedimentation Basin

~	Does the design of an existing presedimentation basin or the lack of a
presedimentation basin contributeto degraded plant performance?

>	The lack of flexibility with a presedimentation basin (i. e., number of basins,
size, bypass) causes excessive algae growth, impacting plant performance.

>	A conventional plant treating water directly from a "flashy" stream
experiences performance problems during high turbidity events.

4.	Raw Water Pumping

~	Does the use of constant speed pumps cause undesirable hydraulic loading on
downstream unit processes?

>	The on-off cycle associated with ra w water pump operation at a plant results
in turbidity spikes in the sedimentation basin and filters.

5.	Flow Measurement

~	Does the lack of flow measurement devices or their accuracy limit plant control
or impact process control adjustments?

>	The flow measurement device in a plant is not accurate, resulting in
inconsistent flow measurement records and the inability to pace chemical
feed rates according to flow.

97

-------
6.	Chemical Storage and Feed Facilities

~	Do inadequate chemical storageand feed facilities limit process needs in a plant?

>	Inadequate chemical storage facilities exist at a plant, resulting in excessive
chemical handling and deliveries.

>	Capability does not exist to measure and adjust the coagulant and flocculant
feed rates.

>	Plant has inability to feed high PAC dose (i.e., > 20 mg/L) to treat for a HAB
event (i. e., storage and feed equipment).

>	Plant has inability to feed PAC because of design limitations (e.g., direct,
pressure filters).

7.	Flash Mix

~	Does inadequate mixing result in excessive chemical dose, insufficient

coagulation, or inability to suspend PAC to the extent that it impacts plant

performance?

>	A static mixer does not provide effective chemical mixing throughout the
entire operating flow range of the plant.

>	Absence of a flash mixer results in less than optimal chemical addition and
insufficient coagulation.

>	High PAC feed in rapid mix results in PAC settling to bottom of basin or
mechanical failure.

8.	Flocculation

~	Does a lack of flocculation time, inadequate equipment, or lack of multiple

flocculation stages result in poor floe formation and degrade plant performance?

Does inadequate mixing in flocculation basinfail to suspect PAC?

>	A direct filtration plant, treating cold water and utilizing a flocculation basin
with short detention time and hydraulic mixing, does not create adequate
floe for filtration.

>	High PAC feed to flocculation results in PAC settling to bottom of basin or
mechanical failure.

9. Sedimentation

~ Does the sedimentation basin configuration or equipment cause inadequate

solids removal that negatively impacts filter performance?

>	The inlet and outlet configurations of the sedimentation basins cause short-
circuiting, resulting in poor settling and floe carryover to the filters.

>	The outlet configuration causes floe break-up, resulting in poor filter
performance

>	The surface area of the available sedimentation basins is inadequate,
resulting in solids loss and inability to meet optimized performance criteria
for the process.

> Inability to frequency dean sedimentation basins during a HAB event.

98

-------
> Lack of sedimentation process limits ability to treat water during a HAB event.

10.	Filtration

~	Do filter or filter media characteristics limit the filtration process performance?

>	The filter loading rate in a plant is excessive, resulting in poor filter
performance.

>	Either the filter underdrain or support gravel have been damaged to the
extent that filter performance is impacted.

~	Do filter rate-of-f low control valves provide a consistent, controlled filtration rate?

>	The rate-of-f low control valves produce erratic, inconsistent flow rates that
result in turbidity and/or particle spikes.

~	Do inadequate surface wash or backwash facilities limit the ability to clean the
filter beds?

>	The backwash pumps for a filtration system do not have sufficient capacity
to adequately dean the filters during backwash.

>	The surface wash units are inadequate to properly dean the filter media.

>	Backwash rate is not sufficient to provide proper bed expansion to properly
clean the filters.

11.	Disinfection

~	Do the disinfection facilities have limitations, such as inadequate detention time,
improper mixing, feed rates, proportional feeds, or baffling, that contribute to
poor disinfection?

>	An un baffled clearwell does not pro vide the necessary detention time to meet
the Giardia inactivation requirements of the SWTR.

>	Plant has inability to treat HAB toxins through oxidation during the
disinfection process (e.g., use of chloramines).

12.	Sludge/Backwash Water Treatment and Disposal

~	Do inadequate sludge or backwash water treatment facilities negatively influence
plant performance?

>	The plant is recycling backwash decant water without adequate treatment or
during an HAB event.

>	The plant is recycling backwash water intermittently with high volume
pumps.

>	The effluent discharged from a sludge/backwash water storage lagoon does
not meet applicable receiving stream permits.

>	Inadequate sludge disposal exists at a plant, resulting in reduced cleaning of
settling basins and recycle of solids back to the plant.

3. Plant Operability

1. Process Flexibility

~	Does the lack of flexibility to feed chemicals at desired process locations or the
lack of flexibility to operate equipment or processes in an optimized mode limit
the plant's ability to achieve desired performance goals?

99

-------
>	A plant does not ha ve the flexibility to feed either a flocculant aid to enhance
floe development and strength or a filter aid to improve filter performance.

>	A plant includes two sedimentation basins that can only be operated in
series.

>	Plant has inability to feed PAC at location not impacted by oxidant(s).

>	Plant does not have the ability to bypass treated water during plant upsets.

2.	Process Controllability

~	Do existing process controls or lack of specific controls limit the adjustment and
control of a process over the desired operating range?

>	Filter backwash control does not allow for the ramping up and down of the
flow rate during a backwash event.

>	During a filter backwash, the lack of flow control through the plant causes
hydraulic surging through the operating filters.

>	The level control system located in a filter influent channel causes the filter
effluent control valves to overcompensate during flow rate changes in a
plant.

>	Flows between parallel treatment units are not equal and cannot be
controlled.

>	The plant influent pumps cannot be easily controlled or adjusted,
necessitating automatic start-up/shutdown of raw water pumps.

>	Plant flow rate measurement is not adequate to allow accurate control of
chemical feed rates.

>	Chemical feed rates are not easily changed or are not automatically changed
to account for changes in plant flow rate.

3.	Process Instrumentation/Automation

~	Does the lack of process instrumentation or automation cause excessive operator
time for process control and monitoring?

>	A plant does not have continuous recording turbidimeters on each filter,
resulting in extensive operator time for sampling.

>	The indication of plant flow rate is only located in the pipe gallery, which
causes difficulty in coordinating plant operation and control.

>	Automatic shutdown/start-up of the plant results in poor unit process
performance.

4.	Standby Units for Kev Equipment

~	Does the lack of standby units for key equipment cause degraded process
performance during breakdown or during necessary preventive maintenance
activities?

>	Only one backwash pump is available to pump water to a backwash supply
tank, and the combination of limited supply tank volume and an unreliable

100

-------
pump has caused staff to limit backwashing of filters during peak production
periods.

5.	Flow Proportioning

~	Does inadequate flow splitting to parallel process units cause individual unit
overloads that degrade process performance?

>	Influent flo w to a plant is hydraulically split to multiple treatment trains, and
uneven flow distribution causes overloading of one
flocculation/sedimentation train over the others.

6.	Alarm Systems

~	Does the absence or inadequacy of an alarm system for critical equipment or
processes cause degraded process performance?

>	A plant that is not staffed full-time does not have alarm and plant shut-down
capability for critical finished water quality parameters (i.e., turbidity,
chlorine residual).

7.	Alternate Power Source

~	Does the absence of an alternate power source cause reliability problems leading
to degraded plant performance?

>	A plant has frequent po wer outages, and resulting plant shutdo wns and start-
ups cause turbidity spikes in the filtered water.

8.	Laboratory Space and Equipment

~	Does the absence of an adequately equipped laboratory limit plant performance?

>	A plant does not have an adequate process control laboratory for operators
to perform key tests (i.e., turbidity, jar testing).

9.	Sample Taps

~	Does the lack of sampletaps on process flow streams prevent needed information
from being obtained to optimize performance?

>	Filter-to- waste piping folio wing plant filters does not include sample taps to
measure the turbidity spike following backwash.

>	Sludge sample taps are not available on sedimentation basins to allow
process control of the sludge draw-off from these units.

3. Operation

1. Testing

1. Process Control Testing

~ Does the absence or wrong type of process control testing cause improper
operational control decisions to be made?

>	Plant staff do not measure and record raw water pH, alkalinity, and turbidity
on a routine basis; consequently, the impact of raw water quality on plant
performance cannot be assessed.

>	Sedimentation basin effluent turbidity is not measured routinely in a plant.

101

-------
>	Plant staff do not measure toxins or surrogates (indicators) for cyanotoxin
removal (e.g., to be determined; EL/SA, NOM, phycocyanin, chlorophyll-a,
streaming current, particle count, turbidity).

2. Representative Sampling

~	Do monitoring results inaccurately represent plant performance or are samples
collected improperly?

>	Plant staff do not record the maximum turbidity spikes that occur during filter
operation and following filter backwash events.

>	Turbidity sampling is not performed during periods when the reclaim
backwash water pump is in operation.

>	Source water sampling does not accurately represent water quality (e.g.,
sampling reservoir to characterize water quality at various depths).

2. Process Control

1.	Time on the Job

~	Does staff's short time on the job and associated unfamiliarity with process
control and plant needs result in inadeguate or improper control adjustments?

>	Utility staff, unfamiliar with surface water treatment, were given
responsibility to start a new plant; and lack of experience and training
contributed to improper coagulation control and poor performance.

2.	Water Treatment Understanding

~	Does the operator's lack of basic water treatment understanding contribute to
improper operational decisions and poor plant performance or reliability?

>	Plant staff do not have sufficient understanding of water treatment processes
to make proper equipment or process adjustments.

>	Plant staff have limited exposure to water treatment terminology, limiting
their ability to interpret information presented in training events or in
published information.

>	Plant staff feed PAC at same location or dose to oxidant feed in process.

>	Plant staff feed algaecide to a reservoir indiscriminately or feed pre-oxidants
in the plant resulting in the possibility of cell lysis during HAB events.

>	Plant staff recycle backwash/sludge decant water to plant during HAB event.

>	Plant staff do not consider sedimentation sludge age and the potential for
toxin release during a HAB event.

3.	Application of Concepts and Testing to Process Control

~	Is the staff deficient in the application of their knowledge of water treatment and
interpretation of process control testing such that improper process control
adjustments are made?

>	Plant staff do not perform jar testing to determine appropriate coagulant
dosages for different water quality conditions.

>	Plant staff do not perform studies to determine most effective PA C type, dose,
and mixing energy to treat for HABs.

102

-------
>	Dedicated studies are not conducted to evaluate treatment options to
optimize plant performance and consider simultaneous treatment objectives.

>	Plant filters are placed back in service following backwash without
consideration for effluent turbidity levels.

>	Filter to waste valves are available but are not used following filter backwash.

>	Plant staff do not calculate chemical dosages on a routine basis.

>	Plant staff do not change chemical feed systems to respond to changes in
raw water quality.

>	Filters are backwashed based on time in service or headloss rather than on
optimized performance goal for turbidity or particle removal.

>	Sedimentation basin performance is controlled by visual observation rather
than process control testing.

3. Operational Resources

1.	Training Program

~	Does inadequate training result in improper process control decisions by plant
staff?

>	A training program does not exist for new operators at a plant, resulting in
inconsistent operator capabilities.

2.	Technical Guidance

~	Does inappropriate information received from a technical resource (e.g., design
engineer, equipment representative, regulator, peer) cause improper decisions or
priorities to be implemented?

>	A technical resource occasionally provides recommendations to the plant
staff; however, recommendations are not based on plant-specific studies.

3.	Operational Guidelines/Procedures

~	Does the lack of plant-specific operating guidelines and procedures result in
inconsistent operational decisions that impact performance?

>	The lack of operational procedures has caused inconsistent sampling
between operator shifts and has led to improper data interpretation and
process control adjustments.

4. Maintenance

1. Maintenance Program
1. Preventive

~ Does the absence or lack of an effective preventive maintenance program cause
unnecessary equipment failures or excessive downtime that results in plant
performance or reliability problems?

> Preventive maintenance is not performed on plant equipment as
recommended by the manufacturer, resulting in premature equipment
failures and degraded plant performance.

103

-------
> A work order system does not exist to identify and correct equipment that is
functioning improperly.

2.	Corrective

~	Does the lack of corrective maintenance procedures affect the completion of
emergency equipment maintenance?

>	A priority system does not exist on completion of corrective maintenance
activities, resulting in a critical sedimentation basin being out of service for
an extended period.

>	Inadequate critical spare parts are available at the plant, resulting in
equipment downtime (e.g., critical parts are not available for mixing and
sludge collection equipment during PAC feed season).

3.	Housekeeping

~	Does a lack of good housekeeping procedures detract from the professional
i mage of the water treatment plant?

>	An unkempt, cluttered working environment in a plant does not support the
overall good performance of the facility.

2. Maintenance Resources

1.	Materials and Equipment

~	Does the lack of necessary materials and tools delay the response time to correct
plant equipment problems?

>	Inadequate tool resources at a plant results in increased delays in repairing
equipment.

2.	Skills or Contract Services

~	Do plant maintenance staff have inadequate skills to correct equipment problems
or do the maintenance staff have li mited access to contract mai ntenance services?

>	Plant maintenance staff do not have instrumentation and control skills or
access to contract services for these ski lis, resulting in the inability to correct
malfunctioning filter rate control valves.

104

-------
Appendix C
Exit Meeting and Final Report

Contents

"Exit Meeting Agenda" Template	2

"Exit Meeting Presenters' Agenda" Template	3

"Possible Further Studiesfor Plant Staff to Conduct to Support Plant Optimization" Template	5

Example "Why Optimize?" Exit Meeting Handout	6

Example HAB CPE Reports	8

Office of Water (MS-140)

EPA 815-B-22-005

June 2022

-------
"Exit Meeting Agenda" Template

City of XXXX Water Treatment Plant
City, State
Date

Optimization Overview: Why Optimize

Performance Assessment - Historical Data

•	Historical turbidity data

•	Historical backwash data

On-Site Studies (as applicable)

•	Sedimentation Basin Backup

•	Filter Assessment Study

•	Filter backwash study

•	Turbidity data integrity assessment

•	PAC Jar Test

•	Source Water Sampling

Major Unit Process Evaluation/Summary

•	Microbial

•	HABs

Path to Optimization: Factors Limiting Performance
Potential follow-up studies
Wrap up

2

-------
"Exit Meeting Presenters' Agenda" Template

City of XXXX Water Treatment Plant
City, State
Date

Assign a moderator for the meeting, who will introduce each speaker, point out take-home
messages and draw connections between each topic. Use of a computer/projector is optional,
but may enhance the data-based discussions (e.g., historical performance data, study data,
etc.).

Optimization Overview: Why Optimize

Set the stage by discussing the optimization goals and the multiple barrier approach. High
level reiteration of key points about optimization and "why optimize"and remind attendees
of the handout from the entrance meeting which provides more information.

Performance Assessment - Historical Data

•	Historical turbidity data

Present the optimization assessment software summary of the raw, settled and finished
water turbidity data. Emphasize raw/settled/finished water trends (i.e., spikes in the raw
water passing through to settled and finished water, performance relative to the
optimization goals).

•	FTW time analysis

Discuss historical filter-to-waste data compared to the optimization goals. The data
indicated the filter-to-waste periods are exceeding the recommended period of 30 minutes.

On-site studies

Discuss any planned on-site studies, relating each to the goals and historical performance data
findings.

•	Study #1

•	Study #2

•	Study #3

•	Etc....

Major Unit Process Evaluation/Summary

3

-------
Three MUPs were developed - one for microbial/turbidity performance, one for HAB adsorption
and one for HAB oxidation. The MUP assessment intends to determine whether the system has
the "concrete and steel"in place to meet the optimization goals. Tie the discussion back to the
goals by assessing if the major unit processes and HAB control process are capable of meeting
the optimization goals.

Explain the assumptions used in the evaluation.

Factors Limiting Performance
Potential follow-up studies
Wrap up

• Present a summary of the evaluation and describe follow-up activities that potentially
exist. This will likely be the responsibility of the host state to make this presentation.

4

-------
"Possible Further Studies for Plant Staff to Conduct to Support Plant

Optimization" Template

City of XXXX WTP HAB CPE

Study #1 - Title

•	Description

o

•	Benefits

o

Challenges

Solutions



•
•

Study #2 - Title



•	Description

o

•	Benefits

o



Challenges

Solutions





Study #3 - Title



•	Description

o

•	Benefits

o



Challenges

Solutions





5

-------
Example "Why Optimize?" Exit Meeting Handout

WHY OPTIMIZE?

¦	Drinking water research indicates that achieving optimized performance goals provides
increased public health protection.

¦	Field work demonstrates that optimization goals are achievable at most plants without
majorcapital expenditures

¦	Optimization is a promising approach for control ling the impacts of HABs (i.e., reducing
cyanobacteria and related cyanotoxins)

OPTIMIZED PERFORMANCE GOALS

Minimum Data Monitoring

¦	Daily raw waterturbidity

¦	Settled waterturbidityat4-hourtime increments from each sedimentation basin

¦	On-line (continuous) turbidity from each filter

¦	One filter backwash profile each month from each filter

Individual Sedimentation Basin Performance Criteria

¦	Settled waterturbidity lessthan 1 NTU 95 percentof the time based on daily maximum
values when annual average raw waterturbidity is lessthan or equal to 10 NTU

¦	Settled waterturbidity lessthan 2 NTU 95 percentof the time based on daily maximum
values when annual average raw waterturbidity isgreaterthan 10 NTU

Individual Filter Performance Criteria

¦	Filtered waterturbidity lessthan 0.10 NTU 95 percent of the time (excluding 15-minute
period following backwashes) based on the maximum dailvvalues

¦	Maximum filtered water measurement of 0.3 NTU

¦	Initiate filter backwash immediately afterturbidity breakthrough has been observed and
before effluentturbidityexceedsO.10 NTU.

¦	Post backwash individual filter effluent turbid itv for filters with filter-to-waste capability:
Minimize individual filtereffluentturbidityduringfilter-to-waste period and record
maximum value. Return the filterto service at < 0.10 NTU.

6

-------
Post backwash individual filter effluent turb id ity for filters without filter-to-waste
capability: Maximum i nd ivid ua I filter effl uent tu rbid ity following backwash <0.30 NTU
and achieve < 0.10 NTU within 15 minutes.

Disinfection Performance Criteria

¦	CT values to achieve required log inactivationof Giardia and virus

OPTIMIZATION UTILIZES THE MULTIPLE BARRIER STRATEGY TO ENHANCE FINISHED WATER
QUALITY:

¦	Key treatment barriers include flocculation/sedimentation, filtration, and disinfection.
Each barrier is important when strivingforoptimized performance

¦	Performance of each barrier can often be assessed using surrogates, such as turbidity;
disinfection effectiveness can be measured directly. Toxin sampling/measurement is
needed on some basis to assess impact.

¦	Treatment objective is to produce a consistent, high quality finished water.

Variable
Quality
Source
Water

Flocculation/Sedimentation
Barrier

Finished
Water

Filtration
Barrier

Disinfection
Barrier

7

-------
Example HAB CPE Reports

8

-------
Results of the
Harmful Algal Bloom
Comprehensive Performance Evaluation
for the
ABC Treatment Plant
Anytown, State

August 1-5, 2016

Prepared By:

Process Applications, Inc.
2627 Redwing Road, Suite 340
Fort Collins, Colorado 80526

USEPA Technical Support Center
26 West Martin Luther King Drive
Cincinnati, Ohio 45268

State Environmental Protection Agency

9

-------
Table of Contents

SITE VISIT INFORMATION	5

INTRODUCTION	14

DESCRIPTION OF WATER TREATMENT PLANT	15

Overview	15

Source Intake and Pump Station	16

Water Treatment Processes	16

PERFORMANCE ASSESSMENT	19

Historical Performance Assessment	19

Administration Assessment	19

Historical Water Quality Performance Assessment	20

Historical Performance Summary	31

Disinfection	32

Cvanotoxins	33

MAJOR UNIT PROCESS EVALUATION	53

Particle Removal and Microbial Disinfection	54

Cvanotoxin Removal and Destruction Treatment	56

PERFORMANCE-LIMITING FACTORS	60

Policies - Administration (A)	61

Application of Concepts and Testing to Process Control - Operations (A)	61

Process Instrumentation/Automation - Design (A)	62

Reliability - Administration/Design (B*)	62

Process Control Testing - Operations (B*)	62

EVALUATION FOLLOW-UP	62

APPENDIX A - Major Unit Process Evaluation Supporting Calculations	60

10

-------
List of Figures

FIGURE 1. Schematic of the ABC Water Treatment Plant 17

FIGURE 2. ABC WTP Turbidity Profile 24

FIGURE 3. Maximum Daily Individual Clarifier Effluent Turbidity 25

FIGURE 4. Daily Individual Clarifier Effluent Turbidity from 12:00 P.M. Samples 27

FIGURE 5. Maximum Daily Filtered Water Turbidity (IFE and CFE) 28

FIGURE 6. Daily Disinfection Inactivation Ratio 24

FIGURE 7. Excavated Area of Filter No. 4 Showing Intermixing of Media 35

FIGURE 8. Filter No. 4 Filter Probing Map 37

FIGURE 9. Filter No. 4 Waste Backwash Water Turbidity Profile 39

FIGURE 10. Filter-to-Waste Profile for Inspected Filter No. 4 40

FIGURE 11. Filter-to-Waste Profile for Filter No. 8 Backwash 41

FIGURE 12. Settling Curves for Water Treated With and Without NaMn04 44

FIGURE 13. Impact of NaMnOi on Cvanotoxin Release and Extracellular Microcystins

Concentration 45

FIGURE 14. Plant Profile for 12:00 Hour Sampling Period on August 3.2016 49

FIGURE 15. Plant Profile for 16:00 Hour Sampling Period on August 3.2016 50

FIGURE 16. Plant Process Percent Removals of Total and Extracellular Microcystins at the

12:00 Hour Sampling Time 50

FIGURE 17. Plant Process Percent Removals of Total and Extracellular Microcystins at the

16:00 Hour Sampling Time 51

FIGURE 18. Total Microcystins and Phvcocvanin Correlation 52

FIGURE 19. Total Microcystins and Chlorophvll-a Correlation 52

FIGURE 20. Major Unit Process Evaluation Graph - Particle Removal and Microbial Disinfection 55

FIGURE 21. Cvanotoxin Treatment Major Unit Process Evaluation Graph 57

FIGURE 22. Predicted PAC Dose Based on Removal Efficiency and

Initial Microcvstin Concentration 66

FIGURE 23. Oxidation Capacity Based on 97 Percent Removal Using AWWA Calculator 80

-------
List of Tables

TABLE 1. CPE Turbidity Performance Analysis; Data Acquisition Description	21

TABLE 2. OAS Summary Statistics	23

TABLE 3. OAS Optimization Trend - Settled Water	26

TABLE 4. OAS Clarifier Effluent Statistics. Daily Max vs. Noon Sample Values	27

TABLE 5. OAS IFE Filter No. 4 Statistics. SCADA vs SC200 Controller	29

TABLE 6. OAS Optimization Trend - Filtered Water	30

TABLE 7. CFE Data Removal of Errant Spikes	31

TABLE 8. ABC WTP Performance Summary	31

TABLE 9. Plant Profile Sampling Locations	47

TABLE 10. Major Unit Process Summary	60

12

-------
SITE VISIT INFORMATION

Site and Mailing Address:

Removed

Date of Site Visit:

August 1 - 5, 2016

ABC Water Treatment Plant Personnel Participating:

Commissioner

Sanitary Engineer

Administrator

Superintendent

Assistant Superintendent

HR Department

Operator

Operator

Operator

Operator

CPE Team:

USEPA Technical Support Center, 26 West Martin Luther King Drive, Cincinnati, Ohio
45268

AlisonDugan - 513-569-7122;DuganAlison@epa.gov
RickLieberman - 513-569-7604; Lieberman.Richard@epa.gov
Tom Waters-513-569-7611; Waters.Tom@epa.gov

USEPA Office of Research and Development, 26 West Martin Luther King Drive, Cincinnati,
Ohio 45268

Craig Patterson- 513-487-2805; Patterson.Craig@epa.gov

Process Applications, Inc., 2627 Redwing Road, Suite 340, Fort Collins, Colorado 80526
Bill Davis - 469-338-1823; waterbilldavis@gmail.com
Larry DeMers -970-223-5787; ldemersco@aol.com

State Environmental Protection Agency

HAB Engineer

HAB Coordinator

Design Engineer

Field Engineers/Staff/Inspectors

13

-------
INTRODUCTION

The Composite Correction Program (CCP)4 is an approach developed by the U. S. Environmental
Protection Agency (USEPA) and Process Applications, Inc. (PAI) to improve surface water
treatment plant performance and to achieve compliance with the Surface Water Treatment Rule
(SWTR). Its development was initiatedby PAI and the state of Montana5, which identified the
need for a program to address performance problems at its surface water treatment plants. The
approach consists of two components, the Comprehensive Performance Evaluation (CPE) and
the Comprehensive Technical Assistance (CTA).

A CPE is a thorough evaluation of an existing treatment plant, resulting in a comprehensive
assessment of the unit process capabilities and the impact of the operation, maintenance, and
administrative practices on performance of the plant. The results of the evaluation establish the
plant capability to meet the optimization goals and list a set of prioritized factors limiting perfor-
mance. A CTA is used to improve performance of an existing plant by systematically addressing
the factors limiting performance identified during the CPE.

The implementation of the Interim Enhanced Surface Water Treatment Rule (IESWTR), promul-
gated in December 1998, required plants that serve greater than 10,000 customers to achieve less
than 0.3 NTU (nephelometric turbidity units) turbidity in 95 percent of the monthly combined
filter effluent samples and to monitor individual filter performance. The requirement went into
effect for all surface water treatment plants in 2005. Research results and field experience have
shown that just meeting the requirements of the IESWTR does not guarantee adequate protection
against some pathogenic microorganisms, as evidenced by some waterborne disease outbreaks.

Producing a finished water with a turbidity of less than or equal to 0.10 NTU provides much
greater protection against pathogens like Cryptosporidium3. This microorganism that passed

4 Hegg, B.A., L.D. DeMers, J.H. Bender, E.M. Bissonette, and R.J. Lieberman, Handbook -Optimizing Water
Treatment Plant Performance Usingthe Composite Correction Program, EPA 625/6-91/027, USEPA,
Washington, D.C. (August 1998).

5 Renner, R.C., B.A. Hegg, and D.F. Fraser, Demonstration of the Comprehensive Performance Evaluation
Technique to Assess Montana Surface Water Treatment Plants, Association of State Drinking Water
Administration Conference, Tucson, AZ (February 1989).

3 Patania et al., Optimization of Filtration for Cyst Removal. American Waterworks Association Research
Foundation. Denver, CO. 1995.

-------
through the public water supply was responsible for a large outbreak of Cryptosporidiosis in
Milwaukee, Wisconsin in April 1993, where 400,000 people became ill, and nearly 100 deaths
occurred. Cryptosporidium cysts are extremely resistantto chlorine disinfection, necessitating
optimization of physical removal of particles.

Since the development of the CCP for optimization of surface water treatment plants for protec-
tion from microbial pathogens, PAI and the USEPA's Technical Support Center (TSC) have
adapted the CCP protocol to the additional public health parameters such as DBP control and
distribution system optimization. Given the recent concerns with harmful algal blooms (HABs)
and their impact on surface water treatment plants in this state and nationwide, the State EPA, in
partnership with TSC, has initiated a project to expand the CCP to include optimization for the
removal of blue-green algae (cyanobacteria) cells and the reduction of associated toxins. This
CPE for the ABC Water Treatment Plant (WTP) represents the first of four developmental CPEs
that will be conducted in the state focused on these performance goals.

The following report presents all of the findings from this CPE and will hopefully provide ABC
Water with valuable information that can be used to enhance and maintain water quality. The
CPE team would like to thank the plant staff and utility management for hosting this event and
for taking the time to assist the team in completingthe evaluation. During the evaluation, utility
staff members were very accommodating in providing plant information and sharing their
experience and knowledge regarding treatment approaches to address HAB events. This type of
attitude represents a strong foundation for development of an optimization approach to address
HAB events that public water systems may face in the future. This report documents the
findings of the CPE conducted at the ABC WTP on August 1-5, 2016.

DESCRIPTION OF WATER TREATMENT PLANT
Overview

The ABC WTP is the main source of potable water for the unincorporated eastern portion of
ABC as well as a nearby city and village. Additionally, ABC Water has interconnections with a
nearby village and another Water and Sewer Authority to provide purchased potable water on an
emergency basis. Potable water is delivered to approximately 17,000 direct consumers and a

15

-------
total of 28,000 consumers, including purchased water in the neighboring communities. ABC
Water operates and maintains the system.

Source Intake and Pump Station

A schematic of the water treatment plant, provided by the utility, is shown in Figure 1. The
source water is supplied to the plant from an intake on a nearby lake, with an alternate supply
available on a nearby river. The Lake intake structure is located approximately 1,500 feet off-
shore at a depth of about eight feet, and the intake pumping station building is located along the
southern bank of the lake. Approximately 2,000 feet of raw water line connects the Lake intake
to the raw water pump station. An additional raw water line from an intake on the River is
located northeast of the raw water pump station and 55 feet into the river. This intake ties into
the raw water line from the Lake intake with a length of approximately 200 feet of raw water
pipe, and it is operating by way of a gate valve. The River is typically only used in cold weather
when frazil ice is a problem in the lake. The water quality from the two sources tends to trend
together. Three raw water pumps within the pump station transport the raw water to the water
treatment plant. The plant has the ability to feed sodium permanganate at the intake pump
station with a chemical feed point located on the pump discharge line. Sodium permanganate
was being fed during the CPE.

A raw water sample line is located in the pump station, and it collects raw water from the wet
well prior to permanganate addition. Monitoring instrumentation includes a turbidimeter and pH
meter. Also located within the wet well is a data sonde, which captures and logs continuous data
including: turbidity, phycocyanin, chlorophyll, pH, and dissolved oxygen. These data are used
by the operators to monitor for HABs and to adjust treatment during these events.

Water Treatment Processes

The ABC WTP utilizes conventional surface water treatment processes, including: coagulation,
flocculation, sedimentation, filtration, and disinfection. The reported plant capacity is 9.0 MGD.
A pretreatment step precedes the conventional plant and includes four pretreatment basins, each
equipped with two top-mounted axial flow impeller mixers, to allow for the addition of
powdered activated carbon (PAC) and a secondary permanganate feed point, if necessary. At the
time of the CPE, no additional chemical or PAC was being added to the pretreatment basins.

-------
FIGURE 1. Schematic Removed.

17

-------
The plant staff have experienced significant plugging problems with the PAC feed system to the
pretreatment basins, and, as a result, have stopping using this feed location. These basins do not
have sludge removal but are drained and cleaned out in the fall.

From the pretreatment basins, the water travels through a flume to the two rapid mix units. At
the time of the CPE, aluminum chlorohydrate (ACH) was being added to the rapid mix units for
coagulation along with PAC for taste and odor and cyanotoxin control. Additional chemicals,
including caustic soda, permanganate, and polymer, can also be fed at this location.

Flocculation and clarification are accomplished with three solids contact clarifiers. Each unit has
an inner flocculating zone, an outer settling zone, and an effluent collection system. The solids
contact clarifiers are operated in parallel. Each clarifier is equipped with a turbidimeter to meas-
ure settled water turbidity, and an additional turbidimeter measures the combined settled water
turbidity from a common outlet flume. An online pH meter also measures the pH of the com-
bined settled water.

From the solids contact clarifiers, settled water travels to the filter building through a common
flume, where flow is divided among two trains of four-cell cluster filters. Filtration in the plant
is achieved through eight filter cells equipped with dual-media anthracite and sand. Each filter
effluent is sampled. The samples are transferred to turbidimeters and particle counters located
on the operating floor, using high suction lift sample pumps. Backwash supply is provided by
the filters in service and supplemented by the high service pump discharge. Air scour is also
provided as part of the filter backwash procedure. The filters have the ability to function in
filter-to-waste mode following a backwash or during filter startup.

Filtered water flows to a common transfer wet well, where the combined filter effluent turbidity
is sampled and directed to a continuous turbidimeter. Sodium hypochlorite is injected into the
wet well before filtered water is pumped to the clearwells. Three vertical turbine transfer pumps
pump water from the transfer wet well to two ground level clearwells. Located on the discharge
line of the transfer pumps are inj ection points for the addition of caustic soda, fluoride, and a
poly/orthophosphate blend corrosion inhibitor. Each clearwell holds a volume of 625,000 gal-
lons of water, and both are constructed of concrete with fiberglass domes. The clearwells are
baffled, operate in parallel, and are utilized for disinfection contact time. Treated water is

-------
pumped to the distribution system from a 30-inch suction line from the clearwellsby way of
three horizontal, centrifugal high service pumps. An additional feed point for chlorine also exists
on the high-pressure discharge line manifoldto boost levels after the clearwells, if needed. A
sampling location after the high service pumps is used to measurepH and chlorine residual of the
finished water as well as to take other compliance samples.

Waste filter backwash water is collected in a backwash holding tank and is pumped to sludge
lagoons. Sludge from the solids contact clarifiers is also sent to the two sludge lagoons. A
NPDES permit allows decant from the lagoons to be discharged to a receiving stream.

PERFORMANCE ASSESSMENT
Historical Performance Assessment

Optimized performance, for the purposes of this CPE report, represents performance beyond the
Surface Water Treatment Rule (SWTR) requirements. To achieve optimized performance, a
water treatment pi ant must demonstrate that it can take a raw water source of variable quality and
produce consistent, high quality finished water. In addition, the performance of each treatment
unit process must demonstrate its capability to act as a barrier against the passage of particles at
all times.

Administration Assessment

An assessment of the administration of the ABC WTP and its possible effect on plant
performance was performed by collecting information through interviews in the following
general areas: utility structure, vision, mission, water quality goals, reporting, data review,
management style, communications, planning, plant coverage, financial management, and spend-
ing authority. Two possible administrative issues were identified that could potentially affect
performance. These issues, as well as others, are considered in subsequent sections of the report:

• Individual filter effluent turbidity data review and reporting, and

• Formal adoption of optimization turbidity goals for unit process performance.

-------
Historical Water Quality Performance Assessment

Turbidity and Disinfection- Historical turbidity data were collected from three sources during
this CPE. Monthly ""Sanitary Engineers Reports", "which are in spreadsheet format and
generated by the ABC Regional WTP SCADA system, served as one source of historical data.
These reports contain data that are collected from online instrumentation and from laboratory
bench analyses entered and stored in the water treatment plant SCADA system. Data from these
monthly reports were provided to the CPE team in electronic format, which allowed for direct
copying and pasting of data into an Optimization Assessment Spreadsheet (OAS) that is used to
assess performance against the optimization goals for turbidity.

Operators' daily log sheets, in hard copy format, were a second source of historical data that
were provided to the CPE team onsite in the form of paper copies. These logs were especially
useful in assessing the performance of the three up-flow clarifiers. Members of the CPE team
determined the maximum daily turbidity values from each individual clarifier and entered these
values into the OAS.

A third source of historical turbidity data used by the CPE team was a Hach SC200 Controller,
which stores and transmits data from the Hach 1720E Turbidimeters, measuring individual filter
effluent (IFE) turbidity values for filter No. 2 and No. 4. These data were downloaded from the
controller and used to compare the IFE values from those two filters against the same IFE values
obtained from the " "Sanitary Engineers Reports" " generated by the SCADA system. This was
done to check on the integrity of the data that were generated by the turbidimeters and then
transmitted electronically to the SCADA system, which stores and generates reports from the
data. Filter No. 4 was selected for this analysis because the data can be accessed from the SC200
Controller. Turbidity data from the other filters were stored and transmitted via SCI00
Controllers, and these data were not accessible by the CPE team.

Historical performance was generally assessed over a 12-month time period, starting on
August 1, 2015 and ending on July 31, 2016. Table 1 describes in more detail the exact source
of the data used in the CPE performance assessment.

-------
TABLE 1. CPE Turbidity Performance Analysis; Data Acquisition Description

Performance Parameter

Data Used in the CPE Performance Analysis

Maximum Daily Raw WaterTur-
bidity Entering the Plant

Maximum daily raw water turbidity data were obtained from monthly
"Sanitary Engineers Reports" in the column labeled
Rpt_RawWaterTurbidity_Max. The values were located on the
spreadsheet tab named data in column EJ, starting in row9. These data
represent values obtained from an online turbidimeter located in the raw
water pump station. The sample tap is located in a raw water line in the
pump station, opposite of the NaMn04 injection point.

Individual Clarifier Effluent, or
"Settled W/afer"Turbidity

Operators' logs of two-hourturbidity test results were utilized to
determine the maximum turbidity value for each day, from each clarifier.
"Sanitary Engineers Reports" included combined clarifier effluent data
from online instrumentation and daily 12:00 P.M. grab samples of
individual clarifier effluent which were entered into the SCADA system.
These values were used for comparison, but the daily maximum individual
clarifier effluent values obtained from operators' logs were used to assess
performance against the optimization goals.

IFE Turbidity

The individual filter effluent (IFE) daily maximum turbidity records were
taken from monthly "Sanitary Engineers Reports" in each column labeled
Rpt_Filter_X_Turbidity_Max, with the "X" representing each of the eight
filter numbers. The values were located on the spreadsheet tab named
data in columns EO through EV, starting in row9.

The CPE team attempted to use the operators' logs to eliminate high tur-
bidity values associated with turbidimeter maintenance, calibration,
sample pump maintenance, filter, and backwashing. Theteam also
attempted to eliminate values associated with filter-to-waste cycles.
However, the operators' logs could not explain all of the irregularities,
and the CPE team could not access additional information from the
SCADA system. Therefore, all IFE data were used in the performance
assessment analysis, even though the CPE team does not believe the data
accurately represent true IFE quality.

CFE Turbidity

The combined filter effluent (CFE) daily maximum turbidity records were
obtained from monthly "Sanitary Engineers Reports"in the column
labeled Maximum Turbidity from the SI/1/77? MOR tab, column S,
beginning in row 18. These data are also available from the data tab in
the column labeled Rpt_TransferWellTurbidity_Max. The values were
located on the spreadsheet tab named data in column FU, starting in
row 9. However, the difference is that the SCADA system uses an
algorithm to eliminate what are considered errant CFE spikes from the
SWTR MOR data. In addition, operators' logs are used to verify CFE spikes
that may not be eliminated by the SCADA algorithm and could be
eliminated manually if there is justification from the operators' logs.
Therefore, the SWTR MOR data were used to populate the OAS rather
than the data from the TransferWellTurbidity column.

21

-------
Performance Parameter

Data Used in the CPE Performance Analysis

Disinfection

Monthly "Sanitary Engineers Reports" include spreadsheets generated by
the plant SCADA report function. Reports were provided covering the
period from January through December 2015.

22

-------
Historical disinfection data were made available to the CPE team in the monthly "Sanitary
Engineers Reports " generated through the plant SCADA report writer function in the form of an
Excel spreadsheet. Each report contains a tab that uses plant data to calculate the daily minimum
CT (disinfectant concentration times effective contact time) and to record the required CT
entered by the operator. The operators determine the required CT via interpolation from
published USEPA CT tables. The disinfection data were availablefrom January 2015 through
December 2015 for this analysis.

The turbidity data described in the table below were entered into an OAS, and these data were
analyzed through the spreadsheet calculations and charts, comparing values to optimization
goals. Figure 2 displays a turbidity profile which is a graphical description of water treatment
plant performance over the past year, and Table 2 shows the OAS summary statistics for the
plant.

The turbidity profile reveals general trends and also a sharp decline in raw water turbidity in
December 2015. There is a possibility of seasonal influences over clarifier performance, as indi-
cated by the black line in Figure 2, and IFE data reveal significant turbidity spikes which need to
be investigated, as indicated by the dashed blue line. There is no visual evidence of significant
pass-through of raw water spikes, and this observationis supportedby the RSQ values in Table 2
below.

TABLE 2. OAS Summary Statistics

ANNUAL DATA

Avg

Min

Max

RSQ

95%

Opt. Goal

Reg.

NTU

% Values

Max. Raw Turbidity

51.1

0.6

500.0

n/a

177.0

n/a

Max. Clarifier Effluent Turbidity

1.6

0.4

4.5

0.00

3.2

n/a

Max. Filtered Turbidity

0.25

0.05

5.00

0.00

0.52

n/a

Combined Filtered Turbidity

0.09

0.03

5.00

0.00

0.15

100

RSQ = Correlation Coefficient for two selected data sets (> ~ 0.25 suggests correlation)

95% = 95th Percentile value for data set

Opt. Goal = % of values in data set that are less than or equal to the selected optimization turbidity goal
Reg. = % of values in data set that are less than or equal to the regulated turbidity requirement

-------
Turbidity Profile

	Raw 	Max Sed 	Max Filter 	Combined

0.01 -I	,	,	,	,	,	,	,	,	,	,	,—

Aug-15 Sep-15 Oct-15 Nov-15 Dec-15 Jan-16 Feb-16 Mar-16 Apr-16 May-16 Jun-16 Jul-16

FIGURE 2. ABC WTP Turbidity Profile.

Individual clarifier performance appears to be better in the colder months of the year. Individual
filter effluent performance is very erratic, with significant spikes throughout the time period. As
described, the IFE data appears to be heavily influenced by issues impacting the accuracy of the
data. The combined filter effluent data show much better performance than the IFE data, and
they are generally below the optimization goal for filtration of 0.10 NTU, although still not
meeting the goal 95 percent of the time, as shown in Table 2.

The statistics in Table 2 are based on the maximum daily values for raw water, individual clari-
fier effluent, IFE, and CFE turbidity during the August 1, 2015 to July 31, 2016 time period.
These statistics are then compared to optimization goals. The optimization program utilizes the
"maximum " daily turbidity readings to assess worst-case performance by each of the barriers. If
the plant can perform within the optimization goals at the time of its worst daily performance,
then the plant staff can be assured that the plant is maximizing its ability to protect public health
against the passage of pathogens and cyanobacteria. Table 2 shows that the daily maximum raw

24

-------
water turbidity values average for the ABC WTP was 50.9 NTU. For raw water conditions such
as this, where the average maximum daily raw water turbidity is greater than 10 NTU, the
optimization goal for settled water turbidity is 2 NTU and the optimization goal for filtered water
turbidity is 0.10 NTU.

The maximum daily clarifier effluent turbidity, as measured with grab samples from the effluent
of each clarifier, met the optimization goal 76 percent of the time. The maximum clarifier efflu-
ent turbidity was 3.2 NTU or lower 95 percent of the days during the evaluation period. A closer
look at the settled water turbidity is shown in Figure 3. The red line in the graph represents
2.0 NTU, the optimization goal for settled water turbidity. It is also more apparent in Figure 3
that the clarifier performance appeared to be the worst in the late spring and summer, as com-
pared to the winter months. The OAS statistics from the ''Optimization Trend" tab (see Table 3)
revealed that clarifier No. 1 effluent turbidity performance was slightly worse than the other clar-
ifiersfrom September2015 to January 2016. Then, clarifier No. 3 had the highest effluent tur-
bidities from February to June 2016. However, the overall performance was fairly even across
the three clarifiers, with the 95th percentile values for the three clarifier effluents listed as 2.6,
2.7, and 2.8 NTU, respectively.

Maximum Daily Settled Water Turbidity

—Max Sed —Goal

FIGURE 3. Maximum Daily Individual Clarifier Effluent Turbidity.

-------
TABLE 3. OAS Optimization Trend-Settled Water



Settled Water Turbidity

95th Percentile Values (NTU)

% Values Meeting Goal

Sed 1

Sed 2

Sed 3

Sed 4

All Sed

3 NTU

2 NTU

1 NTU

Aug-15

2.95

2.95

3.10



3.04

94.62

66.67

14.0

Sep-15

2.89

2.28

2.18



2.50

96.67

88.89

32.2

Oct-15

2.37

2.20

1.50



2.20

98.86

93.18

46.6

Nov-15

1.26

1.00

0.88



1.18

100.00

100.00

93.0

Dec-15

0.92

0.69

0.71



0.81

100.00

100.00

98.9

Jan-16

1.80

1.60

1.40



1.60

100.00

100.00

55.9

Feb-16

1.06

1.16

1.16



1.17

100.00

98.85

88.5

Mar-16

1.50

1.85

1.90



1.88

100.00

97.85

41.9

Apr-16

2.97

3.11

4.08



3.56

90.00

67.78

33.3

May-16

2.40

2.85

3.35



2.94

95.70

78.49

19.4

Jun-16

3.36

3.11

3.37



3.36

91.11

55.56

6.7

Jul-16

2.50

2.20

2.45



2.44

100.00

82.80

18.3

Yr. 95%

2.60

2.70

2.80







Yr. Goal

86.5%

86.5%

84.3%



It is noteworthy that plant staff routinely enter the individual clarifier effluent turbidity values
from daily samples obtained at 12:00 noon. These are the values used for process control deci-
sions. The CPE team entered the noon values into the OAS to compare the performance of the
clarifiers when using these data versus the maximum daily values. When only using the noon
values, the clarifiers met the optimization goal for clarifier effluent 98 percent of the time, as
shown in the statistics of Table 4. Operator interviews revealed that clarifier performance may
degrade in the evening hours. Assessing values from samples obtained only at noon will not
account for particles passing the clarification barrier during worst-case scenarios. Figure 4 was
generated using the values collected at noon, and it reveals the same trend of seasonal perfor-
mance degradation, although much less pronounced, in the late spring and summer months.

26

-------
TABLE 4. OAS Clarifier Effluent Statistics, Daily Max vs. Noon Sample Values

Annual Data

Avg NTU

Min NTU

Max NTU

95% NTU

% Meeting Opt Goal

Max values

1.6

0.4

4.5

3.2

76

Noon values

0.9

0.2

3.4

1.8

98

Maximum Noon Value Daily Sed Turbidity

—Max Sed	—Goal

FIGURE 4. Daily Individual Clarifier Effluent Turbidity from 12:00 P.M. Samples.

For filtered water turbidity, the optimization goal is 0.10 NTU or less 95 percent of the time.
Table 2 shows that the maximum daily IFE turbidity values met the optimization goal 15 percent
of the days analyzed. The maximum IFE values were at 0.52 NTU or less during 95 percent of
the days analyzed. Table 2 also shows that the maximum daily CFE values met the optimization
goal 88 percent of the days analyzed. The maximum CFE values were at 0.16 NTU or less dur-
ing 95 percent of the days analyzed.

Figure 5 depicts the maximum daily filtered water turbidity for IFE and CFE turbidity measure-
ments in relation to the optimization goal of 0.10 NTU, represented by the red line. The graph
shows the maximum IFE turbidity measurements (dashed lines) mostly above the optimization

27

-------
goal throughout the last year, with significant spikes often exceeding 0.3 NTU. By contrast, the
maximum CFE turbidity (solid line) is generally below the optimization goal, although it does
exceed the 0.10 NTU goal on several occasions.

Information included in Table 2 above explains that the IFE data stored in the SCADA system
and captured in the monthly ''Sanitary Engineers Reports " are not scrubbed of errant spikes by
the SCADA system, as in the case of the CFE values. This was verified when some of the IFE
spikes in the "Sanitary Engineers Reports " could not be found on the SCADA screen for the
same time period. Therefore, the spikes were removed from the visual SCADA readout being
monitored by the operators but were captured in the reports.

Maximum Daily Filtered Water Turbidity

	Max Filter	Goal 	Combined

1.0
0.9
0.8
0.7

K 0.6
z

~ 0.5

¦o

In

3 0.4

0.3

0.1

1
1
1
•
•





1
1
1
1
1



Si

¦i
ii
•I

i' S



|

1

•
¦
¦
a





1
1
1
1
1

i

i

¦i
ii
ii
ii







¦
¦

•
N





1

1 i
1 i
1 i
1 I

i

1

i !



ll
l|
1



•

N
N

N

1



!

i

!¦ l j



Ss

ii



II
II
II
II
II

1



i i
•i 1

¦ 1

!|.

!

i

ii
i
¦
i

¦i



II
II

jl

!| !



a

1

s i

: ii ij

i( ii
'! ! !'



1

1
a

ii

is
ii

ii
ii



§

i! !



1 i

il iis: !:
'!i !!i : •'!

11! . H'! 8 i!

i! » ii

i	I ii

ii	a i|
¦ • a i
'!¦!>!



n

5

n

ii
i]

ii



si!

Ill ll

1 ' . >1

*•!¦ £ !' i

it 'V I • ,!



•S j SSi liiijii li'i

1 • ill i ill'u Si •

i> H ii i a
i> H n i. a
ji ii iS,il n
K; ii i i ku

j' (

" 1

ii l
¦l M ,

!' '1

i

i

ii

ii



vr

si

11 "u'vfli in rs*J\i 'j • i! >:' if i»i

Ji

• 4 «

« . 1

1 i

•ni? k Is i

i f!S ,.V1. j'.a*

¦J i, :

i!

i





m

0.0

Aug-15 Sep-15 Oct-15 Nov-15 Dec-15 Jan-16 Feb-16 Mar-16 Apr-16 May-16 Jun-16 Jul-16

FIGURE 5. Maximum Daily Filtered Water Turbidity (IFE and CFE).

28

-------
The CPE team performed an onsite study to compare the IFE data that are captured by the Hach
SC200 Controller (for filter No. 2 and No. 4) to the IFE data reported in the monthly "Sanitary
Engineers Reports. " The SC200 Controller has limited memory for data storage, but the team
was able to access IFE turbidity data for filter No. 4 over the time period of January 26 - July 31,
2016. The results can be found in Table 5. The data from the SC200 Controller represent values
obtained directly from the turbidimeter and would include all values, even during filter back-
washing, filtering-to-waste, or even during sample pump malfunctions, unless taken offline man-
ually. This study reveals that the data transmitted from the SC200 Controller are being altered in
some manner by the SCADA system, even if the SCADA algorithm to "clean up " the data has
not been applied in the way it has for the CFE data. The data collected by the SC200 Controller
revealed higher average and maximum values, which was expected since the SCADA algorithm
was likely designed to remove some of the spikes. However, the 95th percentile values are the
reverse of what was expected, further indicating that the SCADA influence on the recorded IFE
values should be investigated.

TABLE 5. OAS IFE Filter No. 4 Statistics, SCADA vs SC200 Controller

Annual Data

Avg NTU

Min NTU

Max NTU

95% NTU

% Meeting Opt Goal

SCADA values

0.13

0.03

5.00

0.27

SC200 values

0.63

0.03

50.77

0.18

The OAS also plots the performance of each filter based on IFE turbidity values and is one way
of checking to see if certain filters are performing better than others. These data are summarized
in Table 6, and they reveal that each filter profile is unique from the others, but that all eight
filters experience significant spikes and no individual filter appears to be performing better or
worse than the others.

-------
TABLE 6. OAS Optimization Trend-Filtered Water



Filtered Water Turbidity

95th Percentile Values (NTU)

%Values Meeting Goal All Filters

Filter 1

Filter 2

Filter 3

Filter 4

Filter 5

Filter 6

Filter 7

Filter 8

Combined

All Filters

0.3

0.2

0.1

Aug-15

0.27

0.32

0.26

0.31

0.23

0.24

0.40

0.34

0.24

0.30

95.16

63.31

19.4

Sep-15

0.18

0.22

0.13

0.19

0.16

0.17

0.15

0.21

0.14

0.19

100.00

96.67

69.6

Oct-15

0.14

0.16

0.13

0.10

0.17

0.17

0.13

0.15

0.07

0.15

99.19

98.79

83.9

Nov-15

0.12

0.18

0.12

0.12

0.11

0.18

0.10

0.18

0.08

0.16

100.00

98.75

86.3

Dec-15

0.18

0.19

0.14

0.15

0.13

0.21

0.16

0.16

0.09

0.17

99.19

98.39

81.5

Jan-16

0.25

0.72

0.34

0.67

0.17

0.21

0.19

0.21

0.12

0.25

95.97

87.90

72.6

Feb-16

0.15

0.22

0.22

0.20

0.20

0.22

0.22

0.43

0.09

0.22

97.84

91.38

78.4

Mar-16

0.15

0.29

0.17

0.35

0.17

0.22

0.25

0.22

0.09

0.23

97.98

90.73

78.6

Apr-16

0.19

0.15

0.21

0.19

0.14

0.21

0.23

0.25

0.13

0.20

99.58

95.00

82.5

May-16

0.10

0.10

0.09

0.09

0.10

0.11

0.10

0.12

0.07

0.11

100.00

100.00

92.7

Jun-16

0.21

0.10

0.10

0.12

0.10

0.12

0.10

0.13

0.09

0.13

98.33

96.67

88.8

Jul-16

0.16

0.14

0.20

0.62

0.08

0.15

0.10

0.12

0.20

0.16

97.98

95.97

85.9

Yr. 95%

0.21

0.24

0.23

0.24

0.19

0.21

0.21

0.22

0.15



Yr. Goal

78.1%

72.7%

77.3%

73.5%

80.6%

77.0%

80.6%

73.0%

88.0%

In contrast to the IFE data, spikes in the CFE data are removed from the SWTR MOR data set if
there are documented reasons to justify those actions. Table 7 contains some of the CFE values
which were removed from the past year's data set. Plant staff review all elevated CFE turbidity
values against operator log records and make decisions about the authenticity of the data. Errant
spikes are removed from the calculation of the CFE turbidity values reported on the SWTR
MOR if it is determined that the elevated values did not represent actual CFE water quality. Tur-
bidity spikes typically occur during maintenance of sample pumps and turbidimeters and plant
power outages.

30

-------
TABLE 7. CFE Data Removal of Errant Spikes

Date

SWTR MOR

"Sanitary Engineers Reports"

01/27/2016

0.06

0.64

03/22/2016

0.09

0.41

04/27/2016

0.04

0.61

05/18/2016

0.06

1.43

07/15/2016*

5.00

5.00

* The data for the July SWTR MOR had not yet been edited at the time of the CPE.

Historical Performance Summary

The ABC WTP performance is summarized in Table 8.

TABLE 8. ABC WTP Performance Summary

Barrier

Optimization Goal

Performance

Clarification

Settled water turbidity
1.0 NTU or less 95 per-
cent of the time, based
on daily maximum
values

The goal was assessed against individual clarifier effluent turbidity
values. This is the most effective way to assess the clarification bar-
rier. Plant staff are to be commended for sampling, analyzing, and
reporting individual clarifier effluent turbidities. The 95th percentile
of the maximum daily individual clarifier effluent turbidity was above
the goal, at 3.2 NTU, for the year analyzed. The plant met the 2 NTU
goal on 76 percent of the days during the year.

Filtration

IFE and CFE turbidities
0.10 NTU or less 95 per-
cent of the time, based
on daily maximum
values

The IFE data show performance meeting the optimization goal
15 percent of days analyzed during the year, with an annual 95th
percentile of 0.52 NTU. However, the authenticity of the data set is
in question, and the data must be "cleaned up" in a systematic, well-
documented process in order to make appropriate conclusions on IFE
performance.

The CFE data show performance meeting the optimization goal
88 percent of the days analyzed during the year, with an annual 95th
percentile of 0.15 NTU. The performance of the plant, based on CFE
data, fails to meet the filtered water optimization goal.

31

-------
Disinfection

Disinfection is the final barrier in the treatment plant for protection from microbial pathogens.
CT represents the disinfection concentration (C) multiplied by the contact time (T) (adjusted for
basin hydraulics). The plant operators measure parameters to calculate CT daily, and they use a
spreadsheet to compare the daily required CT value to the calculated CT value. The CPE team
used the data from the plant CT calculations to evaluate historical disinfection performance. An
inactivation ratio is determinedby dividing the measured CT value by the required CT value, and
the inactivation ratio values for 2015 are plotted in Figure 6. The optimization goal for disinfec-
tion is an inactivation ratio of at least 1.0 (demonstrating compliance with the regulatory require-
ment). Figure 6 shows that the ABC WTP met the goal every day during the year by a wide
margin, especially during the summer when disinfectant dosages were increased, presumably as
a precaution during a FLAB event.

FIGURE 6. Daily Disinfection Inactivation Ratio.

32

-------
Cvanotoxins

During HAB events, cyanotoxins can enter the plant as either intracellular cyanotoxins (con-
tained within a cyanobacteria; e.gMicrocystis, cell) or extracellular cyanotoxins (outside the
cell, or free). Intracellular cyanotoxins can also be released from cells if they are lysed (broken
apart) during treatment in the plant. Historically, microcystins have only exceeded the current
action level and reporting limit of 0.3 |ig/L in the finished drinking water on one occasion
(0.47 |ig/L on September 18, 2013). At that time, raw water microcystins concentrations were
reported as exceeding 10 |ig/L. Insufficient data are available, however, to evaluate the removal
or destruction of cyanotoxins by each of the individual unit processes in the plant.

On-site Studies

During the CPE, several studies were conducted to assess plant performance and process control.
These studies included: 1) filter media assessment, bed expansion, and waste backwash profile;
2) filter backwash recovery; 3) assessment of sodium permanganate (NaMn04) dose on particle
removal and cyanobacteria cell lysing; and 4) chlorophyll-a, and phycocyanin plant profile.

Study 1: Filter Media Assessment, Bed Expansion, and Waste Backwash Profile
Filter Inspection-

The purpose of the filter inspection is to visually observe physical conditions of the filter media.
A visual examination was made of the media once filter No. 4 was drained. The anthracite
media appeared to be clean and angular. There didn't appear to be mudballs or cracks through-
out the surface of the media. A small section of the filter bed was excavated by hand to observe
the degree of stratification between the anthracite and sand media. Typically, there should be a
distinct layer of anthracite over a short depth of intermixed anthracite and sand media, followed
by a distinct layer of sand underneath. The excavation in filter No. 4, shown in Figure 7,
revealed almost complete mixing of the anthracite and sand layers throughout the depth of the
filter bed profile. Re-stratificationof media following a backwash cycle is a function of media
density, filter bed expansion during backwash, and the approach used to ramp down the back-
wash flow rate. In filter No. 4, the failure of the anthracite and sand to re-stratify back to their
original respective locations following a backwash cycle may be a reflection of poor bed expan-
sion provided during the backwash cycle and the approach used to ramp down the flow rate. It is

-------
important to note that these findings only represent the inspection of one of the eight gravity fil-
ters at the plant.

Further studies could be pursued by plant staff to assess the condition of the media in the remain-
ing filters. In addition, it may be possibleto adjust current backwash procedures and examine
the resulting re-stratification until the most effective backwash and re-stratification configuration
is found. A possible issue with completely mixed filter media is a more rapid increase in filter
headloss and blinding of the filters when operating under high hydraulic and solids loading con-
ditions such as during a HAB event.

34

-------
FIGURE 7. Excavated Area of Filter No. 4 Showing Intermixing of Media.
Filter Probing

35

-------
The purpose of conducting a filter probing study is to evaluate the overall depth of media in the
filter. This is accompli shed by probing the filter at approximately equally-spaced distances
across the plan area of the filter following a grid-like pattern. Once depths are measured at the
various points across the plan area of the filter, the data points can be plotted on a map and used
to determine areas where the bed is uneven or where media loss has occurred. The CPE team
used a metal rod to manually probe and measure the media depth from the support deck to the
top of the media in filter No. 4. The media depth was measured in a grid-like pattern at 24 loca-
tions across the area of the filter, and the total depth of media at these locations ranged from
26.75 inches to 29.25 inches. The filterwas originally installed with a 15-inch layer of anthra-
cite, followed by a 12-inch layer of sand, and a three-inch layer of torpedo sand, for a total media
depth of 30 inches. Therefore, the filter probing study showed a 0.75 to 3.25-inch loss of media
in the filter. Figure 8 shows the map of the filter media bed observed in filter No, 4 during the
CPE.

36

-------
GULLET

Filter No. 4
August 3, 2016

Measurements
taken at 4-foot
increments
measured from the
gullet.

Area of Excavation:
Anthracite and sand
were completely
intermixed along
profile.

Media depths
ranged from
26.75" to 29.25".

Bed Expansion:

1.5 inches.
5% expansion.

28"

27.25"

28.25"

28.25"

tJO
O

4-J

CJ

4—'

03

to
03

27.5" 29"

(Excavation!

27.75" 28.5"

27.5" 29.5"

26.75" 27"

Ladder

tJO

o

CJ

4—'

03

to
03

29.25" 28"

28.5" 27.75'

27.5" 28.25'

28.75" 28.75

tJO
O

-M

QJ

-M
03

cn
03

20"

FIGURE 8. Filter No. 4 Filter Probing Map.

37

-------
Bed Expansion-

The purpose of conducting the bed expansion study is to determine the depth to which the media
expands during a typical backwash cycle being used at the plant. Once the depth of expansion is
determined, it can then be used to calculate the bed expansion percentage. Knowing the percent
of bed expansion helps operators understand how effective the backwash cycle is in cleaning the
media and the ability for the media to re-stratify back to its original location following a back-
wash. A minimum of 20 percent filter bed expansion is desirable; however, filters using air
scour can achieve satisfactory backwashing at lower expansion rates (i.e., 15 percent). The
equipment used to conduct the study during the CPE included a Secchi disk attached to a pole.
The CPE team marked the pole when it was sitting on the anthracite media before the filter back-
wash and again at the high backwash flow rate when the Secchi disk was observed to disappear
below the fluidized anthracite. The distance between these two marks represents the depth of
media expansion.

The CPE team attempted to determine the depth of expansion during the high-rate portion of the
filter No. 4 backwash cycle. The initial bed expansion measurement of 1.5 inches indicated a
bed expansion rate of only five percent. During an attempt to re-check the bed expansion depth,
the Secchi disk detached from the pole and was made unusable. Plant staff members were able
to retrieve the Secchi disk from the top of the filter media bed during a subsequent draining of
the filter. The plant staff is encouraged to repeat the filter bed expansion study to confirm the
bed expansion rate during backwash of each of the filters.

Backwash Waste Turbidity Profile-

The purpose of conducting a waste backwash profile study is to determine the amount of time
necessary for effective media cleaning. The equipmentused to perform the waste backwash pro-
file study during the CPE included a sample collection device and a turbidimeter. During the fil-
ter No. 4 backwash, the CPE team attempted to collect turbidity grab samples from the waste
trough using a long pole with a sample bottle attached at the end. The CPE team was able to col-
lect six samples from the start of the backwash cycle through ten minutes into the cycle, at vary-
ing intervals. Due to the force of the water from the backwash launders, the sample bottle
became detached from the pole and discharged from the filter along with the backwash
wastewater. Due to the loss of the sampler, the six, seven, and eight-minute samples were

-------
missed, but a ten-minute sample was collected. The backwash waste turbidity profile is shown
in Figure 9. These data indicate that the filter was adequately cleaned. This study can also be
used to determine an optimum turbidity level to stop backwashing and to determine if excess
backwash water is being used during filter backwash. The plant operators are encouraged to
periodically conduct this study to support optimization of the filter backwash procedure.

80
70
60
50

3"
z

¦S" 40
la

30
20
10
0

0 2 4 6 8 10 12

Time [min]

FIGURE 9. Filter No. 4 Waste Backwash Water Turbidity Profile.

Backwash Turbidity Profile - Filter No. 4 - 8/3/16

Overview:

• Maximum turbidity may have been missed because of
unexpected overflow during beginning of air scour.

• Initial turbidity of 75 NTU gradually decreased duringthe
backwash to ~ 1 NTU at 10 minutes.

• The 6,1, 8 minutes were missed due to loss of sample cup.
The backwash duration was adequate to clean the filter.

Study 2: Filter Backwash Recovery

The optimization goal for plants with filter-to-waste capability is to return the filter to service at
<0.10 NTU. Following the inspection and backwash of filter No. 4 during the CPE, the tur-
bidity recovery of the filter was monitored during filter-to-waste. The filter-to-waste profile is
shown in Figure 10. The turbidity spiked to 4 NTU and gradually decreased to 1 NTU after 12
minutes; however, the turbidity remained around 1 NTU for several more minutes. Because of
this high turbidity, the filter was not immediately placed in service. The operator commented
that this was not a typical filter-to-waste recovery and thought that the draining of the filter might
have resulted in air entrainment in the filter media and underdrain. The operator also commented

-------
that the pumping of the filter effluent samples to the turbidimeters has frequently resulted in high
turbidity spikes that the operators need to address.

Filter No. 4 Filter-to-Waste Profile 8/3/16

Overview:

•	Filter backwash (BW) ripening from inspected filter.

•	Turbidityduringfilter-to-wastespiked at about

4 NTU and recovered to ~ 1 NTU after 15 minutes. Filter
was not returned to service due to high turbidity.

•	Operator suspected that air was introduced into the pump
suction line due to draining of the filter.

•	The operator said that this was not a typical BW ripening

response duringfilter-to-waste.

6	8	10	12	14

Elapsed Time from Beginning of Filter-to-Waste (min)

FIGURE 10. Filter-to-Waste Profile for Inspected Filter No. 4.

To assess a normal turbidity recovery following a filter backwash, the CPE team evaluated tur-
bidity data from filter No. 8 that was also backwashed on the same day. These data are shown in
Figure 11. During the filter-to-wasteperiod of 15 minutes, the turbidity varied from 0.11 to
0.13 NTU. After the filter was returned to service to the clearwell, the turbidity spiked to
0.14 NTU and did not reach the optimization goal of 0.10 NTU for another 50 minutes. This
filter recovery did not meet the optimization goal of achieving <0.10 NTU by the end of the
filter-to-waste period. Achieving this goal following each filter backwash reduces the number of
particles (includingpathogens and cyanobacteria cells) that pass to the clearwell and, as a result,
enhances public health protection.

40

-------
Filter No. 8 Filter-to-Waste and Filter Run Profile 8/3/16

o -(

10

20	30	40	50	60

Elapsed Time from Beginning of Filter-to-Waste (min)

70

80

90

FIGURE 11. Filter-to-Waste Profile for Filter No. 8 Backwash.

Study 3: Assessing NaMnOj Dose on Particle Removal and Cyanobacteria Cell Lysing

Sodium permanganate (NaMn04) is added at the raw water pump station as a pre-oxidant to
enhance turbidity removal and to oxidize organics to reduce taste and odor in the finished water.
While sodium permanganate addition can provide treatment benefits, there could be negative
impacts of adding this oxidant to the raw water when cyanobacteria cells are present. The poten-
tial exists for the permanganate to disrupt the cyanobacteria cells and release the intracellular
cyanotoxin into the water. The ABC operators currently feed permanganate in the range of 0.7
to 1.1 mg/L and target a residual of 0.3 mg/L at the head of the plant just prior to coagulant
addition.

41

-------
The purpose of this study was to assess the benefits and negative impacts of permanganate
addition at the ABC WTP. Two study hypotheses were identified:

1. The addition of NaMn04 to the raw water will improve floe formation through the coag-
ulation process and improve particle settleability through the clarification process.

2. As the NaMnC>4 dose increases in the raw water, it will contribute to cell disruption and
an increase in extracellular microcystins concentration.

Approach-

This study was conducted using j ar testing equipment provided by the plant. Since the raw water
microcystis cell concentration was relatively low during the CPE, the evaluation team augmented
the cell concentration in the raw water sample. Phycocyanin measurements were taken with a
data sonde from both the river intake site (near bank) and raw water wet well in the pump
station. Since concentrations were higher in the river, a plankton net was used to concentrate
phytoplankton cells from the river source water. The concentratedbiomass collected via the
plankton net was mixed into approximately four gallons of river water for use in the jar test.

The plant's coagulant, aluminum chlorohydrate (ACH), and NaMnC>4 were added to the jars dur-
ing the test. A micropipette was used to deliver the neat ACH dose to the j ars. Knowing that
1 microliter (|iL) of water is equal to 1 mg of water, the ACH volume to deliverto the jars was
calculated using the ACH specific gravity and jar volume. The following sample calculation for
a 24 mg/L ACH dose shows that the required ACH delivered to a 2 liter j ar would be 36 |iL.

24 mg 1 iiL ACH

—;— x ttttt x 2 L jar = 36 \iL

L 1.32 mg ACH

For NaMnC>4, a 0.2 percent stock solution of the chemical was prepared. This stock solution con-
centration resulted in a 1 mg/LNaMn04 dose to the 2-literjarsfor every 1 milliliter of stock
solution added.

The j ar test settings were estimated based on hydraulic detention time through the plant and the
current jar test settings used by the plant staff. The settings used in this study are listedbelow.
The permanganate contact time was reduced from the actual time (approximately one hour)

-------
because of the limited time to conduct the study. The j ars were sampled for turbidity and total
microcystins following the eight-minute settling time.

1. Permanganate contact time - 30 minutes @30 rpm

2. Rapid mix-2 minutes @ 175 rpm

3. Flocculation- 15 minutes @ 30 rpm, then 15 minutes @ 20 rpm

4. Settling time - 8 minutes

In addition to the sampling time of eight minutes that represented the settling rate in the clarifi-
ers, additional sampling times were setforjarNo. 2 and No. 3 starting immediately after the
mixer was turned off (i.e., time zero minutes) and continuing at one, two, four, six, and ten
minutes. These sample times were used to support the development of settling curves from these
two jars.

The j ars were set up with the following conditions:

Jar 1 - blank (no chemicals)

Jar 2 - ACH dose = 24 ppm (plant dose)

Jar 3 - ACH dose - 24 ppm, NaMnC>4 = 1.2 ppm (plant dose)

Jar 4 - ACH dose - 24 ppm, NaMnC>4 = 3 ppm (high dose where cyanotoxin release
may occur)

Testing from the samples collected from the jars included turbidity, total microcystins, and extra-
cellular microcystins. Turbidity was also measured from Jar No. 2 and No. 3 from the samples
collected to develop the settling curves. Following the sampling for turbidity, samples were col-
lected from each of the jars for testing total and extracellular microcystins at the State EPA lab.
A small volume of sodium thiosulfate was added to these samples to quench the permanganate
residual to stop any further oxidation reactions in the samples.

Results and Conclusions-

To assess the impact of NaMnC>4 on the formation and settleability of floe particles, settling
curves were developed from samples collected from Jar No. 2 (coagulant only) and No. 3 (coag-
ulant and NaMnC>4). Figure 12 shows a plot of settling time versus turbidity for these two j ars.
Observations of these two j ars during the testing showed the formation of higher density, larger
diameter floe particles in the j ar with permanganate addition. The initial part of the settling
curve for the j ar with permanganate supports this observation with the j ar having higher initial

-------
turbidity than Jar No. 2 (time zero samples) and faster settling particles, as indicated by lower
turbidity from the two- and four-minute samples. However, this trend reversed in the six- and
ten-minute samples. The reason for this trend change is unknown but could be due to some type
of sampling or testing error or color interference from the jar with permanganate addition. This
trend change makes the study results somewhat inconclusive, and repeating this study would be
needed to confirm the impact of NaMnC>4 on coagulation and particle settleability.

50
45
40

, 30

3
I—

.£* 25

jg
1q

l_ 20
15
10
5
0

0 2 4 6 8 10 12

Time (min)

FIGURE 12. Settling Curves for Water Treated With and Without NaMnC>4.

Settthng Curve Comparison

ACH

- ACH + NaMnO

4 1

lj.

\ x ^

¦7

Expected cur\

The results of the NaMnC>4 addition on the potential for cyanobacteria cell lysing and release of
microcystinsis shown by the results in Figure 13. The first two samples show the total and
extracellular microcystins of the augmented raw water sample and another augmented raw water
sample that was mixed for 30 minutes. Both show similar results, with the majority of the
microcystins being contained within the cells (i.e., intracellular). Sample 3 shows that about
85 percent of the total microcystins were removed through coagulation, flocculation, and sedi-
mentation. Comparing the results from Samples 2, 3, and 4 indicates that, as the NaMnC>4 dose
was increased in the j ars from zero permanganate in Sample 2 to 3 mg/L in Sample 4, the extra-
cellular microcystins concentration increased from about 20 percent of the total concentration to

-------
almost 80 percent of the total concentration. The extracellularmicrocystins concentration
increased slightly in the sample, with 1.2 mg/L of permanganate (i.e., plant dose).

Microcystin Data
Jar test conducted on August 3, 2016

3. 20

O 10

Total Microcystin

Extracellular Mirocystin

0 = Augmented raw

1 = Augmented stirred raw

2 = Augmented coagulated with ACH @ 24 mg/L

3 = Augmented coagulated with ACH @ 24 mg/L & NaMn04 @ 1.2 mg/L

4 = Augmented coagulated with ACH @ 24 mg/L & NaMn04 @ 3 mg/L

FIGURE 13. Impact of NaMn04 on CyanotoxinRelease and Extracellular

Microcystins Concentration.

Implementation-

The results of this study were inconclusive regarding the benefits of NaMnC>4 addition on
coagulation and particle removal. Due to conflicting results from the settling curves developed
from thejar test, it is recommended that this part of the study be repeated. The development of
accurate settling curves requires practice with the sampling techniques used when conducting
multiple sampling events from the same jar. When consistent sampling techniques are followed
during this study, more reliable settling curves should result.

The microcystins testing conducted as part of this study does support optimization of particle
removal as the primary mechanism for removal of total microcystins in the plant. The testing
also indicates that extracellular microcystins concentration does increase when NaMnC>4 is added
to the water; however, the most significant increase occurred at the higher permanganate dose of
3 mg/L. Repeating this study is recommended to confirm these findings.

-------
Study 4: Chlorophyll-a, Phycocyanin, and Microcystins Plant Profile

Process control sampling through the water treatment plant provides information on how each
unit process is performing relative to water quality goals or targets. A plant profile is a useful
way of trending these process control sampling results. Especially during a harmful algal bloom
(HAB), it is important for water utilities to understand how each water treatment unit process is
performing at removing cyanobacteria cells and cyanotoxins, while maintaining other treatment
objectives, such as turbidity and TOC removal and disinfection. Plant profile trending can pro-
vide operators with warning of a source water HAB propagating through the treatment plant such
that incremental process control changes can be made to avoid detection of cyanotoxins in the
finished water.

Approach-

A plant profile was developed from sampling results obtained from a combination of grab sam-
ples and data sonde readings at locations in the water treatment plant indicated in Table 9. Chlo-
rophyll-a, phycocyanin, total and extracellular microcystins, temperature, pH, and turbidity data
were collected and trended in a spreadsheet. The pigments chlorophyll-a and phycocyanin are
used as indicators of total algal biomass (chlorophyll-a) and cyanobacteria biomass (phycocya-
nin). It should be noted that each unit process sample was collected in conjunction with the
plant's normally scheduled sampling for turbidity. As such, the samples do not represent the
same slug of water as it is flowing through the plant. If the water system conducts future unit
process sampling, they should consider timing sample collection to mimic flow through the
plant, since raw water microcystins concentrations can vary over time. For this profile, raw
water samples were collected at the raw water pump station wet well approximately 30 minutes
prior to the first unit process sample. The remaining samples were collected sequentially
through the plant.

-------
TABLE 9. Plant Profile Sampling Locations

Location

Rationale

Raw water (Lake) (surface grab from wet well prior to
sodium permanganate addition)

Determine concentrations of cyanotoxins and cyano-
bacteria biomass indicators that are entering the
water treatment plant.

Pre-sedimentation basin (surface grab from first
chamber after inlet from raw water)

To understand the effect of sodium permanganate
(NaMn04) pre-oxidation on total and extracellular
cyanotoxin concentrations (and biomass indicators).

Clarifier 1,2, and 3 effluent

Determine the concentrations of cyanotoxins and indi-
cators leaving each clarifier. Understand the effect of
coagulation/flocculation/sedimentation process on
cyanobacteria cell removal, cyanotoxin
concentrations, and biomass indicators. Sampling
each clarifier helps indicate any potential performance
issues with individual clarifiers.

Applied/top-of-filter

Represents combined clarifier effluent and is repre-
sentative of water quality being applied to the filters.

Transfer well/combined filter effluent (microcystins
grab sample collected from tap; sonde meas-
urements taken immediately post-chlorine addition).

Determine the concentrations of cyanotoxins and indi-
cators leaving the filters. Understand the effect of the
filters on cyanobacteria cell removal, cyanotoxin con-
centrations, and biomass indicators. Represents the
water quality entering the clearwells for the disinfec-
tion process.

Plant tap(EPOOl)

This represents "finished water." Determine the con-
centrations of cyanotoxins entering the water distribu-
tion system (if any, at this point). The microcystins
grab sample was collected from the tap. No sonde
measurements were taken at this location.

Sludge lagoon inlet and outlet, sludge tower No. 2
well

A combination of filter backwash wastewater and
clarifier sludge blowdown water. Help understand if
cyanobacteria are still viable and potentially producing
cyanotoxins in the clarifier sludge and/or filter media.
Both inlet and outlet were sampled to understand the
effect of the lagoons on improving water quality (spe-
cific to HABs) prior to discharge.

River intake (surface grab off of dock adjacent to
intake)

Understand potential differences in water quality
between the River and Lake intakes.

Samples that were collected at the individual clarifier effluents, applied/top-of-filter, and transfer
well/combined filter effluent locations were matched (samelocation and collection time) with

-------
grab samples collected by the operator for turbidity analysis. Plant profile samples were col-
lected at two different times of day to coincide with normal operator sampling: 12:00 noon and
4:00 p.m.

A YSI 6600 multi-parameter data sonde was used to measure pH, temperature, chlorophyll-a,
and phycocyanin, both as relative fluorescence units (RFU) and estimates of concentration (|ig/L
of chlorophyll-a, and cyanobacteriacells/mL using a pre-programmed phycocyanin calibration
curve). The data sonde's calibration was verified with standards for each parameter prior to use.
The sonde was allowed to stabilize at each sampling location prior to recording output to mini-
mize any error associated with the sonde's transfer from the previous sampling location. On
occasion, the sonde's internal optics cleaning mechanism was utilized to ensure that fouling of
the optical sensors was minimized between sampling locations.

Grab samples were collectedfor total and extracellular microcystins analysisusing 125 mL poly-
ethylene terephthalate (PETG) containers. All samples that had been subjected to an oxidant
(i.e., sodium permanganate or chlorine) were quenched with a 10 mg sodium thiosulfate tablet.
All samples were analyzed using the Ohio EPAMicrocystins-ADDAELISA method6.

Results and Conclusions -

As expected, the plant profile developed from the sampling results depicts a decreasing trend for
all parameters (except pH), indicating that cyanobacteria cell and cyanotoxin removal is occur-
ring through the plant (Figures 14 and 15). The greatest percentage of removals of chlorophyll-a,
phycocyanin, total and extracellularmicrocystins occurred during the coagulation/
flocculation/sedimentation process (i.e., from pre-sedimentationbasins through the clarifiers).
This emphasizes the importance of the settling process for removing cyanobacteria cells in
conventional water treatment plants (see Figures 16 and 17 for percent removals). Removal of
extracellularmicrocystins also occurred to a significant degree during the settling process, likely
because ABC also adds powdered activated carbon (PAC) in the rapid mix prior to the clarifiers.
Most microcystins concentrations were below the reporting limit of 0.3 |ig/L within the

6 Ohio EPA Total (Extracellular and Intracellular) Microcystins - ADDA by ELISA Analytical Methodology. Ohio
EPA DES 701.0. Version 2.2. November 2015. Retrieved April 28,2016, from

http://www.epa.oliio.gOv/Portals/28/documents/rules/draft/01iio%20EPA%20DES%20701.0%20Version%202.2 D
ec2015.pdf

-------
treatment plant; however, the State's lab has a Minimum Detection Limit (MDL) of 0.09 |ig/L
for their ADDA-ELIS A method. Therefore, the data are presented here for the sake of under-
standing the removals being achieved by each unit process. Plant tap samples were analyzed for
total and extracellular microcystins concentrations, but data sonde measurements were not col-
lected. Both the 12:00 noon and 4:00 p.m. finished water samples were non-detect based on the
0.3 |ig/L reporting limit; however, the concentrations were 0.24 |ig/L and 0.14 |ig/L.

12:00 Sampling Period (8/3/16)









































i i

¦





|i

lb

llll 1

ii. ¦ ¦ .i	

Rawwater Pre-sedimentation Clarifierl	Clarifier 2	Clarifier 3 Applied/Top of Transferwell /

basin (post-	filter	combined filter

NaMn04)	effluent

¦ Chlorophyll [RFU] Chlorophyll [ug/L] ¦ Phycocyanin [RFU] ¦ Total MC [ug/L] ¦ Extracellular MC [ug/L]

FIGURE 14. Plant Profile for 12:00 Hour Sampling Period on August 3, 2016.

49

-------
16:00 Sampling Period (8/B/16)

























1







1

1 .1.. ill. ll.ll Jill .III.

Pre-sedi mentation
basin (post-NaMn04)

Clarifier 3	Applied/Top of filter Transfer well /

combined filter effluent

¦ Chlorophyll [RFU] ¦ Chlorophyll [ug/L] ¦ Phycocyanin [RFU] ¦ Total MC [ug/L] I Extracellular MC [ug/L]

FIGURE 15. Plant Profile for 16:00 Hour Sampling Period on August 3, 2016.

Ottawa Raw 2 Pre-Sed Basin Clarifierl	Cla rifier 2	Cla rifier 3	AppliedTOF TransferWell OttawaTap

FIGURE 16. Plant Process Percent Removals of Total and Extracellular Microcystins

at the 12:00 Hour Sampling Time.

50

-------
Ottawa Raw 1 Pre-Sed Basin Clarifierl Clarifier 2 Clarifier 3 AppliedTOF TransferWell OttawaTap
-120 	



l—l

1

¦—1.	¦=-

—1

H

H

1=1	1 I



FIGURE 17. Plant Process Percent Removals of Total and Extracellular Microcystins

at the 16:00 Hour Sampling Time.

Total microcystins data obtained from the sludge lagoons (supernatant) indicated a decrease from
backwash inlet (0.52 |ig/L)tooutlet(0.15 |ig/L). Chlorophyll-a and phycocyanin data followed
a similar decreasing pattern through the sludge lagoons. Total microcystins were concentrated in
the sludge transfer well sampling location (14 |ig/L), demonstrating the effectiveness of the clari-
fiers at removing cyanobacteria cells. Extracellular microcystins were 0.86 |ig/L in the sludge
transfer well, possibly indicating that some cell death and lysis were occurring. The sludge itself
was not sampled, since an analysis method for microcystins in sediments is still under develop-
ment by the State EPA.

Phycocyanin correlated relatively well with total microcystins concentration (R2 = 0.67), as
demonstrated in Figure 18. This suggests that phycocyanin may be a good indicator of cyano-
bacteria cell presence and total microcystins concentrations, especially for developing future
sampling approaches and plant profiles during a HAB (such that 1 phycocyanin RFU ~

0.85 |ig/L total microcystins, which is in line with other published studies). A correlation plot
was also developed for total microcystin and chlorophyll-a (see Figure 19); however, the lower
R2 value for that relationship demonstrated a less direct correlation (R2 = 0.49). Because
chlorophyll-a is produced by all forms of algae, this less direct correlation with total micro-
cystins may indicate that other forms of algae are also present in the raw water (e.g., green algae,
diatoms, etc.).

51

-------
Total Microcystins vs. Phycocyanin - 8/3/16

3.0

2.5

5" 2.0

LL

0£

C

ra 1.5
>

u
O
u

I" l.o

0.5
0.0

0.0	0.5	1.0	1.5	2.0	2.5	3.0

Total Microcystins [|_ig/L]

FIGURE 18. Total Microcystins and Phycocyanin Correlation.

3.5

3.0

„ 2.5
=3

LL
££

to 2.0
>

f,5
o

0.5
0.0

0.0	0.5	1.0	1.5	2.0	2.5	3.0

Total Microcystins [|_ig/L]

FIGURE 19. Total Microcystins and Chiorophyll-a Correlation.

52



















•

y = 0.8548X
R2 =0.6675

	











• •

























• mm'' •

•



























•

























y = 0.6362x
R2 =0.4882

•

	





<

•

~ 	

	

	





	

	





•

.•v	

	

•



•

-------
Further Implementation-

ABC may consider repeating the plant profile sampling process; for example, by using
phycocyanin as an indicator of cyanobacteriacell removal or determining total and extracellular
microcystins concentrations on a regular basis, especially if microcystins concentrations in the
raw water increase. It is important to regularly understand each unit process's performance,
especially during a HAB, to ensure a robust multiple barrier treatment scheme.

MAJOR UNIT PROCESS EVALUATION

Maj or unit processes are assessed with respect to their capability to meet the optimized goals for:

•	Settled water turbidity

•	Filtered water turbidity

•	Disinfection (inactivation ratio goal)

•	Cyanotoxin adsorption through the use of powdered activated carbon

•	Cyanotoxin destruction (oxidation) through the use of chlorination

There is an emphasis on turbidity reduction to remove cyanobacterial cells through the multiple-
barrier treatment process; however, recognizing that toxins will likely be in the raw and settled
water, toxin removal and destruction are also considered. The capability of each individual unit
process is also assessed to verify its capability to provide consistent optimized performance.

Since the treatment processes of the plant must provide multiple effective barriers at all times, a
peak instantaneous operating flow is also determined. The peak instantaneous operating flow
represents conditions when the treatment processes are the most vulnerable to the passage of par-
asitic cysts, microorganisms, and toxins. If the treatment processes are adequate at the peak
instantaneous flow, then the maj or unit processes should be capable of providing the necessary
effective barriers at lower flow rates. The flow through the plant is controlled by raw water
pumps, each equipped with a variable frequency drive that allows operators to adjust the flow
rate up to the pump capacity. Through discussions with operators regarding operational policies
at the plant as well as the review of plant operating records by the CPE team, the rate of 6 MGD
was selected as the peak instantaneous flow rate through the plant under normal operating condi-
tions. This rate was used to assess the capabilities of the major unit processes.

53

-------
Unit process capability is assessed using performance potential graphs, where the projected treat-
ment capability of each maj or unit process is compared against the peak instantaneous operating
flow rate and the plant design flow for comparison. An individual performance potential graph
is developed for each of the treatment objectives evaluated in this report: 1) microbial
removal/inactivation and 2) cyanotoxin removal and destruction.

Particle Removal and Microbial Disinfection

The Maj or Unit Process Evaluation graph for microbiological treatment through turbidity
removal and disinfection, developed for the ABC WTP, is shown in Figure 20. The unit
processes evaluated during the CPE are shown along the vertical axis. The horizontal bars on the
graph represent the projected peak capability of each unit process that would support
achievement of optimized process performance. These capabilities were proj ected based on sev-
eral factors, including: the combination of treatment processes at the plant, the CPE team's
experience with other similar processes, raw water quality, industry guidelines, the ABC WTP
design, and regulatory standards.

Each unit process can fall into one of three categories:

Type 1: Where the bar for the unit process exceeds the peak instantaneous flow

(>100 percent of peak flow), the plant should be expected to achieve the perfor-
mance goals.

Type 2: If the bar for the unit process falls short but is close to the peak instantaneous flow
(80 to 100 percent of peak flow), then operational adjustments may still allowthe
plant to achieve the performance goals.

Type 3: If the bar for a specific unit process falls far short of the peak instantaneous flow
(<80 percent of peak flow), then it may not be possible to achieve the performance
goals with the existing unit process.

54

-------
Flocculation (A)

Sedimentation (B)

Conventional Filtration (C)

Disinfection (Giardia) (D)

0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24

Flow Rate, MGD

(A)	Flocculation: Capacity calculation using the volume under the clarifier sludge recirculation zone, for 30-min detention time.

(B)	Sedimentation: Capacity calculation using the SOR average in the clarifier sedimentation zone to achieve a settling velocity of 0.7 gpm/ft2.

(C)	Conventional Filtration: Calculation based on eight 18x20 ft filters with one out of service. Loading rate to achieve 4 gpm/ft2.

(D)	Disinfection (Giardia inactivation): Assumptions - pH 8.2 (HAB conditions), temperature = 0 °C (winter time), 2 mg/L free chlorine residual. Clearwell volume
(626,133 gallons) based on minimum level (10 ft) and 73 ft. diameter.

FIGURE 20. Maj or Unit Process Evaluation Graph - Particle Removal and Microbial Disinfection.

55

-------
The shortest bar represents the unit process that may limit plant capability the most relative to
achieving optimized plant performance. The major unit processes evaluated include: floccula-
tion, sedimentation, filtration, and disinfection. The flocculation and sedimentation processes
both occur in the solids contact clarifiers at the ABC WTP, and the disinfection process takes
place in the clearwells. The approach and calculations used to determine the rating for each
process are provided in Appendix A.

The unit process performance potential for each of the processes summarized in the Figure 20
graph shows that all unit processes are rated as Type 1 processes, capable of meetingthe particle
removal and microbial treatment obj ectives at the assigned peak instantaneous flow rate of
6 MGD through the facility. The graph also shows that the maj or unit processes can achieve the
particle removal and disinfection goals at the plant design rate of 9 MGD.

Cvanotoxin Removal and Destruction Treatment

In the event that a HAB occurs at one or both of the ABC WTP sources and cyanotoxins appear
in the raw water, the particle removal processes in the water treatment plant would be able to
remove the maj ority of the intracellular cyanotoxins, provided the cells are removed before
release the cyanotoxins. The pre-oxidant sodium permanganate feed may also have to be
carefully controlled to prevent the cyanotoxins from being released before the cells are removed.
The evaluation of the maj or unit processes in the microbiological (turbidity) control section of
this report gives an estimate of the plant capacity to remove cyanobacteria cells through
clarification and filtration to control the intracellular cyanotoxins. Any extracellular cyanotoxins
present in the raw water or released in the plant would have to be removed primarily through
powdered activated carbon (PAC) adsorption or destroyed through chlorine oxidation in the
plant. The Maj or Unit Process Evaluation graph in Figure 20 would apply to ability of the plant
to remove intracellular cyanotoxins.

The Maj or Unit Process Evaluation graph for extracellular cyanotoxin treatment through pow-
dered activated carbon (PAC) adsorption and chlorine oxidation, developed for the ABC WTP, is
shown in Figure 21. A target total microcystins concentration of 100 |ig/L was used in the
evaluation due to historic concentrations observed in the western basin of this lake. The unit
processes evaluated during the CPE are shown along the vertical axis.

-------
Plant Design Capacity - 9 MGD
Peak I nstantaneous Flow - 6 MGD

9.59

0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24

Assumptions

(A1) PAC Feed (current): Assume 30 mg/L PAC dose is needed to achieve 90% removal during an elevated HAB event (e.g., 100 ug/L microcystins
entering the plant), based on Mohamed et. al. equations and Newcombe 2009 charts with a safety factor. Dose was estimated based on medium
auger feed rate of 70 Ib/hr.

(A2) PAC Feed (potential): Assume 30 mg/L PAC dose is needed to achieve 90% removal during an elevated HAB event (e.g., 100 ug/L
microcystins entering the plant), based on Mohamed et. al. equations and Newcombe 2009 charts with a safety factor. Piping and delivery changes
may be necessary for this option. Dose was estimated based on a combined feed rate of the medium and small augers of 100 Ib/hr, potentially
feeding into pre-sed basin and rapid mix simultaneously.

(B1) Cyanotoxin Oxidation: Required CT calculated using the AWWAspreadsheet. Assumptions: microcystins concentrations (10 ug/L entering
clearwell, 0.3 ug/L entry point), chlorine residual of 4 mg/L, temperature of 20 °C, and pH of 8.8. Clearwell volume (313,066 gallons) based on
minimum level (10 ft) and 73 ft. diameter.

(B2) Same assumptions as B1 but using a safety factor of 2.

FIGURE 21. Cyanotoxin Treatment Maj or Unit Process Evaluation Graph.

57

-------
The horizontal bars on the graph represent the proj ected peak flow capability for each unit pro-
cess that would support achievement of optimized process performance. These capabilities were
proj ected based on several factors, including:

PAC Feed Capacity: Discussions with WTP operators and outside experts on PAC feed
rates and their feasibility at the ABC WTP considering: safety, chemical storage and
acquisition, and the anticipated physical demands on operators of a sustained two- to three-
week high feed rate during an extracellular microcystins event.

Microcvstins (Cvanotoxin) Oxidation: An estimate was based on the AWWA Hcizen-
AdamsCyanoTOX Tool for oxidation kinetics (version 1.0)7. For comparison, the estimate
was also performed using a safety factor of 2 to account for uncertainties in applying the
AWWA tool to the actual conditions at ABC (i.e., competing oxidant demand such as
NOM, pre-oxidants like sodium permanganate, etc.).

In evaluatingthe microcystins control processes in the ABC WTP, each process is assigned a
rated capacity, based on a comparison of the rated capacity to the peak instantaneous flow (6
MGD) at the plant. Results of this evaluation are presented in Figure 21, and each unit process
can fall into one of three categories:

Type 1: Where proj ected peak capability for the unit process exceeds the peak instantaneous
flow (>100 percent of peak flow), the plant should be expected to achieve the per-
formance goals.

Type 2: If the proj ected peak capability for the unit process falls short of, but is close to, the
peak instantaneous flow (80 to 100 percent of peak flow), then operational adjust-
ments may still allow the plant to achieve the performance goals.

Type 3: If projected peak capability for a specific unit process falls far short of the peak
instantaneous flow (<80 percent of peak flow), then it may not be possible to
achieve the performance goals with the existing unit process.

7 AWWA Cyanotoxins resource site: http://www.awwa.org/resources-tools/water-kiiowledge/cYanotoxiiis.aspx

58

-------
The lowest proj ected process peak capability (flow rate) represents the unit process that may
most limit plant capability relative to achieving optimized plant performance. The major unit
processes evaluated include PAC feed and oxidation. PAC is fed by two dry chemical feeders (a
third spare unit is provided, which is not currently active, but can be placed into service if neces-
sary). Each feeder has a dedicated hopper, filled manually with bags of PAC, which dispenses
PAC to the feeders. The feeder meters the PAC into a slurry solution through an auger that can
be adjusted to control the PAC feed rate. A carrier, or dilution, water line draws in the PAC to
form a slurry that is then carried and fed into the rapid mix basin. Feed lines also exist and can
be connected to carry the slurry (or additional slurry from one of the feeders acting inde-
pendently) to the presedimentation basin, which was originally intended to be a PAC contact
basin. The oxidation process takes place through the application of chlorine in the clearwells
after the particle removal treatment processes.

The unit process performance potential summarized in Figure 21 shows that the PAC feed and
the oxidation processes are rated as a Type 1. However, another bar for the oxidation process
shows that this process would be in the Type 3 range, if a safety factor of 2 is applied to account
for a lack of real plant data and the uncertainties of applying the AWWA model to predict oxida-
tion. The plant operators could not lower the pH of the water entering the clearwells to try to
make the oxidation reaction more efficient, but they could raise the chlorine concentration
higher, to 4 mg/L, on a short-term basis if conditions called for it. Considering the uncertainties
in the unit process evaluation but the flexibility in raising the chlorine dosage on a short-term
basis if necessary, the CPE team rated the oxidation process a Type 2 process. The plant may
not have the capability to meet the optimization goals under normal operating conditions, but
operational adjustments may make it possible to meet the goals on a short-term basis during a
FLAB event.

The overall major unit process summary for microbial and cyanotoxin removal and destructionis
summarized in Table 10 below. The particle removal and microbial disinfection unit process
ratings are all classified as Type 1, indicating that the plant has the capability to achieve the
microbial optimization goals when excellent process control skills are applied. Due to uncertain-
ties in the assumptions made during the evaluation, the cyanotoxin removal and destruction pro-
cesses are rated as a more conservative Type 2, indicating that the plant has the capability of

-------
achieving the microcystins finished water target, assuming more aggressive attention is given to
plant O&M to prepare for and treat through a significant HAB event.

TABLE 10. Major Unit Process Summary

Microbiological Treatment

Major Unit Process

Rating

Flocculation'1'

Type 1

Sedimentation(1)

Type 1

Filtration(1)

Type 1

Disinfection/Oxidation(1)

Type 1

PAC Adsorption Process,2)

Type 2

Chlorine Oxidation,2)

Type 2

(1)	Microbial treatment

(2)	Extracellular cyanotoxinremoval and destruction

PERFORMANCE-LIMITING FACTORS

The areas of design, operation, maintenance, and administration were evaluated to identify fac-
tors that limit performance. These evaluations were based on information obtained from the
plant tour, interviews, performance and design assessments, studies, and the judgment of the
CPE team. Each of the factors was assessed for the possible classification as A, B, or C
according to the following guidelines:

A Major effect on a long term repetitive basis

B Moderate effect on a routine basis, or major effect on a periodic basis
C Minor effect

The performance-limiting factors identified were prioritized as to their relative impact on perfor-
mance. They are summarized below. While developing the list of factors limiting performance,

60

-------
over 50 potential factors were reviewed, and their impact on the performance of the ABC WTP
was assessed. There were three "A " factors and two "B*" factors identified. Note that the
asterisk on the"B" factor (B*) refers to a performance-limiting factor identified for the specific
situation when the plant is facing a harmful algal bloom in its source water and must remove
cyanobacteriaand cyanotoxins during treatment.

Policies - Administration (A)

• The numerical optimization goals for individual clarifier effluent, individual filter efflu-
ent, and combined filter effluent turbidity have not been officially adopted.

Application of Concepts and Testing to Process Control - Operations (A)

• Documented operational guidelines identify a target turbidity value of 0.25 NTU for initi-
ating a backwash and returning a filter to service after filtering to waste. These individ-
ual filter effluent turbidity values exceed the optimization performance goal of 0.10 NTU.

• Staff are aware of extensive filter media mixing and limited bed expansion, but studies
have not been conducted to investigate problems and possible solutions (e.g., assessing
alternative air scour and backwash procedures). Mixing of media could limit filter run
time and performance during higher hydraulic and solids loading rates (e.g., during HAB
event).

• Staff are not trending the daily maximum raw, settled, IFE, and CFE turbidities over
time.

• Studies are not being conducted to assess HAB control (e.g., carbon feed, NaMnC>4 feed).

• Capability to feed higher carbon doses in the plant for a HAB event has not been
adequately tested (i.e., address treatment limitations).

• IFE particle counters are available for process control but are not calibrated (only
cleaned). Particle counters can be effective for assessing cyanobacteria cell removal
through filters.

-------
Process Instrumentation/Automation- Design (A)

• The location of all IFE and CFE turbidimeters and particle counters and the type of sam-
ple pump requires significant suction lift to transport the sample stream to the instruments
for analysis, resulting in more frequent interruptions in monitoring and potentially erratic
readings.

Reliability- Administration/Design (B*)

• The sustainability of manually adding PAC to the hoppers during a long-term HAB event
is questionable.

• The feasibility of adding PAC at rates above 10 mg/L to the rapid mix is limited by the
existing design of the supply lines (configuration of supply lines, undersized eductor, and
carrierwater pressure).

• PAC feed lines to the presedimentation basin are not connected, and the design is prone
to excessive plugging.

Process Control Testing - Operations (B*)

• The water system's ability to optimize individual treatment processes during a HAB
event is limited by a lack of information concerning concentrations of total and extracel-
lular microcystins throughout the treatment train.

• Phycocyanin measurements are not being obtained throughout the treatment train. This
information could assist in optimization of intact cell removal.

EVALUATION FOLLOW-UP

The State EPA has not established an approach for providing follow-up trainingto CPEs at the
current time. Additional HAB-focused developmental CPEs are planned at other water utilities
over the next year. Following these events, the State EPA will be considering follow-up strate-
gies to support common CPE findings and performance-limiting factors. Plant staff are encour-
aged to contact EPA staff regarding any questions or comments they may have regarding specific
findings from this CPE.

-------
The ABC WTP staff and management are commended on their proactive approach to operation
and maintenance of their treatment plant and to addressing HAB treatment challenges. This CPE
has identified further areas that can be pursued to enhance particle removal performance and be
better prepared for future HAB events. An excellent place to start the optimization process is
collecting and trending optimization data such as the approach demonstrated in the Historical
Water Quality Performance Assessment section of this report. The studies conducted during this
CPE also demonstrate a structured approach for conducting problem-solving activities by plant
staff. The following section includes several study ideas for plant staff to consider and prioritize
based on benefits to plant operation and performance, level of complexity, and available staff
time.

Ideas for Further Study

Study 1: OptimizingNaMnO4 Dose

• Description:

¦ Conduct j ar testing using NaMnC>4 and determine optimum dose based on oxidant
demand.

¦ Conduct j ar testing using NaMnC>4 and assess impact on coagulation and settling
(repeat of CPE study).

• Potential Benefits:

¦ Determine permanganate demand of raw water.

¦ Better understand the benefits of feeding permanganate.

¦ Assist with decision making when considering turning off permanganate during a
HAB event.

¦ The NaMnC>4 demand part of the study is a relatively simpletopic to learn the study
approach (staff develop the study, it is reviewed by EPA, and staff implement and
share documented results with the state EPA).

-------
• Obstacles and possible solutions:

Obstacles

Solutions

Finding operator time to conduct studies

Assess priority of study
benefits relative to other
studies.

NaMn04 demand study very
doable by plant staff.

Becoming familiar with preparing permanganate stock
solution (use Jar Test spreadsheet)

Obtain training on basic jar
test training (local operators,
AWWA manuals/video).

Developing settling curves to assess impacts of
coagulation/flocculation/sedimentation and pre-oxidation
conditions

Practice the sampling and
testingtechniques by sampling
from two jars prepared
identical to each other,
developing settling curves, and
comparingthe results.

Study 2: Investigating Filter Backwash Capability to Improve Media Expansion and
Stratification

•	Description:

¦	Conduct studies during routine backwashing to increase backwash flow to improve
media expansion.

¦	Assess ability to more slowly ramp down high rate wash to better stratify filter media.

•	Potential Benefits:

¦	Improved media stratification.

¦	Longer filter run time during periods with high solids loading to the filters (e.g., dur-
ing aHAB event).

¦	Understand if any design limitations exist.

¦	Excellent study to learn problem-solving skills.

64

-------
• Obstacles and possible solutions:

Obstacles

Solutions

Finding operator time to conduct studies

Assess priority of study benefits
relative to other studies.

Very doable by plant staff.

Requires a bed expansion measurement tool

Can be constructed by plant staff.

Study 3 — Evaluation of IFE and CFE Sample Pump Operation on Turbidity Spikes

•	Description:

¦	Collect data to determine the maximum daily IFE and CFE turbidity values.

¦	Document turbidity spike occurrences for IFE and CFE samples related to sample
pump operation.

¦	Identify possible solutions to eliminate turbidity spikes related to pumping of
samples.

•	Potential Benefits:

¦	Provide more reliable IFE and CFE turbidity data to assess filter performance.

¦	Reduced operator time addressing sample pumping problems.

¦ Excellent study to learn problem-solving skills.
Obstacles and possible solutions:

Obstacles

Solutions

Finding operator time to conduct studies

Assess priority of study benefits



relative to other studies.



Very doable by plant staff.

Identification of daily maximum IFE and CFE turbidity
values

Determine values from daily
review of SCADA screen and
compare with SCADA data logs.

65

-------
66

-------
4 — Carbon Feed Dose Versus Microcystins Removal

Description:

¦	Conduct j ar testing using current carbon type and assess varying doses versus micro-
cystins removal.

Potential Benefits:

¦	Assist staff with determining how much carbon dose would be needed to treat
through a HAB event (e.g., up to 100 |ig/L extracellular).

¦	Supports full-scale study to determine the ability of the plant to feed a higher carbon
dose.

Obstacles and possible solutions:

Obstacles

Solutions

Finding operator time to conduct studies

Assess priority of study benefits
relative to other studies (one-time
study may be better for others to
conduct?).

Microcystins testing

Send to City of Oregon lab (if they
have capacity) or send to other
certified labs.

Obtaining higher microcystins concentrations for
testing (natural versus spiking)

Concentrate sample from natural
raw water (more doable by plant
staff, less costly).

Spike with standards (costly).

Interpreting results and showing relationships between
the PAC dosage and performance

Use the AWWA PAC evaluator
spreadsheet, available for free
download on the AWWA.org
website. The spreadsheet will
compile the results and develop
dosage curves.

67

-------
Study 5 — Assessing the Impact of Full-Scale Feeding of High Carbon Dose on Plant
Performance

•	Description:

¦	Conduct full-scale study to assess the ability to feed an extended high dose of carbon
(supported by previous jar study, determine dose to achieve microcystins
performance goal).

•	Potential Benefits:

¦	Establish capability of plant to feed high carbon dose and assess the impact on clari-
fier and filter performance and sludge handling capability.

¦	Identify plant design limitations.

¦	Better able to assess O&M sustainability of feeding high carbon dose.

•	Obstacles and possible solutions:

Obstacles

Solutions

Finding operator time to conduct studies

Assess priority of study benefits
relative to other studies.

Very doable by plant staff
(short term).

Cost of study (carbon, extra staffing)

Conduct study during HAB sea-
son when higher carbon doses
are likely beneficial for water
quality.

Potential operation and performance issues

Step up carbon feed rate and
assess O&M and performance
issues. Stop study if impacts
become significant.

68

-------
Study 6 - Assessing NaMnOj Feed on Microcystis Cell Disruption and Cyanotoxin Release

•	Description:

¦	Conduct j ar testing using variable NaMnC>4 doses to assess impact on microcystins
cell disruption and cyanotoxin release (repeat of CPE study).

•	Potential Benefits:

¦	Better understand potential impacts of feeding permanganate on cyanotoxin release.

¦	Determine a permanganate dose (if any) that minimizes cyanotoxin release, such that
treatment focus can be on particulate/cell removal.

¦	Assist with decision-making when considering turning off or reducing permanganate
feed during aHAB event.

•	Obstacles and possible solutions:

Obstacles



Solutions

Finding operator time to conduct studies

Assess priority of study benefits rel-
ative to other studies (one-time
study may be better for others to
conduct?).

Becoming familiar with preparing permanganate stock
solution (use Jar Test spreadsheet)

Obtain training on basic jar test
training (local operators, AWWA
manuals/video).

Microcystin testing

Send to city of Oregon lab (if they
have capacity) or send to other
certified labs.

Obtaining higher microcystin concentrations for
testing (natural versus spiking)

Concentrate sample from natural
raw water (more doable by plant
staff, less costly) or spike with
standards (costly).

Other Plant Studies

• Impact of seasonal adjustments to clarifier sludge blanket level on clarifier performance
(winter versus summer).

69

-------
•	Evaluating use of pre-sedimentation basins for PAC addition (including modifications to
reduce PAC feed plugging).

•	Repeat microcystins plant profile study to assess the ability of the treatment processes to
control microcystins breakthrough. This study would be dependent on laboratory support
for microcystins analyses and would be conducted during a HAB event.

•	Studies to verify the theoretical predictions on microcystins oxidation laid out in this
report, and to document the effect that variables, such as pH, have on oxidation per-
formance. Conduct j ar tests initially to understand the relationships, the results of which
can be used to develop full-scale studies.

70

-------
	Appendix A	

Major Unit Process Evaluation Supporting Calculations

FLOCCULATION PROCESS - Particle (Turbidity) Removal

The flocculation process takes place in the three parallel solids contact clarifiers, in the section of
the clarifier under the floe recirculation cone that extends down and out from the center of the
basins, surrounding the inlet riser pipe. The CPE team used a 30-minute hydraulic detention
time (HDT) as a rating criterion for the flocculation process. Many solids contact clarifiers have
been found to be more efficient at flocculation than conventional flocculation basins, and the 30-
minute HDT parameter typically used to evaluate single stage flocculation basins is conservative
for the more efficient solids contact process.

The volume of the flocculation zone in one of the contact clarifiers is the volume under the floe
recirculation cone minus the volume of the riser pipe. The operators at the ABC WTP usually
maintain a sludge blanket two to four inches above the bottom of the cone. Water passing
through the sludge blanket as it travels around the bottom of the cone is an important part of the
flocculation process, so the CPE team used the entire volume under the cone as the flocculation
volume (minus the riser pipe volume). In the summer, operators sometimes allow the sludge
blanket to be lowered to an elevation that is below the bottom of the floe recirculation cone, but
that type of operation is not recommended for these types of units. The potential remains for the
flocculation process to work more efficiently through the operation of the unit with a sludge
blanket that extends above the bottom of the cone. For each of the basins, the equation that
defines the volume under the floe recirculation zone is:

Volumecone = [ri2+(rir2)2+r22]

Where:

h = the height of the cone (17 feet for each of the basins)

ri = the radius of the top of the cone (6.5 feet in each of the basins)

r2 = the radius of the bottom of the cone (18.25 feet in each of the basins)

71

-------
The volume under the cone is 8793.2 cubic feet or 65,773.5 gallons in each basin.

The volume of the riser pipe in each basin is defined by the equation:

Vo111meriser = ^ (k)d2

Where:

h = the height of the riser (17 feet for each of the basins)
d = the diameter of the riser pipe (6 feet for each of the basins)

The volume of the riser in each of the basins is 480.7 cubic feet or 3595.4 gallons.

The volume of the flocculation zone is:

Volumecone - Vo 111 m eriser

= 65773.5 gallons - 3595.4 gallons = 62178.1 gallons.

Using the 30-minuteHDT criterion, the potential capacity of each of the units is:
Flocculation Capacity = Volumefi0ccuiation/HDT

62178.1 gat 1,440 mini day

=	X	

30 min 1,000,000 gal/MG

= 2.98 MGD per clarifier. The total capacity for all three clarifiers would be 8.95 MGD,
well above the peak instantaneous flow of 6 MGD.

SEDIMENTATION PROCESS - Particle (Turbidity) Removal

The settling processes in the three solids contact clarifiers were evaluated by calculating an
average Surface Overflow Rate (SOR,) which would also represent the average settling velocity
of a floe particle traveling up through the solids contact settling zone. The average SOR was
determined by calculating the settling area at the top of the basin (a larger area because the floe
recirculation cone is smaller at the top of the basin, leaving more area for settling) and the
settling area at the bottom of the recirculation cone and averaging the two areas. The capacity of
the basin is determined by determining the flow through the basin that would result in a SOR of

72

-------
not more than 0.7 gpm/sf, a typical rating value used for solids contact clarifiers greater than
14 feet in depth.

The settling zone area at the top of the clarifier = Area of the clarifier- Area of the top of the
cone.

Where, Area of the clarifier = -D2

'	4

Where, D is the Diameter of the basin (70 feet for each basin)

And, Area of the top of the cone = ^di2

where di is the diameter of the top of the cone (13 feet for each basin)

The settlingzone area at the top of the filter = 3848.45 sf- 132.73 sf = 3715.72 sf.

The settlingzone area at the bottom of the floe recirculation cone =

Area of the clarifier - Area of the bottom of the of the floe recirculation cone.

Where, Area of the bottom of the floe recirculation zone = -d22

'	4

where 62 is the diameter of the bottom of the cone (37 feet for each basin)

The settlingzone area at the bottom of the floe recirculation zone = 3848.45 sf- 1075.21 sf =
2773.24 sf.

The average settling area of the top and the bottom of the floe recirculation cone =

————————————— = 3244.48 sf

2

The rated capacity is calculated by:

Sedimentation Capacity = SOR x average settling area =

~ ^	, r-	r- 1,440 min/day _ __ .

0.7 gpm/sf x 3244.48 sf x	 = 3.27 MGD

1,000,000 MG/Gal

The capacity of the three solids contact units in tandem is 3.27 MGD x 3 = 9.81 MGD, which is
well above the peak instantaneous flow of 6 MGD.

73

-------
FILTRATION PROCESS - Particle (Turbidity) Removal

The ABC WTP filters are dual-media filters, using sand and anthracite as filtration media. The
capacity of the filtration process is calculated using a filterloading rate of 4 gpm/sf of filter area,
assuming one of the filters is out of service, to account for filtration rates when one of the filters is
being backwashed. The water treatment plant has eight filters, all measuring 18 feet by 20 feet.
The capacity of the process is calculated by:

Filtration Capacity = 7 x 18 ftx 20 ftx 4 gpm/sf x 1,440 min^day = 14.52 MGD, well above the

^ 1,000,000 gal IMG '

peak instantaneous flow of 6 MGD.

DISINFECTION PROCESS - Microbial Treatment

Calculation of plant disinfection capability is based on chlorineCT values (i.e., chlorine concen-
tration multiplied by chlorine contact time) outlined in the USEPA Guidance Manual8 for meeting
disinfection requirements for inactivation of 0.5 log (85 percent) of Giardia cysts. (For disinfec-
tion with chlorine, the Giardia inactivation requirement is more stringent than the virus disinfec-
tion requirement.) This assumes that the ABC WTP is well operated and can be credited for 2.5
log (99.7 percent) removal of Giardia cysts through the plant's physical treatment processes. This
can be achieved by meetingthe specified CT required for disinfection with chlorine, as used in the
clearwell at the water treatment plant.

For disinfection in the clearwell, a required CT value of 63.7 mg-min/Lis obtained from the
USEPA Guidance Manual, using a maximum chlorine residual of 3.0 mg/L, a maximum pH of
8.2, and a worst-case temperature of 0 °C for disinfection. These data are obtained from review-
ing the previous year of operating data and through discussions with water treatment plant
operators. The total volume used for the clearwell is 626,133 gallons, based on two ground
storage tanks with 73-foot diameters and a 10-foot minimum operating level in each. A baffling
factor of 0.6 was assigned to each of the clearwells, based on their well-baffled configuration.

USEPA Guidance Manual for Disinfection Profiling and Benchmarking. Appendix E, EPA 815 -R-99-013
(August 1999).

-------
Based on these criteria, the disinfection rating of the clearwell is 21.61 MGD, which is well
above the reported peak instantaneous flow of 6 MGD.

PAC ADSORPTION PROCESS - Microcystins Removal

Removal of extracellular microcystins through PAC adsorption would depend on factors not
related to design considerations such as type of carbon used and dose, use of pre-oxidants (e.g.,
NaMnC>4), competing compounds in the water (e.g., natural organic matter), and contact time.
To accurately estimate a necessary PAC dose and/or most effective type of PAC for a given
water, jar tests and/or full-scale studies should be performed. In the absence of these studies, the
published research studies may be able to provide an indication of the PAC dose needed. The
USEPA Drinking Water Health Advisory for Cyanobacterial Microcystins Toxins suggests:

According to Newcombe et al. (2010), a PAC dose of 20 mg/L and a contact time of at
least 45 minutes should be considered for removal of most extracellular microcystins
(with the exception of microcystin-LA).

Given the absence of jar test data for the water and the PAC that ABC WTP uses, the initial PAC
dose was estimated using isotherm equations and constants, which account for the type of carbon
used that has been referenced by the State EPA in their Draft White Paper on Cyanotoxin
Treatment (August 2015), based on work from Mohamed et al9. However, isotherm data
typically underestimate PAC doses required for full-scale water treatment plant operation.

Hence, a multiplying (safety) factor was applied to estimate a baseline dose. As shown in Figure
21, the equations can predict the PAC dosage that would achieve 90 percent and 99 percent
removal of extracellular microcystins for a range of initial concentrations. In these predictions,
the carbon constantused was for a wood/coal blend PAC - similar to that used by the ABC
WTP. Also shown is the prediction multiplied by safety factors (two and three). Figure 22
suggests that over 90 percent removal could be achieved for the entire range of initial
microcystins concentrations at a PAC dosage less than 30 mg/L, even with an applied safety
factor of "3. " However, the equations predict that is it unlikely that 99 percent removal of

9 Mohamed, Z. A., W.W. Carmichael, J. An, and H.M. El-Sharouny, "Activated Carbon Removal Efficiency of
Microcystins in an Aqueous Cell Extract of Microcystis aeruginosa and Oscillatoria tenuis Strains Isolatedfrom
Egyptian Freshwaters," Env. Toxicol., 14(5), 197-201 (1999).

-------
microcystins could be achieved if the initial concentration was greater than 50 |ig/L, for a safety
factor of "2. " Also, in Figure 21, jar test data from a nearby Lake western basin system test,
conducted in the winter of 2015-2016, were added as a point of reference. Those data show
consistent removal of microcystins in the 85 to 90 percent range at a 5 mg/L PAC dosage,
regardless of initial concentration. The uncertainty in this analysis highlights the importance and
need for evaluating the PAC used by the ABC WTP at a range of raw water microcystins
concentrations, (the State recommends that systems in the western basin of this Lake be
prepared to treat 100 |ig/L.)

The contact time at the ABC WTP should be adequate for PAC adsorption of microcystins, since
the current feed location in the rapid mix would allow up to two hours of contact in the clarifier
plus a small amount of additional contact in the rapid mix basin. However, the presedimentation
basin may also need to be used for additional PAC contact time if a high PAC dose is required
due to operational constraints (i.e., challenges related to delivering a consistent high dose to one
location, difficulty keeping the PAC from immediately settling out, etc.). The team estimated the
maximum PAC feed rate possible to be 30 mg/L, assumingthe following:

• One feeder is out of service (thus, the one feeder would deliver the entire dose of
30 mg/L).

• The feed system is consistently able to deliver this thick slurry (30 mg/L dose) to a com-
bination of the presedimentation or rapid mix basin.

• An estimate of how much PAC can be stored onsite.

• An assessment of the physical ability of the operators to feed the 50-lb sacks of PAC into
the dry chemical hopper.

It is possible that the plant could feed more PAC, but this would need to be assessed through
sustained, full-scale operation.

-------
100

90

SO

70

60

50

40

30

20

10

0

Isotherm Predictions for Wood/Coal Powdered Activated Carbon (PAC) Blend
Dosage vs. Initial Microcystes Concentration to Achieve 90% and 99% Removal

• Predicted Dose @ 90% Removal
¦ Predicted X 3 @ 90%
x Predicted X 2 @ 99%

a Predicted X 2 @ 90%
~ Predicted Dose @ 99% Removal
+ Predicted X 3 @ 99%

	•

	•

	•

	•	

	•	

	•	

	•	

• J

	•	A

~	~

~
~
~

~

~
~
~

~
~
~

J

A
A
A

X

~
~



—i	1	1	r~

10

+

15	20	25	30	35	40

Wood/Coal PAC Blend Predicted Dose (mg/L)

45

50

55

60

FIGURE 22. Predicted PAC Dose Based on Removal Efficiency and Initial Microcystin Concentration.

77

-------
Based on the available tools and the data from a nearby system, the team estimated that a maxi-
mum feed rate at the water treatment plant of 30 mg/L would likely remove over 90 percent of
extracellular microcystins. ThePAC feed capacity of the plant currently is about 172 lb/hr,
based on a feed rate of 134 lb/hr through the larger auger feeder and 38 lb/hr through the smaller
auger. The water treatment plant operators report that they have a third, spare auger feeder that
is in between the small and large size, but it has not yet been tested to determine its feed rate.
Using the current plant PAC feed design, the feed rate is evaluated based on the largest auger
feeder being out of service and replaced by the medium feeder and an estimate that the medium-
sized auger feeder can feed at a rate of about 70 lb/hr.

At 70 lb/hr feed, and a 30 mg/L concentration, the plant capacity is:

FR f-jM* 453,000 (^)

Capacity =	

x 3.7854 (—> 1,000,000 (Jj)

Where:

FR = chemical feed rate (70 lb/hr x 24 hrs = 1,680 lb/day)

Dose = PAC dosage (30 mg/L)

Plant capability = 6.7 MGD, greater than the peak instantaneous flow of 6 MGD

With minor plant improvements, PAC could be fed simultaneously to the presedimentation basin
and the rapid mix basin using two of the PAC feeders simultaneously. In this way, the PAC feed
rate could be lowered to facilitate slurry travel and dosage optimization. (Water treatment plant
operators may find, through studies, that increasing the PAC feed to the presedimentati on basin
and then adding a lesser amount at the rapid mix is more effective.) Assuming two feeders are
used, the combined feed rate might be 100 lb/hr, if the larger auger is out of service and the
smaller auger is used with the medium-sized auger. Still targeting 30 mg/L total concentration,
the capacity would be calculated to be the same, except that the feed rate would be 100 lb/hr
(2,400 lb/day) and the plant capacity would be 9.59 MGD.

78

-------
While the PAC feed would be rated a Type 1 process based on the sizing of the feeders, concerns
would include: the contact time, the ability to feed 30 mg/L with the existing facilities, and the
sustainability of adding such a high dosage (approximately thirty 50-lb sacks of PAC into the
hopper per day) for a sustained period safely. An additional concern would be the ability of the
plant to deliver the slurry to the rapid mix or pre-sedimentation basin at such a high concentra-
tion without maintenance problems, such as clogged delivery lines. For this reason, the process
has been downgraded to a Type 2 process. In other words, diligent and careful operation would
be needed in order for this process to achieve the specified goals (removal of 90 percent of extra-
cellular microcystins).

OXIDATION PROCESS - Microcystins Destruction

Chlorine oxidation can destroy any remaining extracellular microcystins that would not be
adsorbed onto the PAC earlier in the treatment processes. Predicting the capacity of the plant to
oxidize certain levels of microcystins can be determined using the AWWA Hazen Adams
CyanoTOX (1.0) calculator spreadsheet. However, the oxidation rate is highly dependent on the
pH of the water, and pH during a HAB event could be much higher than the pH observed at the
plant under non-HAB conditions. Figure 23 shows a graph of the oxidation capacity of chlorine
(from 10 |ig/L to 0.3 |ig/L, or 97 percent removal) at different flow rates and pH levels, assum-
ing the chlorine concentration was increased to 3 mg/L and 4 mg/L (both of which are above the
current operating range of the plant). The target of 0.3 |ig/L was chosen because it is identified
by the State EPA as the level above which a Do Not Drink advisory must be issued. During a
HAB event, an anticipated maximum pH is estimated at 8.8 (based on a review of pH data from
twelve previous months at the ABC WTP and experiences with HAB events at other locations),
which would limitthe plant to 6.85 MGD capacity at 4 mg/L chlorine if a concentration of 10
|ig/L of extracellular microcystins were enteringthe clearwells.

79

-------
18.00
16.00
14.00
12.00
10.00

Q

8.00
6.00 ¦



Microcystins Oxidation Capacity (10 to 0.3 ug/L at 20 C)













































-•-p

'1 ant Capacity at 3 mg/L



















•

lant Capacity at 4 mg/L



















































Peak Instantaneous Flow (6 MGD)











































4.00
2.00
0.00

i

















































3 8.1 8

.2 8.

.3 8

.4 8,

.5 8.

PH

.6 8,

00
00

00

.9 9 9.1

FIGURE 23. Oxidation Capacity Based on 97 Percent Removal Using AWWA Calculator.

80

-------
Results of the
Harmful Algal Bloom
Comprehensive Performance Evaluation
for the

ABC Water Treatment Plant
Anytown, State

January 23 - 27,2017

Prepared By:

Process Applications, Inc.
2627 Redwing Road, Suite 340
Fort Collins, Colorado 80526

USEPA Technical Support Center
26 West Martin Luther King Drive
Cincinnati, Ohio 45268

State Environmental Protection Agency

81

-------
Table of Contents

SITE VISIT INFORMATION	5

INTRODUCTION	7

DESCRIPTION OF WATER TREATMENT PLANT	8

Overview	8

Source Intake and Pump Station	9

Water Treatment Processes	9

PERFORMANCE ASSESSMENT	12

Historical Performance Assessment	12

Administration Assessment	13

Historical Water Quality Performance Assessment: Turbidity	13

Historical Performance Summary	19

Additional Performance Observations	20

Disinfection	25

Cyanotoxins	26

Studies	28

MAJOR UNIT PROCESS EVALUATION	51

Particle Removal and Microbial Disinfection	52

Cyanotoxin Removal and Destruction Treatment	54

PERFORMANCE-LIMITING I AC TORS	58

Policies - Administration (A)	59

Application of Concepts and Testing to Process Control (Operations) (A)	59

Operational Guidelines/Procedures (Operations) (A)	60

Staffing/Number (Administration) (B)	61

Process Controllability (Design) (B)	61

Alarm Systems (Design) (B*)	61

Sample Tap (Design (C)	61

EVALUATION FOLLOW-UP	62

Study Ideas	62

82

-------
List of Figures

FIGURE 1. Schematic Removed	10

FIGURE 2. Typical filter layout in relation to common filter wet well	12

FIGURE 3. Maximum daily sedimentation Basin 2 effluent turbidity	16

FIGURE 4. XYZ WTP turbidity profile	17

FIGURE 5. Maximum daily filtered water turbidity (IFE only)	18

FIGURE 6. Sedimentation Basin 1 turbidity 2016	20

FIGURE 7. Sedimentation Basin 2 turbidity 2016	21

FIGURE 8. Top of filter turbidity 2016	21

FIGURE 9. Filter 4 effluent turbidity profile showing spikes at plant startup	22

FIGURE 10. Filter 1 effluent turbidity profile showing post-backwash spike	23

FIGURE 11. Top of filter turbidity profile versus IFE turbidity for each filter	24

FIGURE 12. Seasonal settled water turbidity versus IFE turbidity for each filter	25

FIGURE 13. Daily disinfection inactivation ratio	26

FIGURE 14. Raw water microcystins concentrations at XYZ WTP, 2014 - 2016	27

FIGURE 15. XYZ WTP microcystins profile on May 9, 2015	28

FIGURE 16. Filter 4 plan view	30

FIGURE 17. Filter 4 probing and excavation locations	31

FIGURE 18. Pictures of mixed media found in Filter 4 bed during excavation	31

FIGURE 19. Filter bed expansion tool	33

FIGURE 20. Filter 4 waste backwash water turbidity profile	34

FIGURE 21. Return-to-service profile for inspected Filter 4	36

FIGURE 22. Online turbidimeters with SC200 controllers on the operating floor of the

filter gallery	37

FIGURE 23. Online turbidimeter flow check and sample line detention time	38

83

-------
FIGURE 24. Jar testing to simulate impact of variable PAC dosages and contact times on

microcystins concentrations	44

FIGURE 25. Velocity gradient versus mixer speed in 2-liter square jar	45

FIGURE 26. Graph showing jar test results for microcystin concentration versus

PAC dose and time	48

FIGURE 27. Bar chart showing jar test results for microcystin concentration versus

PAC dose and time	49

FIGURE 28. Pictures of phytoplankton from control sample	50

FIGURE 29. Major Unit Process Evaluation - XYZ WTP turbidity removal

(microbes, cells) and disinfection	53

FIGURE 30. Major Unit Process Evaluation- XYZ WTP microcystins

adsorption and destruction	55

List of Tables

TABLE 1. CPE Turbidity Performance Analysis; Data Acquisition Description	14

TABLE 2. OAS Summary Statistics	15

TABLE 3. OAS Optimization Trend - Filtered Water	19

TABLE 4. XYZ WTP Performance Summary	19

TABLE 5. Data Integrity Study: Turbidimeter Settings	40

TABLE 6. Summary of Chemical Feeder Calibration and Dose Results	41

TABLE 7. Summary of Jar Test Settings to Replicate Raw Water Transmission Line	44

TABLE 8. Jar Test PAC Dosing Regimen	46

TABLE 9. Major Unit Process Summary	58

84

-------
SITE VISIT INFORMATION
Site and Mailing Address:

Removed

Date of Site Visit:

January 23 - 27, 2017

ABC Water Treatment Plant Personnel Participating:

Mayor

Administrator
Fiscal Officer
Water Billing Clerk

Superintendent

Class 3 Operator (Assist. Supt.)

Class 3 Operator
Operator and Meter Reader

85

-------
CPE Team:

USEPA Technical Support Center, 26 West Martin Luther King Drive, Cincinnati, OH 45268

Alison Dugan - 513-569-7122; Dugan.Alison@epa.gov

Rick Lieberman - 513-569-7604; Lieberman.Richard@epa.gov

Tom Waters -513-569-7611; Waters.Tom@epa.gov

USEPA Office of Research & Development, 26 West Martin Luther King Drive, Cincinnati, OH 45268
Craig Patterson - 513-487-2805; Patterson.Craig@epa.gov

Process Applications, Lnc., 2627Redwing Road, Suite 340, Fort Collins, CO 80526
Bill Davis - 469-338-1823; waterbilldavis@gmail.com
Larry DeMers -970-223-5787; ldemersco@aol.com

State Environmental Protection Agency

HAB Engineer

HAB Coordinator

Design Engineer

Field Engineers/Staff/Inspectors

86

-------
INTRODUCTION

The Composite Correction Program (CCP)10 is an approach developed by the U. S. Environmen-
tal Protection Agency (USEPA) and Process Applications, Inc. (PAI) to improve surface water
treatment plant performance and to achieve compliance with the Surface Water Treatment Rule
(SWTR). Its development was initiatedby PAI and the state of Montana11, which identified the
need for a program to address performance problems at its surface water treatment plants. The
approach consists of two components, the Comprehensive Performance Evaluation (CPE) and
the Comprehensive Technical Assistance (CTA).

Producing a finished water with a turbidity of less than or equal to 0.10 NTU provides much
greater protection against pathogens like Cryptosporidium. This microorganism that passed
through the public water supply was responsible for a large outbreak of Cryptosporidiosis in

10 Hegg, B.A., L.D. DeMers, J.H. Bender, E.M. Bissonette, and R.J. Lieberman, Handbook - Optimizing Water
Treatment Plant Performance Using the Composite Correction Program, EPA 625/6-91/027, USEPA, Washington,
D C. (August 1998).

11 Renner, R.C., B. A. Hegg, and D.F. Fraser, Demonstration of the Comprehensive Performance Evaluation
Technique to Assess Montana Surface Water Treatment Plants, Association of State Drinking Water
Administration Conference, Tucson, AZ (February 1989).

-------
Milwaukee, Wisconsin in April 1993, where 400,000 people became ill and nearly 100 deaths
occurred. Cryptosporidium cysts are extremely resistant to chlorine disinfection, necessitating
optimization of physical removal of particles.

Since the development of the CCP for optimization of surface water treatment plants for protec-
tion from microbial pathogens, PAI and the USEPA's Technical Support Center (TSC) have
adapted the CCP protocol to additional public health parameters such as DBP control and distri-
bution system optimization. Given the recent concerns with harmful algal blooms (HABs) and
their impact on surface water treatment plants in the State and nationwide, the State EPA, in part-
nership with TSC, has initiated a proj ect to expand the CCP to include optimization for the
removal of cyanobacteria cells and the reduction of associated toxins. This CPE for the XYZ
Water Treatment Plant (WTP) represents the second of four developmental CPEs focused on
these performance goals that will be conducted in this State.

The following report presents the findings from this CPE, and it will hopefully provide the XYZ
Water Department with valuable information that can be used to enhance and maintain water
quality. The CPE team would like to thank the plant staff and utility management for hosting
this event and for taking the timeto assist the team in completingthe evaluation. During the
evaluation, utility staff members were very accommodating in providing plant information and
sharing their experience and knowledge regarding treatment approaches to address HAB events.
This type of attitude represents a strong foundation for development of an optimization approach
to address HAB events that public water systems may face in the future. This report documents
the findings of the CPE.

DESCRIPTION OF WATER TREATMENT PLANT
Overview

The XYZ WTP is the main source of potable water for the City, providing treated drinking water
from a nearby lake. The XYZ Water System also has interconnections with a County Water and
Sewer Organization to purchase potable water on an emergency basis, but it does not have a
contract defining how much can be purchased in case of emergency. XYZ also provides finished

-------
water to a nearby Public Water System. Potable water is delivered to approximately 3,857 direct
consumers, including water purchased by the neighboring communities.

Source Intake and Pump Station

A schematic of the water treatment plant (removed from Figure 1). The source water is supplied
to the plant by two intake structures located ten miles away from the water treatment plant at a
Lake. There is an existing backup intake located at a Reservoir, which is currently not approved
for use. The two Lake intake structures are located roughly 10 to 13 feet offshore, at a depth of
about eight feet, and the intake pumping station building is located on the bank of the Lake adja-
cent to the intake structures.

Three raw water pumps within the pump station (with one in use) transport the raw water ten
miles to the water treatment plant. The plant has facilities to feed potassium permanganate and
powdered activated carbon (PAC) at the raw water intake pump station; however, potassium per-
manganate was not being fed during the CPE. A raw water sample line located in the pump sta-
tion is used to collect raw water from the wet well prior to PAC addition. However, this must be
done manually; hence, water quality for process monitoring (e.g., turbidity) is collected after
PAC addition.

Water Treatment Processes

The XYZ WTP utilizes conventional surface water treatment processes, including: aeration,
coagulation, flocculation, sedimentation, filtration, and disinfection. The reported plant capacity
is 1.1 MGD.

Preceding the conventional treatment, the raw water is dosed with PAC from a hopper auger unit
located in the pump station at the raw water source. The water is then sent through roughly nine
miles of 12-inch diameter pipe, followed by one mile of eight-inch pipe to the aeration unit at the
water treatment facility.

-------
FIGURE 3. Schematic Removed

90

-------
After the raw water passes through the aeration unit, the water is sent to the first rapid mix where
lime and soda ash are added with an alum/polymer blend coagulant. There is a potassium per-
manganate feed point in the first rapid mix; however, it was not in use at the time of the CPE site
visit. Water leaving the first rapid mix is dosed again with PAC. A chlorine addition feed point
at this location is currently not used.

Flocculation and sedimentation are accompli shed using two flocculation and sedimentation tanks
operated in series. The two-stage processes are separated by a second rapid mix with no mixer in
operation, and aluminum sulfate is added at this location to further aid coagulation. The sludge
from the sedimentation tanks is removed manually and sent to the onsite sludge holding tank.
Each sedimentation basin effluent has a sampling point where a grab sample is taken to measure
turbidity and pH.

From the last sedimentation tank in series, water bypasses the non-operational recarbonation
basin, and flow is distributed by an inlet weir to four dual-media anthracite/sand filters. Each fil-
ter effluent is continuously sampled via sample streams that feed turbidimeters located on the
operating floor. The backwash supply is provided by the filters in service and supplemented by
the high service pump discharge. Figure 2 shows how a higher water elevation in the wet well
(compared with the filter being backwashed) is used to provide the backwash supply. Air is also
inj ected during the backwash. Because of the common filter wet well design, the filters do not
have filter-to-waste capability.

Filtered water flows to a common transfer wet well, where the combined filter effluent turbidity
is sampled and directed to a continuous turbidimeter. Sodium hypochlorite and fluoride are
inj ected in the discharge of the wet well before filtered water is sent to the clearwells. Each
clearwell holds a volume of 164,500 gallons of water. Both are constructed of concrete and are
subterranean in design. The clearwells are baffled, operate in parallel, and are utilized for disin-
fection contact time. Treated water is pumped to the distribution system from a 30-inch suction
line from the clearwells by way of three horizontal, centrifugal high service pumps. An addi-
tional feed point for chlorine also exists on the high-pressure discharge line manifold to boost
levels after the clearwells, if needed. A sampling location after the high service pumps is used to

-------
measure pH and chlorine residual of the finished water as well as to take other compliance
samples.

Variable Height Water Elevation

Backwash Trough

Anthracite

Sand

Under Drain

Filter

Supplemental Backwash Line From
High Service Pump

r.

Overflow to Clearwell
	>

Two Transfer Lines to Clearwell

Wet Well

FIGURE 4. Typical filter layout in relation to common filter wet well.

PERFORMANCE ASSESSMENT
Historical Performance Assessment

Optimized performance, for the purposes of this CPE report, represents performance beyond the
Surface Water Treatment Rule (SWTR) requirements. To achieve optimized performance, a
water treatment plant must demonstrate that it can take a raw water source of variable quality and
produce consistent, high quality finished water. In addition, the performance of each treatment
unit process must demonstrate its capability to act as a barrier against the passage of particles at
all times.

92

-------
Administration Assessment

An assessment of the administration of the XYZ WTP and its possible effect on plant
performance was performed by collecting information through interviews in the following gen-
eral areas: utility structure, vision, mission, water quality goals, reporting, data review, manage-
ment style, communications, planning, plant coverage, financial management, and spending
authority. Two possible administrative issues were identified that could potentially affect perfor-
mance. These issues, as well as others, are considered in subsequent sections of the report:

• Formal adoption of optimization turbidity goals for unit process performance.

• Staffing issues that may affect performance and innovation within the water treatment plant
processes.

Historical Water Quality Performance Assessment: Turbidity

Historical turbidity data were collected from laboratory data sheets and from SCADA files pro-
vided by the Operator of Record. These reports were provided to the CPE team, and the data
were entered in an Optimization Assessment Spreadsheet (OAS) that was used to assess perfor-
mance against the optimization goals for turbidity.

Historical performance was assessed over a 12-month period, starting on January 1, 2016 and
ending on December 31, 2016. Table 1 describes in more detail the exact source of the data used
in the CPE performance assessment. As indicated in Table 1, the plant raw water sampling site
is not a "true " raw water since the sampling location is downstream of the PAC addition at the
intake. Only the effluent of sedimentation Basin 2 was used to assess performance of the sedi-
mentation treatment process.

-------
TABLE 2. CPE Turbidity Performance Analysis; Data Acquisition Description

Performance Parameter

Data Used in the CPE Performance Analysis

Maximum Daily Raw Water
Turbidity Entering the Plant

Maximum daily plant raw water turbidity data were determined from
data recorded on paper laboratory bench sheets. The plant raw water
turbidity values are obtained by benchtop turbidimeter measurements of
grab samples from the plant raw water tap at the laboratory sink. The
plant raw water sample tap is upstream of rapid mixTank 1 and
approximately ten miles downstream of the intake chamber, where PAC
is being fed. Therefore, the plant raw water samples contain PAC.

Individual Sedimentation Basin
Effluent, or "Settled Water"
Turbidity

Operator logs of settled water turbidity test results were used to
determine the maximum turbidity value for each day, using the effluent
of each sedimentation basin and top ofthe filter boxwhich is above the
filter media. These values were used for comparison, but the daily
maximum individual sedimentation effluent values obtained from
sedimentation Basin 2 were used to assess performance against the
optimization goals.

IFE Turbidity

The individual filter effluent (IFE) daily maximum turbidity records were
obtained in electronic CSV format from the SCADA system. The system
obtains continuous turbidity readings from individual Hach 1720E turbi-
dimeters and stores one value from each individual filter turbidimeter
every 15 minutes into an electronic Microsoft Excel file. Each of these
Excel files are named with aT followed by the four-digit calendar year and
one or two digits for each month. Each file contains the title, MONTHLY
FILE - FILTER TURBIDITY 15 MINUTE SAMPLES and stores 15-minute
readings for each ofthe four filters for each day of the month.

CFE Turbidity

The combined filter effluent (CFE) sampling location is not being used for
compliance nor is it actively recording data. Therefore, CFE turbidity per-
formance was not assessed.

Maximum daily raw, settled, and filtered turbidity data were entered in an Optimization Assess-
ment Spreadsheet (OAS). These data were analyzed through the spreadsheet calculations and
charts which compare historical plant performance to optimization goals. Settled water turbidity
values were obtained from plant bench sheets, while filtered turbidity was collected from the
SCADA system. Individual Filter Effluent (IFE) turbidity data is collected every 15 minutes by
the SCADA system during plant production. From this SCADA database, maximum daily tur-
bidity values were determined for each filter. Table 2 shows the OAS summary statisticsfor the
plant, and Figures 3 and 4 display turbidity profiles that are graphical descriptions of water treat-
ment plant performance over the past year (2016).

-------
TABLE 3. OAS Summary Statistics

ANNUAL DATA

Avg

Min

Max

RSQ

95%

Opt. Goal

Reg.

NTU

% Values

Raw Turbidity

12.5

2.3

115.0

n/a

34.4

n/a

Max. Settled Turbidity

3.3

1.1

10.6

0.21

6.2

n/a

Max. Filtered Turbidity

0.31

0.04

1.29

0.06

0.68

n/a

RSQ = Correlation Coefficient for two selected data sets (> ~ 0.25 suggests correlation)

95% = 95th Percentile value for data set

Opt. Goal = % of values in data set that are less than or equal to the selected optimization turbidity goal

Reg. = % of values in data set that are less than or equal to the regulated turbidity requirement

The statistics in Table 2 are based on the maximum daily turbidity values for raw water, sedi-
mentation Basin 2 effluent, and IFE turbidity during the period of January 1, 2016 to
December 31, 2016. These statistics are then compared to optimization goals. The optimization
program utilizes the "maximum " daily turbidity readings to assess worst-case performance by
each of the barriers. If the plant can perform within the optimization goals at the time of its
worst daily performance, then the plant staff can be assured that the plant is maximizing its abil-
ity to protect public health against the passage of pathogens and cyanobacteria. Table 2 shows
that the daily maximum raw water turbidity average for the XYZ WTP was 12.5 NTU. For raw
water conditions where the average maximum daily raw water turbidity is greater than 10 NTU,
the optimization goal for settled water turbidity is 2 NTU. The optimization goal for individual
filter effluent and combined filter effluent turbidity is 0.10 NTU.

The maximum daily settled water effluent turbidity, as measured with grab samples from the
effluent of Basin 2, met the optimization goal 21 percent of the time. The maximum settled
water effluent turbidity was 6.2 NTU or lower 95 percent of the days during the evaluation
period. Figure 3 offers a closer look at the settled water turbidity, as measured with grab sam-
ples from the effluent of Basin 2. The red line in the graph represents 2.0 NTU, the optimization
goal for settled water turbidity. It is apparent that sedimentation performance appeared to be the
best during the late summer and autumn months, as compared to the winter and spring months.
In addition, sedimentation Basin 2 performance appeared to generally meet the optimization goal
during the late summer and autumn months of 2016, but it did not approach the goal during other
parts of the year.

-------
Maximum Daily Settled Water Turbidity

—Max Sed —Goal
12.0 n	

10.0

8.0

0.0 -I	'	1	1	'	1	1	1	1	1	1	1	

Jan-16 Feb-16 Mar-16 Apr-16 May-16 Jun-16 Jul-16 Aug-16 Sep-16 Oct-16 Nov-16 Dec-16

FIGURE 5. Maximum daily sedimentation Basin 2 effluent turbidity.

Table 2 also shows that the maximum daily IFE turbidity values met the optimization goal
10 percent of the days analyzed. The maximum IFE values were at 0.68 NTU or less during
95 percent of the days analyzed.

The R-squared (RSQ) value in Table 2 above represents the correlation coefficient for two
selected data sets. The lower the values, the less carry-through between treatment barriers and
the greater each barrier's efficiency. A value greater than 0.25 suggests a correlation. The raw
turbidity and maximum settled turbidity data sets have an RSQ value of 0.21, which suggests
some carry-over of turbidity from the raw water through the sedimentation process. However,
the RSQ value between the maximum settled turbidity and maximum filtered turbidity is much
lower, and it reveals that filter performance is not closely associated with sedimentation basin
performance.

96

-------
Figure 4 provides a turbidity profile of plant performance over a 12-month period. The raw
water turbidity, depicted by the solid red line, seems to exhibit an increase in the spring and drop
during the summer months, with spikes in June and October. The settled water data represented
by the solid black line generally trends with the raw water turbidity, especially at the end of
2016. A visual observation of the IFE turbidity data suggestsa slight inverse relationship with
the settled water turbidity data, where the filters appear to have the worst performance over the
summer months, even though the settled water turbidity is the lowest during this period. Again,
the combined, solid blue line is not represented in Figure 4 because the XYZ plant does not col-
lect data nor use the CFE sampling point.

Turbidity Profile

	Raw 	Max Sed 	Max Filter 	Combined

1000

FIGURE 6. XYZ WTP turbidity profile.

Figure 5 depicts the maximum daily filtered water turbidity for IFE measurements in relation to
the optimization goal of 0.10 NTU, represented by the red line. The graph shows the maximum
IFE turbidity measurements (dashed blue lines) mostly above the optimization goal throughout
the last year, with significant spikes up to 1.0 NTU at times. A seasonal influence on perfor-
mance occurs during the late summer and autumn time frame.

97

-------
Maximum Daily Filtered Water Turbidity

	Max Filter	Goal 	Combined

1.00 -
0.90 -
0.80 -
0.70 -

? 0.60 -

z

£ 0.50 -

"D

£ 0.40	-

0.30	-

0.20	-

0.10	-

0.00	-

Jan-16 Feb-16 Mar-16 Apr-16 May-16 Jun-16 Jul-16 Aug-16 Sep-16 Oct-16 Nov-16 Dec-16

FIGURE 7. Maximum daily filtered water turbidity (IFE only).

The data used to develop the profile in Figure 5 are the maximum daily IFE values from all four
filters. However, Table 3 offers a closer look at the individual filtered water turbidity data. The
OAS also plots the performance of each filter based on IFE turbidity values, and it is one way of
checking to see if certain filters are performing better than others. These data are summarized in
Table 3, and they reveal the worst performing filter is Filter 4, meeting the 0.10 NTU optimiza-
tion goal 30.1 percent of the time. Filter 2 experiences similar performance to Filter 4, followed
by Filter 1 and Filter 3 in order of increasing efficiency. It is important to note that Filter 3 con-
tains the oldest media, yet it met the optimization goal more than the other filters with a value of
53.3 percent of the timein2016.

98

-------
TABLE 4. OAS Optimization Trend-Filtered Water



Filtered Water Turbidity



95th Percentile Values (NTU)

% Values Meeting Goal All Filters



Filter 1

Filter 2

Filter 3

Filter 4

All Filters

0.3

0.2

0.1

Jan-16

0,44

0.49

0.43

0.46

0.45

87.10

79.84

46.0

Feb-16

0.25

0.33

0.54

0.36

0.41

90.52

82.76

54.3

Mar-16

0.21

0.19

0.23

0.17

0.21

98.39

92.74

77.4

Apr-16

0.33

0.33

0.30

0.43

0.37

87.50

71.67

40.8

May-16

0.31

0.20

0.44

0.24

0.32

94.35

91.13

62.9

Jun-16

0.19

0.26

0.25

0.46

0.28

95.83

82.50

50.0

Jul-16

0.40

0.49

0.39

0.87

0.53

79.84

56.45

5.6

Aug-16

0.32

0.47

0.51

0.39

0.47

79.03

60.48

3.2

Sep-16

0.64

0.78

0.61

0.67

0.68

63.33

44.17

0.8

Oct-16

0.33

0.52

0.52

0.44

0.49

82.26

74.19

48.4

Nov-16

0.30

0.43

0.38

0.35

0.38

86.67

75.00

56.7

Dec-16

0.39

0.37

0.44

0.84

0.52

81.45

72.58

48.4

Yr. 95%

0.44

0.51

0.47

0.54









Yr. Goal

46.2%

35.2%

53.3%

30.1%









Historical Performance Summary

The XYZ WTP performance is summarized in Table 4.

TABLE 5. XYZ WTP Performance Summary

Barrier

Optimization Goal

Performance

Clarification

Settled water turbidity
2.0 NTU or less 95 per-
cent of the time, based
on daily maximum
values

The goal was assessed against sedimentation Basin 2 effluent
turbidity values. The 95th percentile of the maximum daily individual
clarifier effluent turbidity was above the goal, at 6.2 NTU, for the
year analyzed. The plant met the 2.0 NTU goal on 21 percent of the
days during the year.

Filtration

IFE and CFE turbidities
0.10 NTU or less 95 per-
cent of the time, based
on daily maximum
values

The IFE data show performance meeting the optimization goal
10 percent of days analyzed during the year, with an annual 95th
percentile of 0.68 NTU. The performance of the plant, based on IFE
data, fails to meet the filtered water optimization goal.

99

-------
Additional Performance Observations

Sedimentation

For settled water turbidity, data were analyzed for each day of calendar year 2016 from the plant
bench sheets, and the maximum values from each day were collected. Turbidity data were col-
lected from sedimentation Basin 1, sedimentation Basin 2, and the "top offilter " (TOF) location,
depicted in Figures 6, 7, and 8 respectively. Figure 6 reveals relatively high turbidity values
leaving sedimentation Basin 1 that are comparable to raw water turbidities and, thus, indicate
little removal across sedimentation Basin 1. Sedimentation Basin 2 provides the bulk of the tur-
bidity removal, as shown by Figure 7. The TOF samples are another way of measuring settled
water performance and should be essentially the same as the sedimentation Basin 2 effluent sam-
ples. Slightly elevated turbidities were observed from the TOF with higher values in Figure 8
than in Figure 7, even though the profiles were very similar. These elevated turbidity values
could be due to the sampling method followed at the TOF location. The evaluation team could
have used either the TOF data or the values from sedimentation Basin 2. They chose the latter in
order to develop the performance summaries reported in the previous section to represent the
overall turbidity reduction from the sedimentation process.

Sedimentation Basin 1

10.0

5- 8.0

£ 2.0

COCOCOCOtOCOCDCOCOCOCOCO

Q.
<

Cl
0
(/)

Q
Q

FIGURE 8. Sedimentation Basin 1 turbidity 2016.

100

-------
Sedimentation Basin 2

10.0

CO

>
o
z

0

Q

FIGURE 9. Sedimentation Basin 2 turbidity 2016.

Top of Filter

10.0

FIGURE 10. Top of filter turbidity 2016.

Filtration

On several days, the maximum daily filter turbidity occurred in the morning at plant startup.
Figure 9 below is a graph of the Filter 4 daily turbidity data for four days in January 2016 that

101

-------
support this observation. The absence of filter-to-waste capability means that these turbidity
spikes are being measured on water going to the clearwell. Filter performance following these
plant start-up turbidity spikes often meets the optimization goal of < 0.10 NTU for the remainder
of each day.

Max of TURBIDITY
0.14
0.12
0.10

P 0.08

z

£

3 0.06

H

0.04
0.02



0.00

:45 :45 :45 :45 :45 :45 :45 :45 :45 :45 :45 :01 :01 :01 :01 :01 :01 :01 :01 :01 :01 :45 :45 :45 :45 :45 :45 :45 :45 :45 :45 :45 :49 :49 :49 :49 :49 :49 :49 :49 :49 :49

2 Art! AM) AVffl AK3 PMPM PM PM PNJ PUPTB Alfl AM) AM AH PIMPl\2 PM PM PNJ PIVE AM AU AN AM PIMPI\2 Pl\3 PM PM PBIPM AI0 AM) AMI AM PIMPI\2 PM PM PBIPM

AM I I I I I I I I I I AM	AM	AM

6-Jan	7-Jan	8-Jan	9-Jan

Axis Title

Days Hours * Tlme2 »

FIGURE 11. Filter 4 effluent turbidity profile showing spikes at plant startup.

The data indicate that there were some days when a turbidity spike at plant startup did not occur
at all or the spike occurred later during the day. Figure 10 shows a two-day turbidity profile of
Filter 1. On January 10th, a spike occurs later in the day, and on January 11th there are no spikes
or much smaller ones. At the time of the spike on January 10th, there is about an hour of missing
turbidity data, suggesting that this time corresponds to a filter backwash during which time the
15-minute turbidity data are not being captured in the SCADA monthly turbidity files. Similar
observations were documented for Filters 2, 3, and 4 as well. For each observation, an hour or
more of data is missing prior to the spike, and the following day shows few to no spikes for that

102

-------
individual filter. Therefore, the largest turbidity spike of the day can be associated with plant
startup or filter recovery following a backwash event.

Max of TURBIDITY
0.25

0.05

0.00

:45 :15 :45 :15 :45 :15 :45 :15 :45 :15 :45 :15 :45 :15 :45 :15 :45 :15 :45 :15 :45 :15 :45 :15 :45 :15 :45 :15 :45 :15 :45 :15 :45 :15 :45 :15 :45 :15 :45 :15 :45
7AM8AM 9 AM 10 AM 11AM 12 PM 1PM 2 PM 3 PM 4 PM 5 PIW AM8 AM 9 AM 10 AM 11AM 12 PM 1PM 2 PM 3 PM 4 PM 5 PM
1/10/2016	1/11/2016

Date Time ~ DateTime ~

FIGURE 12. Filter 1 effluent turbidity profile showing post-backwash spike.

It is difficult to understand the magnitude and duration of the post-backwash turbidity spikes,
since the reported backwash time and/or amount of missing data for an individual filter varies.
This corresponds with different operators overseeing the backwash and potentially clicking the
IFE "start" button on the SCADA turbidity collection at different times. Therefore, the exact
time period when data were captured by the SCADA system during the filter ripening period fol-
lowing backwash is unknown.

Filter performance appeared to be inversely proportional to sedimentation basin performance.
The plant staff mentioned that IFE turbidity seemed to degrade when "top offilter "turbidity val-
ues approached 1 NTU or less. This trend appears in Figure 4 above, where sedimentation

103

-------
Basin 2 effluent and IFE data are plotted together. This trend is even more apparent in
Figure 11, which depicts daily maximum TOF data versus daily maximum IFE data.

FIGURE 13. Top of filter turbidity profile versus IFE turbidity for each filter.

Because this observation is counterintuitive and unique to the evaluation team, an additional
analysis was performed using the 15-minute IFE turbidity data rather than the daily maximum
values. Figure 12 displays data that were observed for several days during the months of April
and October, representing times when the settled water turbidity data were generally character-
ized by high and low values, respectively. This analysis compared performance of sedimentation
Basin 2 to each individual filter for approximately the same time period that the sedimentation
basin samples were measured. This analysis reveals that Filters 2, 3, and 4 performed better in
the month of April, which is when settled water turbidity values were high. Only Filter 1 per-
formed better in October, when the settled water turbidity data was lower. It is possible that tur-
bidity carryover from the sedimentation process could act as a "filter aid" that enhances filter

104

-------
performance; however, that turbidity carryover could also contain pathogens. The goal in opti-
mization is to optimize each treatment barrier for pathogen removal, including sedimentation and
filtration. If low settled water turbidity contributes to higher filter effluent turbidity, areas of fil-
ter operation and backwashing should be investigated to identify potential causes and solutions
to the filter performance issue.

0.10

0.09

Turbidity per Filter Number in April vs October

•	April Data (higher settled water turbidity)

•	October Data (lower settled water turbidity)

0.07

£ 0.05

0.03

0.01

0.00

2	3

Filter Number

FIGURE 14. Seasonal settled water turbidity versus IFE turbidity for each filter.

Disinfection

Disinfection is the final barrier in the treatment plant for protection from microbial pathogens.
CT represents the disinfection concentration (C) multiplied by the contact time (T) (adjusted for
basin hydraulics). The plant operators measure parameters to calculate CT daily, and they use a
spreadsheet to compare the daily required CT value to the calculated CT value. The CPE team
used the data from the plant CT calculations to evaluate historical disinfection performance. An
inactivation ratio is determined by dividing the measured CT value by the required CT value, and

105

-------
the inactivati on ratio values for 2016 are plotted in Figure 13. The optimization goal for disin-
fection is an inactivati on ratio of at least 1.0 (demonstrating compliance with the regulatory
requirement). Figure 13 shows that the XYZ WTP met the goal every day during the year by a
wide margin. Although the goal was met, the CPE team noted that a report from June 2016
found an average of five feet of sediment in the bottom of both clear wells. Had the volumes of
the clear wells been decreased accordingly in the calculations, the margin of exceeding the goal
would have been less, but the trend still would have showed the goal to be attained consistently.

Inactivation Ratio at pk flow CI 	Minimum inacti vat ion ratio

18.0 I—

0.0 '			*

^ ^ ^ ^ ^ ^ ^ ^

FIGURE 15. Daily disinfection inactivation ratio.

Cvanotoxins

During HAB events, cyanotoxins can enter the plant as either intracellular cyanotoxins (con-
tained within a cyanobacteriacell, or "cell-bound" [e.g., in Microcystis]) or extracel 1 ular cyano-
toxins (outside the cell, or dissolved). Intracellular cyanotoxins can also be released from cells if
they are lysed (broken apart) or stressed during treatment in the plant. Total microcystins refers
to the sum of both intracellular and extracellular microcystins.

106

-------
Over the past three years total microcystins have peaked in XYZ's raw water in late May and
again in late November - early December, as shown in Figure 14. XYZ experienced a finished
water microcystins detection of 3.4 ug/L on May 7, 2015, but a drinking water advisory was not
issued since microcystins were not detectedin repeat finished water or distribution system
samples. Resample results showed that extracellular microcystins were between 8-40% of the
total microcystins concentration in the raw water on May 8, 2015.

1 A

"5th o

iA
C:

1|_» Q

a 5b

vho

)

$ %
s0 O

~ Q

c
r
c

7 <©

samples
Collected

h>rQ<

¦QrrrrrrJP 9 Q,

D
H

29/2014 ,

18/2014 i

77/2014

26/2015

17/2015

i/6/2015 <

25/2015 (

14/2015 i

1/3/2015

22/2015 ^

11/2016

i/1/2016

20/2016

i/9/2016

29/2016

17/2016 t

./6/2016

oc/or\i c. .¦

FIGURE 14. Raw water microcystins concentrations at XYZ WTP, 2014 - 2016.

Treatment train sampling was conducted on May 9, 2015 (see Figure 15) and showed 0-52%
removal of total microcystins following powdered activated carbon addition at intake (assuming
either blend of lower and upper intake or just the lower intake was in use at the time). The "In-
Plant 1" sampling location is post powdered activated carbon addition. An additional 72%
removal was achieved through settling processes and 68% removal was achieved through
filtration and chlorine disinfection. The 0.1 ug/L of microcystins present in the finished water on
May 9 is below State EPA's microcystins reporting limit of 0.3 ug/L. Following the initial
finished water microcystins detection, Cadiz increased their powdered activated carbon dose and
has not had any repeat finished water microcystins detections above State EPA's reporting limit.

107

-------
State EPA conducted additional treatment train sampling during bloom events in May 2016 and
in December 2016. During the December sampling event, 92-95% of the microcystins were
intracellular in the raw water, and microcystins were reduced to non-detectable concentrations
following sedimentation.

3

2.5

FIGURE 15. XYZ WTP microcystins profile on May 9, 2015.

It is not known if the plant was just utilizing the lower intake or a blend of the upper and lower
intake at the time.

On-site Studies

During the CPE, several studies were conducted to assess current plant performance and process
control. These studies included: 1) filter probing, media assessment, bed expansion, and
backwash waste turbidity profile; 2) filter backwash recovery; 3) online turbidimeter flow rate
and sample time assessment; 4) chemical feeder calibration check, and 5) jar testing to assess
impact of PAC addition on cyanotoxin removal.

108

-------
Study 1: Filter Probing, Media Assessment, Bed Expansion, and Waste Backwash Profile

During the CPE, Filter 4 was drained by the plant staff to allow the CPE team to conduct filter
evaluation studies. Each of the studies is described in the sections that follow.

Filter Probing-

The purpose of conducting a filter probing study is to evaluate the overall depth of media in the
filter. This is accompli shed by probing the filter at approximately equally-spaced distances in a
grid-like pattern across the plan area of the filter. Once depths are measured, the data points can
be plotted and used to determine areas where the bed is uneven or where media loss has
occurred. However, due to the design of XYZ's filters, a complete filter probe study could not
be conducted. As seen in Figure 16, the backwash collection troughs have shrouds in place to
keep media from being lost over the trough. These shrouds covered a large area of the filter,
making maneuverability across the filter bed difficult. Therefore, only the center-line was
probed in six-inch to one-foot increments, as shown in Figure 16. The filter was originally
installed with a 3-inch layer of torpedo sand followed by a 12-inch layer of sand and then a 15-
inch layer of anthracite, for a total media depth of 30 inches. At no point were team members
able to probe down to the full depth of the media, due to the tightness of the media.

109

-------
FIGURE 16. Filter 4 plan view.

Filter Inspection-

The purpose of the filter inspection is to observe physical conditions of the filter media. A visual
examination was made of the media once Filter4 was drained. The media was excavated to
depths ranging from 30 to 34 inches. No mudballs or cracks were observed throughout the
media. Two small sections of the filter bed were excavated by hand to observe the degree of
stratification between the anthracite and sand media. The excavation sites selected were areas of
the filterthat exhibited the most resistance when insertingthe probe. These locations were at
eight inches from the right wall and at two feet from the left wall, as shown in Figure 17.
According to the filter design, there should be a distinct layer of anthracite over a layer of sand,
typically with a small layer of intermixing between the two distinct layers. The excavation in
Filter 4, shown inFigure 18, revealed almost complete intermixing of the anthracite and sand
layers throughout the depth of the filter bed profile. Also, lime particles were found throughout
the media.

110

-------
Filter 4

January 25, 2017

Probed 15 areas along
centerline of filter.

Area of Excavation:
Highest area of resistance
to probing found
approximately 8 inches
from right wall and 2 feet
from left wall.

Top several inches showed
mostly anthracite but
some sand mixed in.

As depth progressed the
anthracite and sand were
completely intermixed
along profile.

Excavated media depths
ranged from 30" at Site 1
to 34" at Site 2.

Centerline
of filter

FIGURE 17. Filter 4 probing and excavation locations.

FIGURE 18. Pictures of mixed media found in Filter 4 bed during excavation.

Ill

-------
Re-stratificationof media following a backwash cycle is a function of media density and size,
filter bed expansion during backwash, and the approach used to ramp down the backwash flow
rate. In Filter 4, the failure of the anthracite and sand to re-stratify back to their distinct layers
following a backwash cycle could be due to low bed expansion provided during the backwash
cycle. As observed from the backwash study, only one inch of bed expansion was achieved dur-
ing backwash (see further discussion of bed expansion below). The current bed expansion is
limited by the backwash system design. The flow rate available for backwashing is determined
from the water level in the filter effluent wet well. Also, the backwash flow rate cannot be
ramped down since there is no automatic control of the rate in the system design.

A possible issue with completely mixed filter media is more rapid increase in filter headloss and
binding of the filters when operating under high hydraulic and solids loading conditions. Under
normal operating conditions, this can potentially be managed operationally, by reducing hydrau-
lic loading and reducing filter run times. However, the mixed filter media could be challenged to
produce optimized water quality (turbidity) when operating with higher hydraulic rates and sol-
ids loading conditions (e.g., during a HAB event). Instead, the achievable rate of filtration
(gpm/sf) would likely decrease, and shorter filter run times may result.

During this CPE, only Filter 4 was evaluated; however, the media condition could be similar in
the other filters. Also, additional studies to optimize filter run times to improve filter perfor-
mance and to optimize backwash duration to improve filter return to service can be conducted.
Design modifications to allow for further adjustment of backwash flow rate and to improve re-
stratification of the media may be an option.

Bed Expansion-

The purpose of conducting the bed expansion study is to determine the extent to which the filter
media expands during a typical backwash cycle. The percent bed expansion can be calculated
from the measurement of media expansion. The percent of bed expansion helps operators under-
stand the effectiveness of the backwash cycle in cleaning the media and the ability for media re-
stratification following a backwash. A minimum of 20 percent filter bed expansion is desirable;
however, filters using air scour can achieve satisfactory backwashing at lower expansion rates
(i.e., 15 percent). During this study, a Secchi disk attached to a pole was used to measure the
extent of media expansion. The CPE team marked the pole when the Secchi disk was sitting on

112

-------
the media before the filter backwash and again at the high backwash flow rate when the Secchi
disk was observed to disappear below the fluidized media. The distance between these two
marks represents the depth of media expansion.

FIGURE 19. Filter bed expansion tool.

The CPE team attempted to determine the extent of expansion during the high-rate portion of the
Filter 4 backwash cycle. Due to the depth of the filter and limited light in the filter, determining
the top of the media layer was difficult. Through a combination of visual observation and feel-
ing the difference in water density at the media interface, a bed expansion of about one inch was
estimated. Based on 30 inches of filter media, the bed expansion was approximately 3 percent.
This bed expansion is at the low end of the typical range, even for filters with combined air and
water backwash. Due to the challenges of measuring the bed expansion, a repeat of the study is
suggested for all of the filters. Alternative bed expansion measurement tools, such as the one
shown in Figure 19, may provide more conclusive results for the XYZ filter design.

Backwash Waste Turbidity Profile-

The purpose of conducting a backwash waste turbidity profile study is to determine the amount
of time necessary for effective media cleaning. The equipment used to perform this study

113

-------
included a sample collection device for grab samples, a YSI EXO sonde, and a portable turbi-
dimeter. During the Filter 4 backwash, the CPE team collected eight turbidity grab samples at
varied times from the discharge of the waste trough, using a long pole with a sample bucket
attached at the end. In addition, the EXO sonde was lowered into the filter and placed in the
water on top of the filter, next to the washwater launder. The continuous turbidity data from the
EXO sonde and the turbidity data from the grab samples are shown below in Figure 20. These
data show that the waste backwash water turbidity spiked to about 400 NTU immediately after
the start of the backwash and gradually decreased to the 100 NTU range. After the blower was
turned on, the turbidity spiked to the 150 NTU range and then gradually decreased to the end of
the backwash when the turbidity was less than 5 NTU. The results suggest that most of the parti-
cles trapped in the filter have been removed during the backwash.

Backwash Turbidity - 1/25/2017

FIGURE 20. Filter 4 waste backwash water turbidity profile.

114

-------
This study approach can be used by plant staff to determine an optimum turbidity level to stop
backwashing and to determine if inadequate or excess backwash water is being used during filter
backwash. The plant operators are encouraged to periodically conduct this study to support opti-
mization of the filter backwash procedure.

Study 2: Filter Backwash Recovery

The optimization goal for plants with filter-to-waste capability is to return the filter to service at
< 0.10 NTU. Following the inspection and backwash of Filter 4, the filter effluent turbidity was
monitored as the filter was returned to service, and these performance data from the continuous
turbidimeter are shown in Figure 21. The challenge with the XYZ filter backwash design is
knowing when the filter returns to filtering mode. Based on discussions with the plant staff, it
was estimated that this occurs about ten minutes after the backwash waste valve closes and the
filter begins to fill. The blue data points in Figure 21 represent turbidity from the Filter 4
turbidimeter when backwash is occurring, while the red data points represent turbidity from the
filter when the water is going to the combined filter wet well (i.e., return to service). During
return to service, the turbidity spiked to 0.7 NTU and gradually decreased to 0.22 NTU after
35 minutes. For filters without filter-to-waste capability, the optimization goal isto limitthe
turbidity spike to less than 0.30 NTU and return to < 0.10 NTU within 15 minutes. These goals
were not achieved during this filter backwash. Achieving these goals following each filter
backwash can reduce the number of particles (including pathogens and cyanobacteria cells) that
pass to the clearwell and, as a result, can enhance public health protection.

115

-------
Filter 4 Return-to-Service Profile 1/25/16

Overview:

• Filter backwash ripening from inspected filter.

• Turbidity of backwash supply water was ~ 0.05 NTU and spiked to ~ 0.6 NTU during middle of air scour.
Reason not known, but could be related to turbulence (air or particles) in the filter wet well.

• Turbidity spiked at a bout 0.7 NTU and recovered to ~ 0.22 NTU after 35 minutes.

• Operator logged filter online at 3:41; however, likely delayed due to filter investigation.

2:24 2:38 2:52 3:07 3:21 3:36 3:50 4:04

Return to Service (hr:min)

FIGURE 21. Return-to-service profile for inspected Filter 4.

The time when the data logger was turned back on in the SCADA system is also shown in
Figure 21. It should be noted that the operator was likely delayed in completingthis final back-
wash activity due to the study that was occurring at the time. However, through discussions and
interviews with plant staff, it was determined that there was not a consistent approach for deter-
mining when the data logger is turned back on in the SCADA system (e.g., one hour after filter
backwash starts, after a downward trend is observed on the turbidimeter). Establishing a con-
sistent, data-based approach for completingthis task is important for monitoring water quality
that is representative of all water going to the combined filter wet well.

Study 3: Online Turbidimeter Flow Rate and Sample Detention Time Assessment

To assess the flow rate of the online Hach 1720E turbidimeters in the filter gallery, the CPE team
used a graduated cylinder and a stop watch to measure the flow rates from the individual filter

116

-------
effluent (IFE) turbidimeter drain lines (Figure 22). IFE Turbidimeter 1 (284 mL/min), Turbidim-
eter 2 (282 mL/min), and Turbidimeter 3 (284 mL/min) fell within the manufacturer's recom-
mended flow rate range of 250 to 750 mL/minute, as shown in Figure 23. The flow rate of IFE
Turbidimeter 4 (200 mL/min) fell below the manufacturer's recommended minimum flow rate of
250 mL/minute. With the flow rate below minimum, the particles in the turbidimeter sample line
and meter may settle out, causing a slightly lower effluent turbidity reading for Filter 4 (i.e., false
low value).

FIGURE 22. Online turbidimeters with SC200 controllers on the
operating floor of the filter gallery.

117

-------
Online Turbidimeter Flow Study 1/26/17

800

700

600

500

o 400

300

Online turbidimeter flows were measured
and showed Filter 4 turbidimeter was below
manufacturer's recommended minimum
flow rate of 250 mL/min.

IFE 1 sample DT
IFE 2 sample DT
IFE 3 sample DT
IFE 4 sample DT

1.9 minutes
1.9 minutes
1.7 minutes
2.3 minutes

Optimization recommendation < 1 minute

l Flow Rate
¦ Low Range

High Range

200

100

Turbidimeter

FIGURE 23. Online turbidimeter flow check and sample line detention time.

The CPE team calculated the sample detention time of the online Hach 1720E turbidimeters in
the filter gallery. The CPE team estimated and measured the distance from the effluent ports of
individual filter underdrain pipes to the influent ports of the online turbidimeters. The sample
lines are comprised of 1/2-inch and 1/4-inch tubing. The length of 1/2-inch tubing from the
effluent ports of individual filter underdrain pipes to a 1/2-inch tubing manifold on the basement
floor of the filter gallery was estimated using XYZ plant design plans. A tape measure was then
used to measure the remaining length of 1/2-inch tubing from the tubing manifold to the influent
ports of four PulsafeederChemtech mechanical metering pumps on the basement wall of the fil-
ter gallery. The length of 1/4-inch tubing was then measured from the effluent ports of the
metering pumps through the operating floor of the filter gallery to the influent ports of the on-
line IFE turbidimeters. Usingthe distances and flow rates from the filter underdrain pipes to the

118

-------
turbidimeters, a detention time was calculated for each individual filter as listed in the legend
box in Figure 23. The detention times varied between 1.9 and 2.3 minutes, which is greater than
the recommended sample travel time of one minute or less for IFE turbidimeters. The sample
line distances have already been minimized at the XYZ plant; however, the sample detention
time can be reduced by increasing the flow rate settings on the metering pumps or adjusting flow
control valves.

The CPE team checked the settings on the online Hach 1720E turbidimeter SC200 controllers, as
listedin Table 5. Each SC200 controller communicates with two turbidimeters. A source of his-
torical turbidity data is available from the controller, which stores IFE turbidity values based on
the selected datalogging interval (30 seconds, 1 minute, 5 minutes, 10 minutes, or 15 minutes).
All four turbidimeters were set on a datalogging frequency of 15 minutes (the Hach default
value), allowing storage of about four to six months of data. These data can be downloaded from
the controller using an SD card and used to compare the IFE SC200 turbidity values against the
IFE turbidity values in the SCADA system. While the turbidimeter controller is storing data in
15-minute increments, the controller transmits continuous data (i.e., value approximately every
second) to the plant SCADA system, where it is displayed on the control screen and used to gen-
erate reports.

Signal averaging was set on 90 seconds (the Hach default value is 30 seconds) on all four turbi-
dimeters. Signal averaging every 90 seconds reduces the impact of outliers by averaging
datasets over a 90-second period. Higher signal averaging values produce a smoother signal but
increase the time it takes for a signal to respond to a change in the process value. All four turbi-
dimeters had their bubble reject value on (the Hach default setting) to prevent the recording of
erratic readings caused by air in the sample lines. Error Hold Mode and Output Span are
explained in detail in the notes listedbelow Table 5. The XYZ plant has good operation and
maintenance practices for their turbidimeters, with replacement of sample lines every three
months and cleaning of turbidimeters monthly. Turbidimeters are also calibrated quarterly by
operators and verified monthly with a Hach Ice Pic, as confirmed in the XYZ plant calibration
history.

119

-------
TABLE 6. Data Integrity Study: Turbidimeter Settings

Turbidimeter Location

Filter Effluent
No. 1

Filter Effluent
No. 2

Filter Effluent
No. 3

Filter Effluent
No. 4

Turbidimeter Model

1720 E

1720 E

1720 E

1720 E

Controller Model and Data
Logging Setting (1)

SC200
15 minutes

SC200
15 minutes

SC200
15 minutes

SC200
15 minutes

Signal Averaging (2)

90 seconds

90 seconds

90 seconds

90 seconds

Bubble Reject (3)

On

On

On

On

Error Hold Mode (4)

Hold Outputs

Hold Outputs

Hold Outputs

Hold Outputs

Output Span (5)

Oto 100 NTU

Oto 100 NTU

Oto 100 NTU

Oto 100 NTU

Other

Date stamp check: Off by 1 hour

Operational practices:

•	Replace sample lines every three months

•	Clean turbidimeters monthly

•	Calibration quarterly by operators and verification monthly
with Ice Pic (confirmed in Calibration History)

(1) Check to see if current data and time are correct. Check frequency of data logging. Default is 15 minutes
for Hach models.

(2)	Default for Hach models is 30 seconds. This is acceptable in most cases.

(3)	Default is On for Hach models. This is acceptable in most cases.

(4)	Specific to Hach 1720E and FilterTrak 660 models. Default is to Hold Outputs (HO) and send last known
value to SCADA when turbidimeter loses communication with controller. Better option is Transfer Outputs
(TO) to send an operator-selected value to SCADA (e.g., 0, 99) to make operator aware of problem.

(5)	To avoid "capping" of data to SCADA, the output span should be at least 0 to 5.1NTU (applicable to ana-
log signals). Span for all XYZIFE turbidimeters 0 to 100 NTU (no data capping issues).

Accessing output span for Hach SC200 controller: Menu/SC200 setup/Output setup (select 1 or 2; select
Source to see which turbidimeter is highlighted and then Backbutton)!Activation (low value; high value).

Study 4: Chemical Feeder Calibration Check

A study was conducted to determine the feed rate and dose of chemical additions in the plant.
Results are summarized in Table 6. The CPE team collected powdered activated carbon (PAC),
lime, and soda ash from their respective feeding points over a measured time period and weighed
the samples on a triple beam balance available in the plant's laboratory. The weight of PAC at
the in-plant and at the Lake intake were found to be 58.4 grams per two minutes and 174 grams
per five minutes, respectively. Based on a plant flow of 0.79 MGD, this equated to a PAC dose

120

-------
of 14 mg/Lin-plantand 17 mg/L at the intake. The PAC feed rates were comparable to the total
usage over time reported by the operators.

Lime and soda ash additions were measured in a similar fashion and were found to be 262 grams
per two minutes for lime and 473 grams per two minutes for soda ash. This equated to a
63 mg/L lime dose and a 114 mg/L dose of soda ash.

In addition to the dry chemical additives, the CPE team also measured the liquid coagulant and
alum feeds. The coagulant (AS3040) is added at the primary rapid mix and was measured at
103 mL per 30 seconds for a dose of 130 mg/L. Alum is added at the second-stage rapid mix
basin. Alum was measured at 12 mL per minute for a dose of 7.6 mg/L.

The operators do not routinely check their chemical feed dosage rates, and they also reported that
chemical feed dosages are seldom changed. Routinely recording and confirming plant chemical
feed dosages can provide useful information to support plant optimization.

TABLE 7. Summary of Chemical Feeder Calibration and Dose Results

Chemical Feeder

Feed Rate,
g/min or
mL/min

Feed Rate,

lb/day
(24 hr basis)

Flow Rate,
MGD

*Chemical Dose,
mg/L

PAC plant
(small pulley)

58.4 grams/
2 minutes

0.79

PAC at lake (small
pulley)

174 grams/
5 minutes

110

0.79

Coagulant
(AS3040)(1)

103 mLs/
30 seconds

860 (wet basis)

0.79

130

Alum'2'

12 mLs/
1 minute

50 (wet basis)

0.79

7.6

Lime

262 grams/
2 minutes

416

0.79

Soda Ash

473 grams/
2 minutes

749

0.79

114

^Reported specific gravity = 1.31; wet weight = 10.93 lb/gal
^Reported specific gravity = 1.307; wet weight = 10.9 lb/gal

121

-------
Study 5: Jar Testing to Assess Impact of PAC Addition on Cyanotoxin Removal
Introduction-

Jar testing is a valuable tool for assessing and optimizing water treatment plant chemical dosing,
especially during a harmful algal bloom (HAB) event. When source water quality changes, it is
important to determine if chemical dosing needs to be adjusted or added to remove cyanotoxins
while maintaining other treatment objectives, such as turbidity and TOC removal. Operators can
conduct jar testing in advance of, or during, the initial stages of HAB occurrences. Utilizing
concentrated raw water samples or spiking commercially-available stock cyanotoxins can be
beneficial to prepare for the changes in chemical dosing that may be necessary to address the
HAB.

Jar testing simulates the water treatment plant's processes by setting mixer speeds. Current plant
chemical dosing performance can be evaluated and compared with alternative chemical dosing
scenarios for changing raw water quality. A simple jar test was conducted to simulate the addi-
tion of powdered activated carbon (PAC) at the raw water pumping station and the 10-mile raw
water transmissionmainto the plant. Various PAC doses were evaluated to assess cyanotoxin
removal.

Approach-

Raw water and concentrated sample

To mimic the conditions XYZ might experience during a HAB event, including elevated micro-
cystins and natural organic matter (NOM), a composite sample was prepared using raw water
from XYZ's Lake and raw water from another lake. Cyanobacteria and their associated
intracellular cyanotoxins were concentrated from another lake source water using a
phytoplanktonnet, and they were transferred to four unpreserved 1-liter (L) amber glass jars on
December 19, 2016. Based on historic sampling data, the cyanobacteria genera and microcystin
variants present in the lake water are similar to those present in XYZ's lake water. Total
microcystins concentrations in the other lake concentrated samples ranged from 100 to 200 |ig/L,
and extracellular microcystins ranged from 2.2 to 2.7 |ig/L. The samples were frozen to lyse
(break apart) cyanobacteria cells and increase the extracellular percentage of microcystins in the
sample, to better evaluate extracellular microcystins removal by PAC. The samples were held
frozen, then thawed the day prior to the jar test. Equal volumes of raw water from the XYZ's

122

-------
upper and lower Lake intakes were collected in one-gallon amber glass jars on January 24, 2017.
Equal volumes of upper and lower intake water were combined with the concentrated alternate
Lake samples. A sample of this prepared raw water was determined by ADDA-ELISA to be 23
|ig/L total microcystins concentration, with 11 |ig/L being in the form of extracellular
microcystins. This composite sample was stirred using a metal spoon to keep the cellular matter
suspended, and it was split between the six jars utilized for the jar test.

Jar testsettiri2s determination

The j ar test was set up to replicate the dosing of PAC at the raw water pump station at XYZ's
lake and the travel time in the 12-inch/8-inch transmission main to the water treatment plant
(Figure 24). The jar test time was based on the volume of the 12-inch and 8-inch pipe diameter
lengths of pipe estimated at 42,103 ft3. Mixing energy in the jars was assumed to be equivalent
to the headloss in the transmission line. Using the average flow rate in the raw water line of 550
gpm (or 0.8 MGD), headloss was calculated to be approximately 44 feet. This results in a
mixing energy expressed as velocity gradient (G) of 52 sec"1. Table 7 below summarizes the
calculated values for the jar test settings. Using Figure 25 below and a velocity gradient (G) of
52, a jar test setting of 61 rpmwas determined to replicate the mixing in the raw water line. The
calculated detention time in the transmission main, assuming a 550 gpm average flow rate, is 9.6
hours. The jar test was run for 24 hours at 61 rpm, with samples taken at times 0, 1,2,4,9.6
(detention time), and 24 hours. The jar test was extended beyond the detention time in the pipe
(9.6 hours) to determine if additional contact time with PAC affects the adsorption of
microcystins for this prepared raw water.

123

-------
FIGURE 24. Jar testing to simulate impact of variable PAC dosages and contact times on
microcystins concentrations.

TABLE 8. Summary of Jar Test Settings to Replicate Raw Water Transmission Line

Parameter

Input

Units

Raw Water Line Volume

42,103

ft3

Flow Rate

0,8

MGD

Detention Time

9.6

Hr

Head Loss

44

Ft

Velocity Gradient,1)

52

Sec1

Jar Mixer Speed

61

RPM

® Based on water temperature of 8°C.

124

-------
Mixing Curves for Flat Paddle in 2-Liter Square Jar

Impeller Speed (rpm)

FIGURE 25. Velocity gradient versus mixer speed in 2-liter square jar.

PAC slurry and jar dosing

A stock PAC slurry was created using the plant's PAC and deionized (DI) water at a concentra-
tion of 10 mg/mL such that each 1 mL of slurry equaled 5 mg/L PAC concentration in the 2-liter
(L) jars. A total of six 2-liter Phipps & Bird jars were run using a Phipps & Bird PB-900 Pro-
grammable Jar Tester with the following concentrations of PAC: 0, 10, 20, 30, and 40 mg/L.
The j ar with zero PAC added was the control j ar. A duplicate 20 mg/L PAC j ar was run for
quality assurance. Table 8 on the following page summarizes the PAC dosing regimen for the
jar test.

125

-------
TABLE 9. Jar Test PAC Dosing Regimen



PAC

PAC Slurry

Jar No.

Dose

Dose



mg/L

mL

1

0

0

2

10

2

3

20

4

4

20

4

5

30

6

6

40

8

Eachjarwas dosed, as described above, at 10-minute staggered intervals to allow timefor sam-
ple processing from eachjar (i.e. vacuum filtration, glassware rinsing, and pH determination) at
the specified time intervals.

Jar Sampling-

Approximately 90 mL of sample was collected from eachjar at each time interval, vacuum fil-
tered with 0.6-micron glass fiber filters (Whatman, grade GF/F), and transferred to 125 mL
PETG containers. Samples were then analyzed using the Ohio EPA ADDA-ELISA method12 for
determining extracellular microcystins. The initial composite sample was analyzed for both total
and extracellular microcystins (23 |ig/L total microcystins, 11 |ig/L extracellular microcystins -
see discussion above). pH was analyzed using a Hach SL1000 handheld instrument on an addi-
tional 30 mL of sample that was collected from eachjar at each time step through four hours.

12 Ohio EPA Total (Extracellular and Intracellular) Microcystins - ADDA by ELISA Analytical Methodology.
OhioEPADES 701.0. Version2.2. November 2015. Retrieved April 28,2016, from

http://www.epa.oliio.gOv/Portals/28/documents/rules/draft/01iio%20EPA%20DES%20701.0%20Version%202.2
Dec2015.pdf

126

-------
PAC jar test troubleshooting

• When conductingjar tests with PAC for extended periods (i.e. greater than four hours, as
observed during this study), PAC can accumulate in the j ar sampling ports and cause a slug
of high-concentration PAC to come through the sampling line during sampling at later time
steps. This is likely because the mixing paddles are at the same level as the sampling port.
To avoid taking unrepresentative samples with high concentrations of PAC, the sampling line
was flushed of the high-concentration PAC and poured back into the jar prior to taking a
sample for cyanotoxin and pH analysis.

• When using a concentrated raw water sample for jar testing, ensure the sample is well-mixed
prior to adding to the j ars. With high concentrations of cyanobacteria cells and other algae or
organic matter, it is important to add equal concentrations in each j ar. These suspended
materials may begin to settle out, resulting in varied concentrations added to each jar if the
stock raw water sample is not well-mixed.

• Staggering the PAC dosing by five to ten minutes is recommended to provide time in
between taking jar samples for any sample processing, such as filtering the sample to remove
PAC, taking measurements on the sample (pH, temperature, turbidity, TOC/DOC, etc.), and
cleaning sampling or filtration glassware and other sampling equipment.

Results and Discussion-

For this section, refer to Figures 26 and 27, which are different graphical representations of the
same data. PAC performance varied based on dose and, in some cases, time. At the lowest dose
evaluated, 10 mg/L PAC, extracellular microcystins were not appreciably reduced throughout the
course of the experiment. At the 20 mg/L dose, reduction in extracellular microcystins was
observed, with additional reduction at later time steps in one jar, but not the other. At the 30 and
40 mg/L PAC doses, microcystins removal occurred more rapidly, with little difference between
one hour and 24-hour time steps. The 40 mg/L PAC dose was most effective at microcystins
removal, achieving an 86 - 95% reduction.

127

-------
The concentration of extracellular microcystins increased over time in the control j ar, and
increased between the 9.5 and 24-hour time points in 4 of the 5 jars dosed with PAC. This may
be due to degradation of Planktothrix filaments in the j ars over time, causing additional cell lysis
and microcystins release. Visual microscopic observation of Planktothrix filaments from the
control sample show cell rupture and degradation (see images Figure 28). The concentration of
microcystins in the initial composite sample was 23 |ig/L total (intra- and extracellular) and 11
|ig/L extracellular. Some of the microcystins that were still in an intracellular form at the
beginning of the experiment could have leached from these degraded filaments and entered an
extracellular form by the end of the experiment, as evidenced by an increase from 5.6 to 9.9 |ig/L
extracellular microcystins concentration in the control jar from hour 0 to hour 24.

The initial concentration of extracellular microcystins for eachjar varied (4.7 to 8.5 |ig/L), and
there are multiple hypotheses for this variability. First, Planktothrix filaments are quite variable
in length, and even a well-mixed sample is difficult to proportion evenly. This variability may
be reduced if the concentrated sample was added as a spike to eachjar instead of mixed into the
raw water sample prior to splitting into the individual j ars. Alternatively, the filaments may be
filtered out prior to adding extracellular microcystins to the water sample; however, this will also
remove organic matter that may affect PAC removal capacity. pH was also analyzed in the jar
test samples to determine if the addition of PAC to the raw water impacts pH. Results through
four hours indicate that pH is likely not affected by the PAC.

128

-------
Cadiz Jar Test

1/25/2017

5	10	15	20	25

Time (hr)

¦	Control	•— 10 mg/L PAC	~	20 mg/L PAC

¦	20 mg/L PAC —•— 30 mg/L PAC	~	40 mg/L PAC

FIGURE 26. Graph showing jar test results for inicroevstin concentration versus

PAC dose and time.

129

-------
15

PAC / Microcystins Jar Test Results

00

Control	10 mg/L PAC 20mg/LPAC 20mg/LPAC 30mg/LPAC 40mg/lPAC

Time [hrs] 0 «1 ml «4 «9.5 ¦ 24

FIGURE 27. Bar chart showing jar test results for microcystins concentration versus PAC dose and time.

130

-------
Diatoms

Detritus



*2J

Planktothrix spp.
filaments

VI

131

-------
Pictures of phytoplankton from control sample preserved with Lugol's iodine solution. Left image shows phytoplankton community i
dominated by Planktothrix spp. filaments (280x magnification). Right panel shows a partially degraded filament of Planktothrix with
ruptured cells (560x magnification).

FIGURE 28. Pictures of phytoplankton from control sample.

132

-------
133

-------
MAJOR UNIT PROCESS EVALUATION

The plant's maj or unit processes are assessed with respect to their capability to meet the optimized
goals for:

•	Settled water turbidity

•	Filtered water turbidity

•	Disinfection (inactivation ratio goal)

•	Cyanotoxin adsorption with powdered activated carbon

•	Cyanotoxin oxidation with chlorination

There is an emphasis on turbidity reduction to remove cyanobacterial cells through the multiple-
barriers of the treatment process; however, recognizing that toxins will likely be in the raw and set-
tled water, toxin removal and destruction are also considered. The capability of each individual
unit process is assessed to estimate its ability to provide water that consistently meets the opti-
mized performance goals.

Since the treatment processes of the plant must provide multiple effective barriers at all times, a
peak instantaneous operating flow is also determined. The peak instantaneous operating flow
represents conditions when the treatment processes are the most vulnerable to the passage of par-
asitic cysts, microorganisms, and toxins. If the treatment processes are adequate at the peak
instantaneous flow, then the maj or unit processes should be capable of providing the necessary
effective barriers at lower flow rates. The flow through the plant is controlled by the raw water
pumps, and each of the pumps is rated at about 570 gpm. Through discussions with operators
regarding operational policies at the plant as well as the review of plant operating records by the
CPE team, the equivalent flow rate of 0.8 MGD was selected as the peak instantaneous flow rate
through the plant under normal operating conditions. This rate was used to assess the capabili-
ties of all the maj or unit processes except for disinfection, which was established by the high ser-
vice pumps since they control the detention time in the clearwell. The peak instantaneous flow
of the high service pumps to be used in the disinfection evaluation was determined to be
1.1 MGD, based on a review of recent pumping records at the plant.

Unit process capability is assessed using performance potential graphs, where the projected treat-
ment capability of each maj or unit process is compared against the peak instantaneous operating

134

-------
flow rate and the plant design flow for comparison. An individual performance potential graph
is developed for each of the treatment objectives evaluated in this report: 1) microbial
removal/inactivation and 2) cyanotoxin removal and destruction.

Each unit process can fall into one of three categories:

Type 1: Where proj ected peak capability for the unit process exceeds the peak instantaneous
flow (>100 percent of peak flow), the plant should be expected to achieve the perfor-
mance goals.

Type 2: If the proj ected peak capability for the unit process falls short of, but is close to, the
peak instantaneous flow (80 to 100 percent of peak flow), then operational adjust-
ments may still allow the plant to achieve the performance goals.

Type 3: If projected peak capability for a specific unit process falls far short of the peak

instantaneous flow (<80 percent of peak flow), then it may not be possible to achieve
the performance goals with the existing unit process.

The results of the assessment, relative to these categories, are discussed below.

Particle Removal and Microbial Disinfection

The Maj or Unit Process Evaluation graph for microbiological treatment through turbidity
removal and disinfection, developed for the XYZ WTP, is shown in Figure 29. The unit
processes evaluated during the CPE are shown along the vertical axis. The horizontal bars on the
graph represent the projected peak capability of each unit process that would support
achievement of optimized process performance. These capabilities were proj ected based on sev-
eral factors, including: the combination of treatment processes at the plant, the CPE team's
experience with other similar processes, raw water quality, industry guidelines, the XYZ WTP
design, and regulatory standards and guidelines.

135

-------












Flocculation (A)

i



1.78

















Sedimentation (B)







2.05









Plant design capacity = 1.0 MGD 	



Conventional Filtration (C)

¦



L05













Peak Instantaneous Flowl = QS MGD (based on rav w ate rpump rate)	

Peak InstantaneousFlow2= 11 MGD (based on high service pump rate)

i

Disinfection (Giardia Inactivation) (D)



i

i

9.01



l l l l

i i i







0.0 1.0 2.0 3.0 4.0 5.0 6.0 7.0 8.0 9.0 10.0

MGD

Assumptions:

(A)	Flocculation: Assume both clarifier processes could be run in parallel or could be run in series as independent processes. Selected
20-minute HDT to allow adequate floe buildup and softening.

(B)	Sedimentation: Selected 10gpm/ft2SORfor softening clarification process, and the depth is greaterthan 14ft in both basins.

(C)	Conventional Filtration: Assume one filter is out of service. Selected 2 gpm/ft2 loading rate since filter inspection indicated that
filter media was mixed, low backwash rate, inability to ramp backwash to stratify in dual-media filter, and limited bed expansion
during backwash.

(D)	Disinfection (Giardia inactivation): Volume is based on a 12.85 ft lowest operating level. Baffling factor is assigned at 0.6 by
State EPA. Assumed pH of 9 and a maximum residual of 4 mg/L.

FIGURE 29. Major Unit Process Evaluation - XYZ Water Treatment Plant turbidity removal

(microbes, cells) and disinfection.

136

-------
The major unit processes evaluated include: flocculation, sedimentation, filtration, and disinfec-
tion. The flocculation and sedimentation processes are both in series, and the disinfection pro-
cess takes place in the clearwells. The shortest bar represents the unit process that may limit
plant capability the most relative to achieving optimized plant performance; and, in this case, it is
associated with filtration. The flocculation, sedimentation, and disinfection processes are rated
as Type 1, indicating that they should be capable of meeting the particle removal and microbial
treatment obj ectives at the assigned peak instantaneous flow rate of 0.8 and 1.1 MGD through
the facility. The filtration process is also rated Type 1, but it is the process that limits plant
capacity and would require careful operation to meet the optimization goals for particle removal
at the peak flow.

Cvanotoxin Removal and Destruction Treatment

If a HAB occurs in the XYZ WTP source and cyanotoxins appear in the raw water, the particle
removal processes in the water treatment plant would be able to remove most of the intracellular
cyanotoxins, provided the cells are removed before release of the cyanotoxins. The evaluation of
the major unit processes in the microbiological (turbidity) control section of this report gives an
estimate of the plant capacity to remove cyanobacteria cells through clarification and filtration to
control the intracellular cyanotoxins (Figure 29). Any extracellular cyanotoxins present in the
raw water or released in the plant would have to be removed primarily through powdered
activated carbon (PAC) adsorption or destroyed through chlorine oxidation in the plant clearwell.

The Major Unit Process Evaluation graph for extracellular cyanotoxin treatment through pow-
dered activated carbon (PAC) adsorption and chlorine oxidation, developed for the XYZ WTP, is
shown in Figure 30. A challenge target of 50 |ig/L microcystins at the raw water intake was used
in the evaluation, with the assumption that all of the toxins are extracellular. This assumption
was based on anticipated baseline occurrence data for inland lakes per State EPA's Guidance for
Developing a Harmful Algal Bloom General Plan13. The unit processes evaluated during the
CPE are shown along the vertical axis.

13 Guidance for Public Water Systems; Developing a Harmful Algal Bloom (HAB) General Plan, State EPA,
Division of Drinking and Groundwaters, Version 1.0, September 2016.
http://epa.ohio.gov/Portals/28/documents/habs/HABGeneralPlanGuidance.pdf.

137

-------
PAC Feed {Intake plus Rapid Mix) {A)

Cya no toxin Oxidation {B1)

Cyanotoxin Oxidation with a Safety Factor
of 2{B2)

27E

a co

0.50

LOO

L50

2.00

250

100

Pea* Instantaneous Flow — ~». —

Plant Design Capacity - — — — —

Assumptions:

(A) PAC feed is based on 40 mg/L. Dose is split between intake and rapid mix locations. Intake feed rate assumption (174 lb/day) will
require management of clogging issues. In-plant feed rate assumption (224 lb/day) is projected based on feeder potential.

(Bl) Assumes plant effluent residual = 4.0 mg/L, pH = 9, and temp. = 10°C. AWWA CyanoTOXcalculationsfor MC-LRare used to support
this assessment.

(B2) Same assumptions with Safety Factor of 2 applied. Basis for this is extrapolating the required CTto a higher pH (9.5) and lower tem-
perature (5°C) than the CyanoTOX spreadsheet can support.

FIGURE 30. Major Unit Process Evaluation - XYZ Water Treatment Plant microcystins adsorption and destruction.

138

-------
The horizontal bars on the graph represent the proj ected peak flow capability for each unit pro-
cess that would support achievement of optimized process performance. These capabilities were
proj ected based on several factors, including:

• PAC Feed Capacity: PAC is fed by two dry chemical feeders - one at the intake and one at
the rapid mix. The State EPA's Guidance for Developing a Harmful Algal Bloom General
Plan recommends that plants be capable of feeding a minimum of 40 mg/L of PAC and
have two feed locations. The XYZ WTP PAC feeds can be split between the intake and the
rapid mix locations. However, during the CPE site visit, the intake feeder was measured to
be operating at about 17.6 mg/L and was experiencing clogging issues that necessitated
daily manual unclogging by operators. To reach a total dose of 40 mg/L, the intake feeder
may not be able to function at a higher rate, but the remaining dose (22.4 mg/L) could be
added at the plant where operators would be onsiteto manually address clogging issues on
a temporary basis during the HAB event. The two PAC feeders are identical, and they
were estimated to each have a feed capacity of 34 mg/L notwithstanding the clogging that
might occur. The PAC would have over two hours' detention time even from the in-plant
feeder, which should be adequate for microcystin adsorption. Operating at such a high
dose (40 mg/L) would require careful oversight and management to prevent clogging of the
feed system.

• Microcvstins (Cvanotoxin) Oxidation: The oxidation process takes place through the appli-
cation of chlorine in the clearwells after the particle removal treatment processes. An esti-
mate was based on the AWWA Hazen-Adams CyanoTOX Toolfor Oxidation Kinetics
(Version 1.0)14 for microcystin-LR. However, since the calculator doesn't apply in the pH
and temperature range that XYZ's water treatment plant sometimes experiences during a
HAB, a safety factor of "2" was also applied to this estimate to account for these parame-
ters. Additionally, the team assumed that the plant could run with a plant effluent residual
of 4.0 mg/L, which is higher than normal but operationally possible.

14 AWWA Cyanotoxins resource site: http://www.awwa.org/resources-tools/water-knowledge/cyanotoxins.aspx
Office of Water (MS-140) EPA 815-B-22-005 June 2022

-------
Similar to the turbidity assessment, in evaluating the microcystins control processes in the XYZ
WTP, each process is assigned a rated capacity, based on a comparison of the rated capacity to
the peak instantaneous flows at the plant (using the raw water flow for the PAC assessment and
the high service flow for the oxidation assessment). The lowest projected process peak
capability (flow rate) represents the unit process that may most limit plant capability relative to
achieving optimized plant performance. Both the PAC adsorption process and the toxin
oxidation process were rated as Type 1. However, many operational assumptions were made for
the PAC assessment, and the oxidation estimates were made by extrapolation of the CyanoTOX
calculator results for MC-LR only and may not be reliable. A more reliable assessment of the
adsorption and oxidation processes could be made through in-house studies to evaluate the
removal and destruction efficiency of the plant processes on its unique water quality.

The overall major unit process summary for microbial and cyanotoxin removal and destructionis
summarized in Table 9 on the following page. The particle removal, microbial disinfection,
cyanotoxin adsorption, and chlorine cyanotoxin oxidation process ratings are all classified as
Type 1, indicating that the plant has the capability to achieve the microbial and cyanotoxin opti-
mization goals when excellent process control skills are applied. The filtration process in partic-
ular would have to be carefully operated to reliably and consistently meet the filtration goals.
Due to uncertainties in the assumptions made during the evaluation, the cyanotoxin removal and
destruction processes are rated as a more conservative Type 2, indicating that the plant has the
capability of achieving the microcystins finished water target, assuming careful attention is given
to plant O&M to prepare for and treat through a significant HAB event.

140

-------
TABLE 10. Major Unit Process Summary

Major Unit Process

Rating

Flocculation'1'

Type 1

Sedimentation(1)

Type 1

Filtration(1)

Type 1 (with careful operation)

Disinfection/Oxidation(1)

Type 1

PAC Adsorption Process,2)

Type 2

Chlorine Oxidation,2)

Type 2

(3)	Microbial treatment

(4)	Extracellular cyanotoxinremoval and destruction

PERFORMANCE-LIMITING FACTORS

The areas of design, operation, maintenance, and administration were evaluated to identify fac-
tors that limit performance. These evaluations were based on information obtained from the
plant tour, interviews, performance and design assessments, studies, and the judgment of the
CPE team. Each of the factors were assessed for the possible classification as A, B, or C
according to the following guidelines:

A Major effect on a long term repetitive basis

B Moderate effect on a routine basis, or major effect on a periodic basis
C Minor effect

The performance-limiting factors (PLFs) identified were prioritized as to their relative impact on
performance. They are summarized below. While developing the list of factors limiting perfor-
mance, over 50 potential factors were reviewed, and their impact on the performance of the XYZ
WTP was assessed. There were three "A " factors, three "B" factors, and one "C" factor
identified. Note that the asterisk on one "B" factor (B*) refers to a performance-limiting factor
identified for the specific situation when the plant is facing a harmful algal bloom in its source
water and must remove cyanobacteria and cyanotoxins during treatment. All other factors that
are listed apply to performance limitations for reduction of turbidity, which can also impact the

141

-------
ability of the water treatment plant to perform during a HAB event. Specific impacts on the
ability of the water treatment plant to perform during a HAB event are also indicated in the
description of each PLF.

Policies - Administration (A)

• The numerical optimization goals for individual clarifier effluent, individual filter effluent,
and combined filter effluent turbidity have not been established and relayed to staff. The
commitmentto produce water quality that is not only required by regulation, but is the best
quality that the plant can produce, must typically be embraced by the top administrators to
create the culture needed to optimize treatment processes.

• Due to the existence of a consistent quality source water, plant staff are not typically chal-
lenged to make significant process control changes which can lead to a lack of prepared-
ness for a HAB event. Striving to meet the water quality goals associated with optimized
treatment and empowering staff to achieve those goals can keep staff skills sharp and
enhance preparedness for potential HAB events.

• Plant staff confidence and capability to respond to unusual water quality conditions exists
primarily with the Operator of Record (e.g., the superintendent who is soon retiring has the
most institutional and operational knowledge at the plant; however, there is no detailed pro-
cedure in place to transfer this capability to other members of the plant staff).

• The utility is limited to one water source during a HAB event. Availability of the alternate
water source, Sparrow Reservoir, is uncertain because it has been inactive since April
2013, and a policy has not been established to pursue developing this source.

Application of Concepts and Testing to Process Control (Operations) (A)

• Plant staff do not perform jar testing to determine optimum softening and coagulant dos-
ages for different water quality conditions.

• Plant staff do not perform studies to determine most effective PAC type, dose, and mixing
energy to treat for HABs (e.g., wood-based versus coal).

142

-------
• Studies are not conducted to evaluate treatment options to optimize plant performance and
consider simultaneous treatment obj ectives.

• Plant staff do not calculate chemical dosages on a routine basis.

• Plant staff do not change chemical feed systems to respond to changes in raw water quality
(e.g., constantPAC feed at lake and plant).

• Filters are backwashed based on time in service or headloss rather than on optimized per-
formance goals for turbidity and particle removal.

• The depth of sludge in the clearwell and storage tanks is not monitored and removed on a
regular basis.

• The second-stage rapid mix is out of service, without plans for replacement. This may
limitthe ability to optimize softening and coagulation.

• The recarbonation basin is out of service, without plans for replacement. This may limit
the ability to stabilize the water following softening and reduce calcium carbonate deposi-
tion in filters and clearwell. Lower pH water would also be more effective for disinfection
and would be more effective for oxidizing cyanotoxins.

• Data logging may not accurately capture the turbidity when water is going to the clearwell
at plant startup and immediately after returning to service after a backwash.

Operational Guidelines/Procedures (Operations) (A)

• Inconsistent approaches are used by staff for logging turbidity data following backwash
(i.e., one hour following start of the backwash versus when turbidity starts to trend down).

• There is not a consistent procedure used by operators to perform chemical feed pump
calibrations and determine chemical dosages.

• There is no operational procedure describing the process for completing monthly reporting.

143

-------
Staffing/Number (Administration) (B)

•	Plant staff are responsible for operation and maintenance of both the water treatment and
wastewater treatment plants as well as distribution system maintenance, revenue collection,
and meter reading. This limits their availability for process control improvements and
adjustments.

•	During emergency conditions, such as distribution system line breaks, personnel are asked
to work long hours which stretch staff resources.

•	The current pay scale is not comparable to that offered by the surrounding water systems
which are competing for personnel with the same level of experience.

Process Controllability (Design) (B)

•	The filter design limits the backwash rate and does not allow the backwash flow rate to
be ramped up and down during a backwash.

•	There are no rate-of-flow controllers to limit the filtration rate of the filters.

Alarm Systems (Design) (B*)

•	There is no alarm system to notify operators if the PAC feed at the lake fails. The PAC
feed may be critical to remove cyanotoxins.

Sample Tap (Design) (C)

•	Each clear well receives a separate chlorine dose, resulting in two separate disinfection
zones. The current monitoring location does not monitor each clear well separately.

Each disinfection zone should be monitored and controlled to optimize its performance.

144

-------
EVALUATION FOLLOW-UP

State EPA has not established an approach for providing follow-up trainingto CPEs at the cur-
rent time. Additional HAB-focused developmental CPEs are planned at other the State's water
utilities over the next several months. Following these events, State EPA will be considering
follow-up strategies to support common CPE findings and performance-limiting factors. Plant
staff are encouraged to contact the state EPA staff regarding any questions or comments they
may have regarding specific findings from this CPE.

The XYZ WTP staff and management have specific challenges outlined in the Performance
Limiting Factors section above. Other issues pertaining to compliance were found that are
outside the scope of the CPE, which will be addressed separately.

This CPE has identified further areas that can be pursued to enhance particle removal perfor-
mance and be better prepared for future FLAB events. An excellent place to start the optimization
process is collecting and trending optimization data such as the approach demonstrated in the
Historical Water Quality Performance Assessment section of this report. The studies conducted
during this CPE also demonstrate a structured approach for conducting problem-solving
activities by plant staff. The following section includes several study ideas for plant staff to
consider and prioritize, based on benefits to plant operation and performance, level of
complexity, and available staff time.

Study Ideas

Study No. 1 — Investigating Impact of Filter Run Time on Filter Return to Service Turbidity
• Description:

¦ Conduct study to assess impact of reducing filter run time on filter backwash turbidity
recovery.

¦ Current run time on filters with newest media is about 80 hours.

• Benefits:

¦ Reduced solids loading to the filter media.

145

-------
¦	Potential for lower backwash recovery turbidity spikes.

¦	Understand if any design limitations exist.

Challenges

Solutions

None identified

Very doable by plant staff

Study No. 2 - Investigating Filter Backwash Procedure to Improve Filter Return-to-Service
Turbidity

• Description:

¦	Conduct study during routine backwashing to optimize backwash duration.

¦	Assess impact of longer filter backwash duration on filter recovery.

• Benefits:

¦	Potential for lower backwash recovery turbidity spikes

¦	Understand if any design limitations exist.

Challenges

Solutions

None identified

Very doable by plant staff

Study No. 3 - Determine Optimal PAC Dose and/or Type for a HAB Event
• Description:

¦ Conduct a simplified version of a PAC study which would evaluate treatment effective-
ness of PAC feed at the intake location accounting for the calculated approximate
9.5-hour detention time in the raw water transmission main. All four jars would have
the same detention time of 9.5 hours, but the PAC dosages could vary between jars.
For example, doses of 10, 20, 30, and 40 mg/L PAC could be evaluated. Samples could
be pulled and filtered at the 9.5-hour time step and analyzed for microcystins.

146

-------
¦	This exercise could also be repeated with different types of PAC. The type of PAC cur-
rently being used at the plant is a coal-basedPAC, but studies have shown different
treatment performance based on the type of PAC being used.

¦	If cyanotoxins are only present at lower concentrations (<5 |ig/L), the cyanobacteria in
the raw water could be concentrated using a phytoplankton net or similar device, and
then the concentrated sample could be added to the raw water sample used in the jar
test.

¦	In addition to microcystins testing, testing for pH, turbidity, TOC/DOC, phycocyanin,
etc. can be included to better understand the PAC performance for other water quality
parameters.

• Benefits:

¦ This approach could help determine optimal PAC dosage and/or PAC type for a HAB
event with higher microcystins concentration prior to a more severe event occurring.

Challenges

Solutions

Learning jar testing techniques

Obtain instructional videos (AWWA, others).

Obtain support from local water treatment plants.

Review jar testing spreadsheet provided by HAB CPE
team.

Review PAC Jar Testing Protocol developed by AWWA
(provided by HAB CPE team).

Obtaining samples of different types of PAC to
evaluate (coal, wood-based, coconut-based)

Reach out to PAC vendors for samples.

Lab support for microcystins analysis

Obtain training on ELISA procedure and develop in-
house capability.

State EPA or contract lab - Check State EPA's website
for list of contract labs.

147

-------
More Advanced Plant Studies

•	Investigating first-stage softening treatment followed by second-stage coagulation usingjar
testing.

•	Investigating impacts of recarbonation on settled water turbidity and stability (pH).

•	Determining optimum in-plantPAC, softening, and coagulant dosages.

148

-------