Professional Documents
Culture Documents
ASSESSMENT OF CURRENT
PRACTICES AND GAP ANALYSIS
TECHNICAL MEMORANDUM
Flint Drinking Water Distribution System Optimization
FINAL
March 2017
ASSESSMENT OF
CURRENT PRACTICES
AND GAP ANALYSIS
TECHNICAL
MEMORANDUM
Flint Drinking Water Distribution System
Optimization
Prepared for:
City of Flint
1101 S. Saginaw Street
Flint, Michigan 48502
Prepared by:
Our Ref.:
20616001.0000
Date:
March 2017
arcadis.com
ASSESSMENT OF CURRENT PRACTICES AND GAP ANALYSIS TECHNICAL MEMORANDUM
CONTENTS
Acronyms and Abbreviations ........................................................................................................................ vi
Executive Summary....................................................................................................................................... 1
3.0 Methodology.......................................................................................................................................... 28
arcadis.com i
ASSESSMENT OF CURRENT PRACTICES AND GAP ANALYSIS TECHNICAL MEMORANDUM
arcadis.com ii
ASSESSMENT OF CURRENT PRACTICES AND GAP ANALYSIS TECHNICAL MEMORANDUM
TABLES
Table ES-1: Distribution System Pipe Characteristics .................................................................................. 1
Table ES- 2: Summary of Industry Best Practices for Key Distribution System Optimization Categories .... 3
Table ES-3: Areas that Meet or Exceed Industry Standards ...................................................................... 10
Table ES- 4: Areas that Need Minor Strengthening .................................................................................... 11
Table ES- 5: Areas that Need Major Strengthening .................................................................................... 14
Table ES-6: Areas Where No Program Exists ............................................................................................ 17
Table 3-1: Summary of the Partnership for Safe Water Distribution System Improvement Variables
Assessed ..................................................................................................................................................... 29
Table 3-2: Summary of Industry Best Practices for Key Distribution System Optimization Categories...... 30
Table 3-3: Description of Key Distribution System Activities Performed by Regulatory Agencies ............. 37
Table 3-4: Summary of Information Reviewed ............................................................................................ 39
Table 4-1: Summary of Main Breaks per Geographic Location .................................................................. 62
Table 4-2: Main Break Classifications (Source: Kirmeyer et al., 2014) ....................................................... 74
Table 4-3: Summary of USEPA Sequential Sampling Rounds ................................................................... 81
Table 4-4: Summary of Paired and Unpaired Locations by Season ........................................................... 85
Table 4-5: Proposed Preventative Maintenance and Rehabilitation and Replacement Goals ................. 107
Table 4-6:Recommended Valve/Hydrant SOPs to be Developed ............................................................ 109
Table 4-7:Water Quality Problems Affected by Retention Time (Source: Brandt, 2004) .......................... 121
arcadis.com iii
ASSESSMENT OF CURRENT PRACTICES AND GAP ANALYSIS TECHNICAL MEMORANDUM
FIGURES
Figure ES-1: Gap Analysis Categories.......................................................................................................... 3
Figure 2-1: Purchased/Treated Water Flow (2010 – 2016)......................................................................... 27
Figure 3-1: Assessment and Gap Analysis Approach ................................................................................. 37
Figure 3-2: Gap Analysis Categories .......................................................................................................... 43
Figure 4-2: Monthly Free Chlorine Residual at TCR Sites and System Entry Point (2013 – 2016)............ 48
Figure 4-3: Monthly Free Chlorine Concentrations at USEPA Sites (Feb. – Nov. 2016) ............................ 49
Figure 4-4: Free Chlorine Concentrations at TCR Sites ............................................................................. 50
Figure 4-5: Free Chlorine Residual at TCR Sites (July – Dec. 2016) ......................................................... 50
Figure 4-6: Free Chlorine Residual at USEPA Sites (Feb. – Nov. 2016) .................................................... 51
Figure 4-7: Free Chlorine Residual at USEPA Sites (July – Nov. 2016) .................................................... 52
Figure 4-8: TTHM LRAAs at Historical Compliance Sites (2010 - 2016) .................................................... 53
Figure 4-9: HAA5 LRAAs at Historical Compliance Sites (2010 - 2016) ..................................................... 53
Figure 4-10: Individual TTHM Concentrations at Historical Compliance Sites (2013 - 2016) ..................... 54
Figure 4-11: HAA5 Concentrations at Historical Compliance Sites (2013 - 2016)...................................... 54
Figure 4-12: THM Concentrations at USEPA Sites (Mar. – Oct. 2016) ...................................................... 55
Figure 4-13: HAA Concentrations at USEPA Sites (Mar. – Oct. 2016) ....................................................... 56
Figure 4-14: Issues That Influence Frequency of Main Breaks................................................................... 58
Figure 4-15: City Main Break and Distribution System Pressure Data Availability ..................................... 58
Figure 4-16: Summary of Flint Water System Main Break Locations (2008 – 2015) .................................. 60
Figure 4-17: Flint Water System Main Breaks Plotted Annually ................................................................. 61
Figure 4-18: Main Breaks Per Year ............................................................................................................. 63
Figure 4-19: Summary of Main Breaks Per Month ...................................................................................... 64
arcadis.com iv
ASSESSMENT OF CURRENT PRACTICES AND GAP ANALYSIS TECHNICAL MEMORANDUM
Figure 4-20: Breaks Per Month Characterized according to their Percentage of the Total Breaks Per Year
.............................................................................................................................................................. 67
Figure 4-21: Breaks in 2014 and 2015 Compared to Trend from 2008 – 2013 .......................................... 68
Figure 4-22: Distribution System Pressure Plotted with Main Breaks (June – Aug. 2015; Flint WTP in
operation over entire period) ................................................................................................................. 70
Figure 4-23: Cedar Street Reservoir Operations Plotted with Main Breaks (June – Aug. 2015; Flint WTP in
operation over entire period) ................................................................................................................. 71
Figure 4-24: Selected Operation Summaries for West Side Reservoir Fill Valve ....................................... 73
Figure 4-25: Example profile showing lead concentration and sample volumes from USEPA Sequential
dataset .................................................................................................................................................. 84
Figure 4-26: Comparison of differential between upper and lower control limits for orthophosphate control
charts generated using all of 2016 data vs. post-July 1, 2016 data ..................................................... 90
Figure 4-27: Current and potential WQP monitoring sites .......................................................................... 93
Figure 4-28: Comparison of Conventional vs. Unidirectional Flushing ....................................................... 99
Figure 4-29: Automated flushing and manual flushing devices (photos courtesy of Matt Alexander,
USEPA Region 5) ............................................................................................................................... 100
Figure 4-30: Programmatic Approach for Cost-Effective Water Loss Control .......................................... 126
Figure 4-31: Utilities Department Organization Chart (May 2016) ............................................................ 136
APPENDICES
A Chlorine Residual Management Porgram
B Corrosion Control Figures
C Water Distribution Model Audit
D Interim Final Revised Total Coliform Rule Monitoring Plan
arcadis.com v
ASSESSMENT OF CURRENT PRACTICES AND GAP ANALYSIS TECHNICAL MEMORANDUM
arcadis.com vi
ASSESSMENT OF CURRENT PRACTICES AND GAP ANALYSIS TECHNICAL MEMORANDUM
arcadis.com vii
ASSESSMENT OF CURRENT PRACTICES AND GAP ANALYSIS TECHNICAL MEMORANDUM
arcadis.com viii
ASSESSMENT OF CURRENT PRACTICES AND GAP ANALYSIS TECHNICAL MEMORANDUM
EXECUTIVE SUMMARY
The City of Flint (City) purchases finished water from the Great Lakes Water Authority (GLWA), formerly
known as the Detroit Water and Sewerage Department (DWSD). For several decades GLWA provided
the sole source of supply for the City. However, on April 25, 2014 the City stopped purchasing water from
GLWA and began treating water at the Flint Water Treatment Plant (WTP), which is maintained by the
City and treats water from the Flint River as a backup emergency supply. Several water quality issues
soon arose, leading to a boil water advisory in August 2014, a Safe Drinking Water Act (SDWA) violation
related to total trihalomethanes (TTHM) in December 2014, an observed increase in Legionnaire’s
disease in the City, and an increase in lead levels at customers’ taps. As a result, the City resumed
purchasing water from GLWA on October 17, 2015.
A state of emergency was declared by the City of Flint on December 14, 2015, by the State of Michigan
on January 14, 2016, and by the President of the United States on January 16, 2016. The United States
Environmental Protection Agency (USEPA) issued an emergency administrative order on January 21,
2016 to address outstanding lead and copper rule (LCR) violations. As a part of this administrative order,
the City was to develop and implement a distribution system water quality optimization plan. The City has
retained a team led by Arcadis of Michigan, LLC (Arcadis) and including Environmental Engineering &
Technology, Inc. (EE&T), Confluence Engineering Group, LLC, and McConnell Communications, Inc. to
develop a Distribution System Optimization Plan, which consists of three main tasks: (1) assessment and
gap analysis, (2) resource analysis and needs assessment, and (3) development of the Plan.
The purpose of this report is to summarize the approach and findings from the first task, the Assessment
and Gap Analysis, which consisted of a review of the City’s existing distribution system practices and
procedures as compared to industry best practices to identify areas for improvement. Also as part of this
task, the Arcadis team worked with the City to submit a Revised Total Coliform Rule (RTCR) plan and
expand the number of unique weekly sampling sites from which data is gathered. New sample sites were
added to the original Total Coliform Rule (TCR) sites to create a sample pool that is more representative
of the entire distribution including residential areas and sensitive populations.
System Overview
The City currently purchases treated water from GLWA and boosts the chlorine and orthophosphate
residuals, and if needed, adjusts pH via caustic addition immediately prior to distribution. The distribution
system consists of approximately 580 miles of pipe, which is predominantly unlined cast iron, as shown in
Table ES-1.
Item Description
arcadis.com 1
ASSESSMENT OF CURRENT PRACTICES AND GAP ANALYSIS TECHNICAL MEMORANDUM
Item Description
96.6% is unlined cast iron pipe (564 miles)
Pipe Material 2.6% is ductile iron (31 miles) and
Remaining material includes steel, concrete, galvanized, and unidentified
63% installed between 1900 and 1930
Pipe Age 33% installed between 1930 and 1960
4% installed after 1960
The City’s distribution system includes two ground storage facilities and three pumps stations. Cedar
Street Reservoir, which houses Pumping Station 3, was constructed in 1948 and expanded in the 1960’s.
The reservoir has a total capacity of 20 MG. West Side Reservoir, which houses Pumping Station 5, was
constructed in 1970 and has a capacity of 12 MG. Both facilities have the ability to boost chlorine on the
fill line to the reservoir using liquid sodium hypochlorite. The City also maintains a third booster station
known as the Torrey Road Pumping Station. In addition to the distribution storage, the City has 26 MG
available storage at the currently unused WTP, for a total of 58 MG.
The population of Flint has dropped significantly since its peak in the 1960s. Most recently, it has
endured a 20 percent decrease in population since 2000, and has a current estimated population of
98,310. This decline has left the Water Department with an oversized distribution system, creating
physical, hydraulic and water quality challenges, including water age and chlorine residual management.
Approach
The Arcadis team coordinated with the appropriate regulatory agencies, reviewed data from historical
sources, and conducted multiple days of on-site interviews and site visits to assess the City’s existing
distribution system against industry best practices for optimization, specifically standards that have been
developed as the foundation for the Partnership for Safe Water Distribution System Optimization Program
(Friedman et al. 2010 and AWWA 2011). The Partnership for Safe Water’s program is comprised of a
methodology for utilities to optimize their distribution systems through a phased process of commitment to
the program, annual data reporting to the Partnership, self-assessment, and optimization, and to provide
a path for continuous improvement for even the most highly functioning system.
Using the Partnership’s self-assessment approach, the distribution system was evaluated based on three
major performance variables:
arcadis.com 2
ASSESSMENT OF CURRENT PRACTICES AND GAP ANALYSIS TECHNICAL MEMORANDUM
A summary of industry best practices for the optimization categories applicable to the City of Flint’s water
distribution system is presented in Table ES- 2.
Table ES- 2: Summary of Industry Best Practices for Key Distribution System Optimization Categories
arcadis.com 3
ASSESSMENT OF CURRENT PRACTICES AND GAP ANALYSIS TECHNICAL MEMORANDUM
arcadis.com 4
ASSESSMENT OF CURRENT PRACTICES AND GAP ANALYSIS TECHNICAL MEMORANDUM
arcadis.com 5
ASSESSMENT OF CURRENT PRACTICES AND GAP ANALYSIS TECHNICAL MEMORANDUM
Develop and maintain an asset inventory Friedman et al. (2010b). Criteria for
Pipeline Installation, Optimized Distribution Systems
Track and assess performance
Rehabilitation and AWWA. (2011c). Partnership for Safe
Replacement Develop a pipeline renewal program and Water Self-Assessment Guide for
written plan Distribution System Optimization
Friedman et al. (2010a). Assessment of
Assess existing conditions and vulnerability Inorganics Accumulation in Drinking
Address existing deposits Water System Scales and Sediments
Reduce contaminant and solids loading Friedman et al. (2010b). Criteria for
Clean areas of precipitation accumulation at Optimized Distribution Systems
least annually (i.e., storage tanks, low water AWWA. (2011c). Partnership for Safe
use areas) Water Self-Assessment Guide for
Distribution System Optimization
Post-Precipitation Iron < 0.1 mg/L (all points in the distribution
Control system) Kirmeyer et al. (2002). Guidance Manual
for Monitoring Distribution System Water
Manganese < 0.02 mg/L (all points in the Quality
distribution system)
Friedman et al. (2005). Development of
Turbidity 0.5 – 1.0 NTU (2.0 NTU at flushed
Distribution System Water Quality
hydrants)
Optimization Plans.
pH > 7.0 (unless State approved)
Smith. (2005). Water Distribution System
Calcium carbonate precipitation potential Assessment Workbook
(CCPP) within established guidelines
arcadis.com 6
ASSESSMENT OF CURRENT PRACTICES AND GAP ANALYSIS TECHNICAL MEMORANDUM
arcadis.com 7
ASSESSMENT OF CURRENT PRACTICES AND GAP ANALYSIS TECHNICAL MEMORANDUM
arcadis.com 8
ASSESSMENT OF CURRENT PRACTICES AND GAP ANALYSIS TECHNICAL MEMORANDUM
Notes:
1. Under the Partnership program, no optimization goals are provided for administrative topics (i.e., administrative policies, funding,
and staffing) and asset management. The Partnership recommends that utilities use benchmarks to form a quantitative basis for
assessing administrative topics.
Gaps
Through the analysis against the Partnership optimization method and other industry standards, gaps in
performance were identified for each of the categories evaluated. Tables ES-3 through ES-6 provide a
summary of the areas that meet or exceed industry standards, need minor strengthening, need major
strengthening, or do not have applicable programs, respectively.
arcadis.com 9
ASSESSMENT OF CURRENT PRACTICES AND GAP ANALYSIS TECHNICAL MEMORANDUM
arcadis.com 10
ASSESSMENT OF CURRENT PRACTICES AND GAP ANALYSIS TECHNICAL MEMORANDUM
arcadis.com 11
ASSESSMENT OF CURRENT PRACTICES AND GAP ANALYSIS TECHNICAL MEMORANDUM
arcadis.com 12
ASSESSMENT OF CURRENT PRACTICES AND GAP ANALYSIS TECHNICAL MEMORANDUM
arcadis.com 13
ASSESSMENT OF CURRENT PRACTICES AND GAP ANALYSIS TECHNICAL MEMORANDUM
arcadis.com 14
ASSESSMENT OF CURRENT PRACTICES AND GAP ANALYSIS TECHNICAL MEMORANDUM
arcadis.com 15
ASSESSMENT OF CURRENT PRACTICES AND GAP ANALYSIS TECHNICAL MEMORANDUM
arcadis.com 16
ASSESSMENT OF CURRENT PRACTICES AND GAP ANALYSIS TECHNICAL MEMORANDUM
arcadis.com 17
ASSESSMENT OF CURRENT PRACTICES AND GAP ANALYSIS TECHNICAL MEMORANDUM
arcadis.com 18
ASSESSMENT OF CURRENT PRACTICES AND GAP ANALYSIS TECHNICAL MEMORANDUM
Continue to revise and expand the current RTCR monitoring program beyond the City’s current
20 sites to 25 sites.
Implement surge control at reservoirs and pump stations, and conduct analysis of future WTP
operations on distribution system pressures.
Use of hydraulic model to evaluate water age and opportunities for minimizing.
Siting and installation of distribution system pressure data loggers to verify pressure control.
arcadis.com 19
ASSESSMENT OF CURRENT PRACTICES AND GAP ANALYSIS TECHNICAL MEMORANDUM
Purchase and install online distribution system water quality monitoring panels. Prior to
installation, use the hydraulic model to confirm optimal locations for installation.
Develop and implement a distribution system operator training program.
Develop and implement a UDF program.
Develop SOPs for routine maintenance activities and for those activities that impact water quality
(flushing, chlorine residual maintenance, etc.).
Increase funding for main replacement activities.
Conduct a local utility salary survey and adjust operator salaries as necessary to be competitive
in the local market.
Develop a hiring plan to fill vacant positions within the WSC Distribution Department.
Provide “whole house” flushing guidance to residents and encourage them to flush their homes
regularly until water quality within their home is restored.
The remaining gaps identified in this analysis will be grouped and combined with information obtained in
the upcoming Resource Analysis and Needs Assessment to develop a tailored set of improvements for
the City’s distribution system. The improvements will be prioritized using selection criteria and weightings
and will be presented in the final Distribution System Optimization Plan.
arcadis.com 20
ASSESSMENT OF CURRENT PRACTICES AND GAP ANALYSIS TECHNICAL MEMORANDUM
1.0 INTRODUCTION
The City of Flint (City) distributes drinking water to a population of approximately 98,310 through
approximately 580 miles of distribution system mains (Rowe, 2016a). The City currently purchases
finished water from the Great Lakes Water Authority (GLWA), formerly known as the Detroit Water and
Sewerage Department (DWSD), boosting the concentration of chlorine and orthophosphate for corrosion
control prior to distribution with capabilities to add sodium hydroxide for pH control as needed.
Additionally, the City owns and operates the Flint Water Treatment Plant (WTP) as a backup supply with
the ability to intake from the Flint River.
The City of Flint has retained a team led by Arcadis of Michigan, LLC (Arcadis) and including
Environmental Engineering & Technology, Inc. (EE&T), Confluence Engineering Group, LLC, and
McConnell Communications, Inc.to perform an analysis of system components and organizational
practices compared to industry standards and best practices to identify and prioritize necessary
improvements to optimize the distribution system.
A boil water advisory was issued on August 15, 2014 due to Escherichia coli (E. coli) detection in
distribution system sample.
A Safe Drinking Water Act (SDWA) tier two quarterly violation on was issued on December 16,
2014 due to concentrations of total trihalomethanes (TTHM) greater than the maximum
contaminant level (MCL) at four distribution sampling locations when calculated as a locational
running annual average (LRAA).
An increase of Legionnaire’s disease was observed in Genesee County including 42 potential
cases reported to the Michigan Department of Health and Human Services (MDHHS) (Flint
Advisory Task Force, 2016).
Rising levels of lead detected in distribution samples collected between July 2014 and July 2015.
arcadis.com 21
ASSESSMENT OF CURRENT PRACTICES AND GAP ANALYSIS TECHNICAL MEMORANDUM
The City resumed purchasing finished water from GLWA on October 17, 2015 because of the rising water
quality concerns. On December 9, 2015, the City began boosting orthophosphate at the WTP to re-
stabilize the distribution system piping network.
A state of emergency was declared by the City of Flint on December 14, 2015, by the State of Michigan
on January 14, 2016, and by the President of the United States on January 16, 2016.
On January 21, 2016, the United States Environmental Protection Agency (USEPA) issued an emergency
administrative order to address outstanding lead and copper rule (LCR) violations. The order was
directed to the Michigan Department of Environmental Quality (MDEQ), the City of Flint, and the State of
Michigan and stipulated that the respondents complete the following action items:
Create a publicly available website publishing information and sampling results relevant to the
lead crisis,
Respond to requests and recommendations by the USEPA Flint Task Force,
Provide water quality parameter measurements from distribution samples,
Provide an inventory of lead service lines, water interruptions, and unoccupied homes,
Cooperate with the USEPA LCR sampling,
Develop and maintain a chlorine residual throughout the distribution system,
Establish and maintain a corrosion control plan,
Ensure appropriate staffing at the WTP,
Develop and implement a distribution system water quality optimization plan, and
Assemble an Independent Advisory Panel (IAP) to make recommendations to ensure safe
drinking water.
In accordance with the order, the Arcadis team been retained to develop a Distribution System
Optimization Plan. Development of the Optimization Plan consists of three main tasks, assessment and
gap analysis, resource analysis and needs assessment, and development of the Plan. The purpose of
this report is to summarize Assessment and Gap Analysis, which included the following main tasks:
1. Assessment and gap analysis – This process involved reviewing the City of Flint’s existing
distribution system assets and practices and comparing to industry standards to identify areas for
improvement. This has been accomplished by a series of workshops in Flint with USEPA, IAP,
MDEQ, and City of Flint. Arcadis has performed a review of relevant reports and documents,
existing standard operating procedures (SOPs), and current plant practices including interviews
with staff at all levels within the organization.
2. Review of the hydraulic model – A hydraulic model was developed by USEPA for the City of Flint
to assist in water quality management. Arcadis analyzed this model against best practices and
provides recommendations on improvements in this memo.
3. Water quality data and sampling plan review – The Arcadis team provided extensive analysis of
available water quality parameter data including chlorine, lead, and orthophosphate.
4. Arcadis worked with the City to submit a Revised Total Coliform Rule (RTCR) plan and expand
the number of unique weekly sampling sites from which data is gathered. New sample sites were
added to the original Total Coliform Rule (TCR) sites to create a sample pool that is more
representative of the entire distribution including residential areas and sensitive populations.
arcadis.com 22
ASSESSMENT OF CURRENT PRACTICES AND GAP ANALYSIS TECHNICAL MEMORANDUM
5. In addition to assessment of chlorine residuals, pressure control and main break occurrences
were compared to industry optimization metrics.
6. Assessment of corrosion control and treatment – The Arcadis team analyzed all available lead
and corrosion control treatment data to evaluate the current standing of the City of Flint with
respect to LCR compliance. The Arcadis team also compared data gathered throughout the
crisis to large utilities across the United States for benchmarking.
Section 2.0 Water System Overview provides a snapshot of the current water system including
the current and future supply and treatment, demand, storage, booster stations and pressure
zones.
Section 3.0 Methodology outlines the approach taken to compare the existing practices to
industry best standards generally using the methods and criteria outlined in the Partnership for
Safe Water.
Section 4.0 Results reviews the various categories that were investigated and analyzed per the
Partnership for Safe Water described in Section 3. Operations and equipment are compared to
industry standards. This section also reviews and categorizes the identified gaps.
Section 5.0 Summary and Recommendations presents a summary of the approach and results
of the gap analysis and recommends early action items for immediate implementation.
arcadis.com 23
ASSESSMENT OF CURRENT PRACTICES AND GAP ANALYSIS TECHNICAL MEMORANDUM
2.1.1 Supply
When operational during quarterly test runs and during the transition from the GLWA pipeline to the KWA,
the WTP drew water from the Flint River via a 72-inch intake pipe. While the City of Flint was purchasing
water from GLWA, the 72-inch transmission main bypassed the plant and entered the distribution system
directly.
At the time of this report, the KWA pipeline is anticipated to be complete in late 2017. At that time, the
City of Flint plans to begin purchasing finished water from Genesee County (who will also receive water
from KWA) until the water plant can be upgraded to sufficiently treat raw water from Lake Huron.
2.1.2 Treatment
The primary processes at the Flint WTP are ozone, flocculation, sedimentation, softening, recarbonation,
filtration, and chlorine disinfection. The rated capacity of the plant is 24 million gallons per day (MGD)
(Sleeping Giant, 2016). Following intake from the Flint River, the ozone system consisting of two
generators and three contact basins pre-treats the raw water with a capacity of 36 MGD through ozone
gas injection (Rowe, 2016a). Following the ozone injection, the three-stage rapid mix and flocculation
process is accomplished using adjustable turbine mixers and ferric chloride with a capacity of 37 MGD
(Sleeping Giant, 2016).
The sedimentation process consists of three basins, each containing four 71-foot (ft). by 5-ft. plate settler
units. With an estimated loading rate of 4.0 gallons per minute per square foot (gpm/ft2), the
sedimentation capacity is estimated at approximately 24 MGD (Sleeping Giant, 2016). From the settling
tanks, process water is fed to two softening basins and slaked lime is added. Sludge from the lime
softening process is pumped to an onsite lagoon for disposal. Softened water is then recarbonated prior
to filtration.
Twelve 3 MGD filters housing sand and granular activated carbon (GAC) filter the process water prior to
final disinfection with chlorine. Disinfection is provided in a 3 million gallon (MG) clear well prior to
distribution.
It should be noted that the City is in the process of updating the equipment and treatment processes at
the plant. At the time of this report, these changes are anticipated to be in place by 2019.
arcadis.com 24
ASSESSMENT OF CURRENT PRACTICES AND GAP ANALYSIS TECHNICAL MEMORANDUM
Recently, the City has installed temporary sodium hypochlorite and phosphoric acid chemical feed
systems to boost the disinfectant and orthophosphate residuals, respectively, on the purchased water
from GLWA prior to distribution to the City. A 25% caustic soda chemical feed system has also been
installed to boost the pH of the purchased water prior to distribution, if needed.
arcadis.com 25
ASSESSMENT OF CURRENT PRACTICES AND GAP ANALYSIS TECHNICAL MEMORANDUM
collected by lab personnel at the same time as RTCR sampling. At the time of this report, West Side
Reservoir is out of service for leak repairs and inspection.
When considering the 26 MG available storage on the WTP property, the total available storage is 58
MG. According to plant staff, the water treatment plant began practicing deep cycling of its reservoirs to
manage water age in the distribution. Due to the high number of main breaks in the system, WSC and
WTP staff are generally uncomfortable with scaling down the amount of storage regularly used in the
distribution system.
Pipe Material
Pipe Age
arcadis.com 26
ASSESSMENT OF CURRENT PRACTICES AND GAP ANALYSIS TECHNICAL MEMORANDUM
98,310. This decline has left the Water Department with an oversized distribution system, creating
physical, hydraulic and water quality challenges, including water age and chlorine residual management.
Prior to the water crisis, system demands averaged 25.5 MGD from 2010-2013. During the water crisis,
demands dropped to 17.6 MGD, and have hovered near 13 MGD since the system reconnected to GLWA
(see Figure 2-1). Contributing to the decline in demand is the loss of commercial customer General
Motors in October 2014 over water quality concerns. As the water system stabilizes and water quality
concerns decrease, demand may increase. However, given continuing decreasing population trends,
demands may not return to pre-crisis levels.
arcadis.com 27
ASSESSMENT OF CURRENT PRACTICES AND GAP ANALYSIS TECHNICAL MEMORANDUM
3.0 METHODOLOGY
This section describes the methodology used to identify industry best practices for optimizing distribution
systems and to assess existing distribution system practices and protocols as compared to those best
practices.
Chlorine residual of at least 0.20 milligrams per liter (mg/L) and no greater than 4.0 mg/L in 95%
of all routine monthly readings,
Routine sample locations should include known problem areas and all storage facilities
No routine sample locations to have consecutive readings outside this range
Disinfection byproducts (DBPs) meet regulatory limits for each sample tested (not based on a
running annual average), specifically:
o TTHM ≤ 80 micrograms per liter (µg/L)
o Haloacetic acids five (HAA5) ≤ 60 µg/L
Physical Integrity. The physical soundness indicator is based upon the frequency of distribution system
main breaks and leaks. Optimization of infrastructure integrity includes the goal of meeting the following
criteria:
No more than 15 reported main breaks and leaks per 100 miles of pipeline per year, and
Reducing main break and leak frequency (based upon a rolling, 5-year trend).
Hydraulic Integrity. The hydraulic soundness indicator is based upon pressure management through the
distribution system. Pressure must be monitored continuously from sensors located within the distribution
system, ideally at low and high pressure locations. The goal for pressure management includes the
following:
arcadis.com 28
ASSESSMENT OF CURRENT PRACTICES AND GAP ANALYSIS TECHNICAL MEMORANDUM
All TTHM and HAA5 routine test results for 12 months with annual DBP trending indicating
running annual average,
Flushing procedures,
Valve and hydrant exercise and repair records,
Internal corrosion testing practices and records,
Pipeline renewal and replacement records,
Storage tank cleaning records,
Water age records indicating annual maximum water age at key locations, and
Distribution system map and asset inventory.
The improvement variables most relevant to and assessed as part of this Gap Analysis for the City of
Flint’s Water Distribution System are presented in Table 3-1. Administrative factors, specifically
administrative policies, funding, and staffing, in addition to asset management were also reviewed as part
of this Gap Analysis, but are not specifically listed in the table below as they impact all aspects of a water
utility, not just the distribution system.
Table 3-1: Summary of the Partnership for Safe Water Distribution System Improvement Variables Assessed
Water
Physical Hydraulic
Improvement Variable Quality
Integrity Integrity
Integrity
Cross-Connection Control
Customer Complaint Tracking and Response
Disinfection Byproduct Compliance
Flushing
Hydraulic Modelling
arcadis.com 29
ASSESSMENT OF CURRENT PRACTICES AND GAP ANALYSIS TECHNICAL MEMORANDUM
Water
Physical Hydraulic
Improvement Variable Quality
Integrity Integrity
Integrity
Installation, Operation and Maintenance (O&M) of Valves and
Hydrants
Internal Corrosion Control
Online Monitoring
Pipeline Installation, Rehabilitation and Replacement
Post-Precipitation Control
Pump Station Design, Operation and Maintenance
Security and Emergency Management
Storage Facility Design, Operation and Maintenance
Water Age Management
Water Loss Control
Water Quality Sampling and Response
Adapted from Water Research Foundation Criteria for Optimized Distribution Systems (Friedman et al., 2010) and AWWA
Partnership for Safe Water Self-Assessment Guide for Distribution System Optimization (AWWA, 2011).
Administrative factors, specifically administrative policies, funding, and staffing, in addition to asset management are included in the
Partnership program self-assessment process and were also reviewed as part of this work, but are not listed in the table above as
they impact all aspects of a water utility, not just the distribution system.
A summary of industry best practices for the optimization categories applicable to the City of Flint’s water
distribution system is presented in Table 3-2. As mentioned above, these are based on standards
presented in the AWWA Partnership for Safe Water Self-Assessment Guide for Distribution System
Optimization (AWWA, 2011), which includes principles from other key industry guidance including various
AWWA standards and manuals of practice, the Water Research Foundation (WRF) Criteria for Optimized
Distribution Systems (Friedman et al., 2010), and Recommended Standards for Water Works, better
known as ‘Ten States Standards’ published by the Great Lakes – Upper Mississippi River Board of State
and Provincial Public Health and Environmental Managers (GLUMRB) (GLUMRB, 2012). Both metric-
based (performance) and non-metric (process) measurements are included.
Table 3-2: Summary of Industry Best Practices for Key Distribution System Optimization Categories
arcadis.com 30
ASSESSMENT OF CURRENT PRACTICES AND GAP ANALYSIS TECHNICAL MEMORANDUM
arcadis.com 31
ASSESSMENT OF CURRENT PRACTICES AND GAP ANALYSIS TECHNICAL MEMORANDUM
arcadis.com 32
ASSESSMENT OF CURRENT PRACTICES AND GAP ANALYSIS TECHNICAL MEMORANDUM
Develop and maintain an asset inventory Friedman et al. (2010b). Criteria for
Pipeline Installation, Optimized Distribution Systems
Track and assess performance
Rehabilitation and AWWA. (2011c). Partnership for Safe
Replacement Develop a pipeline renewal program and Water Self-Assessment Guide for
written plan Distribution System Optimization
arcadis.com 33
ASSESSMENT OF CURRENT PRACTICES AND GAP ANALYSIS TECHNICAL MEMORANDUM
arcadis.com 34
ASSESSMENT OF CURRENT PRACTICES AND GAP ANALYSIS TECHNICAL MEMORANDUM
arcadis.com 35
ASSESSMENT OF CURRENT PRACTICES AND GAP ANALYSIS TECHNICAL MEMORANDUM
Notes:
1. Under the Partnership program, no optimization goals are provided for administrative topics (i.e., administrative policies, funding,
and staffing) and asset management. The Partnership recommends that utilities use benchmarks to form a quantitative basis for
assessing administrative topics.
arcadis.com 36
ASSESSMENT OF CURRENT PRACTICES AND GAP ANALYSIS TECHNICAL MEMORANDUM
Coordinate with
regulatory
agencies
Review avaialble
information
Table 3-3: Description of Key Distribution System Activities Performed by Regulatory Agencies
arcadis.com 37
ASSESSMENT OF CURRENT PRACTICES AND GAP ANALYSIS TECHNICAL MEMORANDUM
arcadis.com 38
ASSESSMENT OF CURRENT PRACTICES AND GAP ANALYSIS TECHNICAL MEMORANDUM
The Arcadis project team also began participating in the standing administrative order conference call
with USEPA, MDEQ and the City of Flint to obtain and provide updates on the various aspects of the
administrative order.
Performed/ Developed
Item Date
By
Data / Databases
Cedar Street Reservoir Online Pressure
June 2015 – March 2016 City of Flint
Data
Control Station No. 2 Online Chlorine
June 2015 – March 2016 City of Flint
Residual Data
Control Station No. 2 Online Pressure Data June 2015 – March 2016 City of Flint
DBP Compliance Data January 2010 – December 2016 City of Flint
Enhanced Water Quality Parameter Data
January 2010 – December 2016 City of Flint
(Quarterly)
Enhanced Water Quality Parameter Data
November 2015 – December 2016 City of Flint
(Weekly)
Heterotrophic Plate Count Data January – July 2016 City of Flint
Hydrant Database (Excel) Unknown City of Flint
Investigative Chlorine Residual Monitoring
March – November 2016 USEPA
Data
LCR Compliance Tap Sampling Data January 1992 – June 2015 City of Flint
Main Break Tracking Spreadsheet January 2008 – February 2016 City of Flint
Meter Spreadsheet (Active Accounts Only) Unknown City of Flint
Monthly Operating Reports January 2010 – December 2016 City of Flint
Recurring Chlorine Residual Monitoring
March – November 2016 USEPA
Data
Recurring DBP Monitoring Data March – October 2016 USEPA
Residential Lead Sampling Results September 2015 – October 2016 MDEQ
Sentinel Testing Results February – September 2016 MDEQ
Sequential Lead Sampling Results January – November 2016 USEPA
TCR Chlorine Residual Data January 2010 – December 2016 City of Flint
arcadis.com 39
ASSESSMENT OF CURRENT PRACTICES AND GAP ANALYSIS TECHNICAL MEMORANDUM
Performed/ Developed
Item Date
By
Valve Database (Excel) Unknown City of Flint
Water Service Center Inventory June 2016 City of Flint
West Side Reservoir Online Pressure Data December 2015 – March 2016 City of Flint
Relevant Reports
ROWE Professional
Asset Management Report October 2016
Services Company
Cedar Street Reservoir Sampling Study
November 2016 USEPA
Summary of Findings
City of Flint Strategic Plan 2014-2019 Unknown City of Flint
City of Flint Strategic Plan 2016-2020 Unknown City of Flint
City of Flint Strategic Plan Second Quarter
Unknown City of Flint
Report Oct 1 – Dec 2015
Quality Assurance Plan 2016 City of Flint
Distribution System Water Quality
Unknown City of Flint
Parameter Monitoring Plan
Drinking Water Revolving Fund – Draft ROWE Professional
April 2016
Project Plan Services Company
EchoShore-M Leak Detection Report June 2016 Echologics
Elevated Tank Inspection Report June 2011 Utility Service Co.
Emergency Response Plan January 2013 City of Flint
ePulse Condition Assessment Report June 2016 Echologics
Center for Public Safety
Fire and EMS Operational Report December 2014
Management
Flint Administrative Order January 2016 USEPA
Flint Amended Administrative Order November 2016 USEPA
Flint Water Advisory Task Force Final Flint Water Advisory Task
March 2016
Report Force
Flint Water Quality Report March 2015 Veolia
Flint Water Rate Analysis May 2016 Raftelis
Flint Water Treatment Plant Evaluation April 2016 Sleeping Giant
KWA Supply Options to Flint Water Lockwood, Andrews &
February 2016
Treatment Plant Newnam, Inc.
KWA Supply to Flint Water Treatment Plant Lockwood, Andrews &
June 2016
- Coagulant Study Newnam, Inc.
Laboratory Quality Assurance Plan Unknown City of Flint
Operational Evaluation Report – Lockwood, Andrews &
August 2015
Trihalomethane Formation Concern Newnam, Inc.
Sample Siting Plan – Bacteriological June 2016 MDEQ
arcadis.com 40
ASSESSMENT OF CURRENT PRACTICES AND GAP ANALYSIS TECHNICAL MEMORANDUM
Performed/ Developed
Item Date
By
Sanitary Survey Evaluation December 2012 – January 2013 MDEQ
Valve Work Order Reports May – July 2015 WachsWater Services
Violation Notice – MCL for TTHM
Operational Evaluation – 4th Quarter 2014 December 2014 MDEQ
Monitoring Period
Water Audit and Non-Revenue Water Final
September 2016 Echologics
Report
Water Audit and Non-Revenue Water
September 2016 Echologics
Summary Report
Water Reliability Study December 2013 Rowe
Water Service Center Fiscal Year 2016
October 2016 City of Flint
Report
Tucker, Young, Jackson,
Water Supply Assessment February 2013
Tull Inc.
Water Valve Assessment, Mapping and
May – July 2015 WachsWater Services
Data Management Program
Other
City of Flint Routine Distribution System
Sampling Sites & Additional Water Quality Unknown City of Flint
Monitoring Sites
City of Flint Utilities Organizational Chart Unknown City of Flint
City of Flint Utilities Position List September 2016 City of Flint
City of Flint Water Treatment Plant
October 2016 City of Flint
Organizational Chart
Corrective Work Order Maintenance Form Unknown City of Flint
Recent update to original
Distribution System Hydraulic Model 2016
model performed by USEPA
Select SOPs Unknown City of Flint
Water Distribution Map 1980 Black & Veatch
Water Information Management Solution
and Online Water Quality Monitoring
2016 Hach
Equipment Quotations and Draft
Specification
Water Main and Curb Boxes Maps Unknown City of Flint
Water Supply Map January 1991 City of Flint
arcadis.com 41
ASSESSMENT OF CURRENT PRACTICES AND GAP ANALYSIS TECHNICAL MEMORANDUM
WSC Supervisor
WSC Deputy Supervisor
WTP Supervisor
WTP Operations Supervisors
WTP Operations Foremen
WTP Operator / Maintainers
WTP Maintenance Supervisors
Instrument Technician
Electrician/Supervisory control and data acquisition (SCADA) Technician
Water Quality and Laboratory Supervisor (interim)
Plumbing and Mechanical Inspector (interview conducted via phone)
In addition, field visits were held during this period to conduct visual inspections of assets and observe
selected process, practices, data and systems. Facilities toured included:
arcadis.com 42
ASSESSMENT OF CURRENT PRACTICES AND GAP ANALYSIS TECHNICAL MEMORANDUM
arcadis.com 43
ASSESSMENT OF CURRENT PRACTICES AND GAP ANALYSIS TECHNICAL MEMORANDUM
4.0 RESULTS
This section presents a summary of the City’s current programs, practices and tools in place along with
key findings and gaps based on a comparison to industry best practices (refer to Section 3.1) for each
optimization topic assessed. The topics have been organized to first present the three performance
indicators that represent the distribution system integrity based on the Partnership for Safe Water
program (i.e., disinfectant residual, main breaks, pressure management) followed by internal corrosion
control, which is critical to public health. Subsequent sections address the relevant influencing variables
from the Partnership program plus several supporting practices and programs essential to utility
management and operation.
4.1.1 Background
Maintenance of a disinfectant residual is critical to ensuring the integrity of the distribution system. In
addition to providing microbial control, disinfectant residuals provide oxidizing conditions to help stabilize
pipe scales, and can serve as an indicator of distribution system integrity. Therefore, a key aspect of
distribution system water quality management and optimization is to identify appropriate disinfectant
residual level(s) and strategies for monitoring and maintaining them on an on-going basis. Many factors
may influence the disinfectant residual, or may be influenced by disinfectant residual, as shown in Figure
4-1. With the exception of nitrification, all of the factors identified in this diagram have had a significant
role in the City’s distribution system chlorine residual management.
MDEQ regulations are consistent with the
Surface Water Treatment Rule (SWTR)
and require that no more than 5% of
monthly samples have an undetectable
disinfectant concentration for any two
consecutive months. Furthermore,
disinfectant concentrations are not
permitted to exceed maximum residual
disinfectant levels in samples collected for
DBP analysis.
Some states have set or are proposing
minimum numeric values for chlorine
residual. These include Colorado (0.2
Figure 4-1: Chlorine Residual Influence Diagram
mg/L), Minnesota (0.1 mg/L), Nebraska (Source: Friedman et al., 2010)
(0.1 mg/L), and Pennsylvania (0.3 mg/L,
proposed).
Compliance with disinfection byproduct MCLs is based on locational running annual averages. MCLs are
set at 80 µg/L for the sum of four regulated trihalomethanes (TTHM) and 60 µg/L for the sum of five
regulated haloacetic acids (HAA5).
arcadis.com 44
ASSESSMENT OF CURRENT PRACTICES AND GAP ANALYSIS TECHNICAL MEMORANDUM
Many factors can affect levels of DBPs in the distribution system, including disinfectant type, dosage,
contact time, and residual concentrations, point(s) of application, presence and characteristics of natural
organic matter (NOM) in the water supply, pH, temperature, and water age. Some strategies for
optimizing DBPs in the distribution system include:
Identification of sample sites representative of system entry point(s) and distribution system
Chlorine residual ≥0.2 mg/L and ≤4.0 mg/L in 95% of samples collected on a monthly basis
Inclusion of sample sites that target known problem areas and all storage facilities
No consecutive monthly readings outside performance goals
DBPs meet regulatory limits for all samples collected
arcadis.com 45
ASSESSMENT OF CURRENT PRACTICES AND GAP ANALYSIS TECHNICAL MEMORANDUM
arcadis.com 46
ASSESSMENT OF CURRENT PRACTICES AND GAP ANALYSIS TECHNICAL MEMORANDUM
during the re-stabilization process and to assist with control of microbial growth beyond the customer’s
meter.
Variability in chlorine measurements was well-controlled prior to the switch to Flint River as shown by the
height of the bars and overall range in percentiles in Figure 4-2. Variability in chlorine measurements has
decreased considerably since switching back from Flint River water to GLWA, but is still somewhat
greater than pre-Flint River. It is expected that as the system continues to stabilize, and boosting
operations at the reservoirs are optimized, the variability in chlorine measurements at specific sites and
across the system will decrease. Opportunities for tighter control of variability and minimizing peaks while
still meeting minimum residual goals are being evaluated under the Chlorine Residual Management
Program.
arcadis.com 47
ASSESSMENT OF CURRENT PRACTICES AND GAP ANALYSIS TECHNICAL MEMORANDUM
Figure 4-2: Monthly Free Chlorine Residual at TCR Sites and System Entry Point (2013 – 2016)
arcadis.com 48
ASSESSMENT OF CURRENT PRACTICES AND GAP ANALYSIS TECHNICAL MEMORANDUM
3.0
2.5
Free Chlorine Residual (mg/L)
2.0
1.5
1.0
0.5
0.0
Feb‐16 Mar‐16 Apr‐16 May‐16 Jun‐16 Jul‐16 Aug‐16 Sep‐16 Oct‐16 Nov‐16
Figure 4-3: Monthly Free Chlorine Concentrations at USEPA Sites (Feb. – Nov. 2016)
The following figures present site-specific data during the three operational periods (before, during, after
switch) that were developed to identify potential problem areas and will be used to identify site-specific
and area-specific solutions (use of autoflushers, confirmation of valve positions, water age management,
etc.). Figure 4-4 shows generally higher concentrations and more variable levels between and within
TCR sites post Flint River compared to pre-Flint River. However, when focusing on just the most recent
six months of data since boosting has been modified (July through December 2016), the 95th percentile
residual levels are significantly elevated, and variability within each site has decreased. These results are
shown in Figure 4-5. Monthly and site-specific trends will continue to be evaluated as part of the Chlorine
Residual Management Program.
Figure 4-6 shows the history of site-specific chlorine residual levels from the 24 USEPA sites, and Figure
4-7 focuses on the most recent data (July through November 2016). These graphs show significant
variability between sites, but decreasing variability within sites (smaller bars) in the more recent results.
Opportunities for site-specific and area specific improvements will be identified with the goal of increasing
minimum residual levels while hopefully reducing maximum levels to enhance customer acceptance of
the delivered water in order to meet disinfection goals.
arcadis.com 49
ASSESSMENT OF CURRENT PRACTICES AND GAP ANALYSIS TECHNICAL MEMORANDUM
3.0
Flint River Operational Period
2.5
Free Chlorine Residual (mg/L)
2.0
1.5
1.0
0.5
0.0
Plant Tap
Plant Tap
Plant Tap
TCR Site WS
TCR Site WS
TCR Site 1
TCR Site 2
TCR Site 3
TCR Site 4
TCR Site 5
TCR Site 6
TCR Site 7
TCR Site 8
TCR Site CS
TCR Site WS
TCR Site 1
TCR Site 2
TCR Site 3
TCR Site 4
TCR Site 5
TCR Site 6
TCR Site 7
TCR Site 8
TCR Site CS
TCR Site 1
TCR Site 2
TCR Site 3
TCR Site 4
TCR Site 5
TCR Site 6
TCR Site 7
TCR Site 8
TCR Site CS
Period 1: Pre‐Flint River (Jan 1, 2010 ‐ Apr 24, Period 2: Flint River (Apr 25, 2014 ‐ Oct 16, Period 3: Post‐Flint River (Oct 17, 2015 ‐
2014) 2015) present)
3.0
2.5
Free Chlorine Residual (mg/L)
2.0
1.5
1.0
0.5
0.0
Figure 4-5: Free Chlorine Residual at TCR Sites (July – Dec. 2016)
arcadis.com 50
ASSESSMENT OF CURRENT PRACTICES AND GAP ANALYSIS TECHNICAL MEMORANDUM
3.0
2.5
Free Chlorine Residual (mg/L)
2.0
1.5
1.0
0.5
0.0
Figure 4-6: Free Chlorine Residual at USEPA Sites (Feb. – Nov. 2016)
arcadis.com 51
ASSESSMENT OF CURRENT PRACTICES AND GAP ANALYSIS TECHNICAL MEMORANDUM
3.0
2.5
Free Chlorine Residual (mg/L)
2.0
1.5
1.0
0.5
0.0
Figure 4-7: Free Chlorine Residual at USEPA Sites (July – Nov. 2016)
arcadis.com 52
ASSESSMENT OF CURRENT PRACTICES AND GAP ANALYSIS TECHNICAL MEMORANDUM
140
120
100
LRAA (ug/L)
80
60
40
20
Flint River
Compliance is RAA until Stage 2 Operational Period
monitoring begins in 10/2012
0
Q1 Q2 Q3 Q4 Q1 Q2 Q3 Q4 Q1 Q2 Q3 Q4 Q1 Q2 Q3 Q4 Q1 Q2 Q3 Q4 Q1 Q2 Q3 Q4 Q1 Q2 Q3 Q4
2010 2011 2012 2013 2014 2015 2016
MCDONALDS, 3719 DAVISON RD BP GAS, 822 S DORT HWY
7‐11/Liquor Palace, 3302 S DORT HWY TACO BELL, 3606 CORUNNA RD
2501 FLUSHING RD SALEMS HOUSING, 3216 MLK BLVD
RITE AID, 5018 CLIO RD NORTH FLINT AUTO, 6204 N SAGINAW ST
MCL Stage 1 RAA
70
Flint River
Operational Period
60
50
LRAA (ug/L)
40
30
20
10
Compliance is RAA until Stage 2
monitoring began in 10/2012
0
Q1 Q2 Q3 Q4 Q1 Q2 Q3 Q4 Q1 Q2 Q3 Q4 Q1 Q2 Q3 Q4 Q1 Q2 Q3 Q4 Q1 Q2 Q3 Q4 Q1 Q2 Q3 Q4
2010 2011 2012 2013 2014 2015 2016
MCDONALDS, 3719 DAVISON RD BP GAS, 822 S DORT HWY, FLINT
7‐11/Liquor Palace, 3302 S DORT HWY TACO BELL, 3606 CORUNNA RD
2501 FLUSHING RD SALEMS HOUSING, 3216 MLK BLVD
RITE AID, 5018 CLIO RD NORTH FLINT AUTO, 6204 N SAGINAW ST
MCL Stage 1 RAA
arcadis.com 53
ASSESSMENT OF CURRENT PRACTICES AND GAP ANALYSIS TECHNICAL MEMORANDUM
200
Flint River
180 Operational Period
160
TTHM Concentration (ug/L)
140
120
100
80
60
40
20
0
Q1 Q2 Q3 Q4 Q1 Q2 Q3 Q4 Q1 Q2 Q3 Q4 Q1 Q2 Q3 Q4
2013 2014 2015 2016
MCDONALDS, 3719 DAVISON RD BP GAS, 822 S DORT HWY
7‐11/Liquor Palace, 3302 S DORT HWY TACO BELL, 3606 CORUNNA RD
2501 FLUSHING RD SALEMS HOUSING, 3216 MLK BLVD
RITE AID, 5018 CLIO RD NORTH FLINT AUTO, 6204 N SAGINAW ST
MCL
Figure 4-10: Individual TTHM Concentrations at Historical Compliance Sites (2013 - 2016)
70
Flint River
Operational Period
60
HAA5 Concentration (ug/L)
50
40
30
20
10
0
Q1 Q2 Q3 Q4 Q1 Q2 Q3 Q4 Q1 Q2 Q3 Q4 Q1 Q2 Q3 Q4
2013 2014 2015 2016
MCDONALDS, 3719 DAVISON RD BP GAS, 822 S DORT HWY
7‐11/Liquor Palace, 3302 S DORT HWY TACO BELL, 3606 CORUNNA RD
2501 FLUSHING RD SALEMS HOUSING, 3216 MLK BLVD
RITE AID, 5018 CLIO RD NORTH FLINT AUTO, 6204 N SAGINAW ST
MCL
arcadis.com 54
ASSESSMENT OF CURRENT PRACTICES AND GAP ANALYSIS TECHNICAL MEMORANDUM
The high levels of THMs found in samples collected after the switch to Flint River water resulted in
several water quality violations. Following the return to Detroit water, the levels of THMs and HAAs
dropped to levels consistent with those measured before the switch. DBP concentrations in water sourced
from GLWA are consistently less than 50% of the MCL for both groups of DBPs, despite increased
chlorine concentrations throughout the distribution system from chlorine boosting at the plant and
reservoirs.
TTHM and HAA5 concentrations seasonally increase in the summer; a result of increased temperatures
and seasonally higher raw water NOM. This trend is observed during the 3rd quarter sample collection
that occurs in August of each year, regardless of which source of supply is used or disinfection boosting
and operational targets.
Monthly concentrations of THMs and HAAs at 24 sites sampled by USEPA are presented in Figure 4-12
and Figure 4-13. These sites are not DBP compliance locations; however, they do provide more
information on DBP behavior throughout the distribution system. No single location tested by the USEPA
since March 2016 has exceeded an MCL for TTHM or HAA5. Moreover, the average concentrations for
TTHM and HAA5 have remained below 50% of their respective MCLs, despite increased chlorine dosing
at the WTP and distribution system reservoirs.
Based on the review of DBP compliance data, DBP data collected at USEPA locations, and chlorine data,
it is recommended that the City maintain the current operational approach. Additional operations
improvements are not needed at this time to optimize DBP levels. Optimization of DBP levels should be
reviewed following the switch to water from KWA to determine if improvements are needed.
100
90
TTHM Concentration (µg/L)
MCL = 80 µg/L
80
70
60
50
40
Max
30
Ave
20
Min
10
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
EPA Site
arcadis.com 55
ASSESSMENT OF CURRENT PRACTICES AND GAP ANALYSIS TECHNICAL MEMORANDUM
100
90
HAA5 Concentration (µg/L)
80
70
MCL = 60 µg/L
60
50
40
30
Max
20
Ave
10
Min
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
EPA Site
4.1.4 Summary
The City is currently meeting the Partnership optimization goal of ≥ 0.2 mg/L in 95% of samples
collected on a monthly basis, included challenging locations within the distribution system.
DBP levels are below 50% of MCLs at compliance locations and at additional more challenging
locations monitored by USEPA during 2016.
The City will likely qualify for reduced DBP monitoring under the Stage 2 DBPR, even with
chlorine boosting at the WTP and reservoirs.
Begin recording chlorine residual concentrations measured when collecting samples for DBP
analysis.
Track chlorine residual management performance against optimization criteria using spreadsheet
tool (enter measurements on weekly basis from plant entry point, coliform sampling sites, DBP
sampling sites, and additional surveillance sampling sites)
Identify optimal chlorine dosing to target 0.2- 0.5 mg/L in 95% of samples throughout distribution
system
Identify seasonal optimal boosting strategies at plant and reservoirs
arcadis.com 56
ASSESSMENT OF CURRENT PRACTICES AND GAP ANALYSIS TECHNICAL MEMORANDUM
Continue flushing in areas where maintaining a detectable chlorine residual has been challenging
and identify opportunities to improve residual levels without further increasing dosages
Continue efforts and approach outlined in Chlorine Residual Management Program to optimize
residual levels while minimizing variability within and between sites to improve customer
acceptance of water throughout distribution system while meeting disinfectant residual goals.
4.2.1 Background
The analysis of main breaks falls under the guidance outlined in the Partnership’s physical integrity
category. As shown below in Figure 4-14, there are many operations practices, infrastructure
characteristics, and environmental factors that may influence a City’s infrastructure integrity. This
assessment is based upon the frequency of distribution system breaks and can be characterized as
optimized if the following criteria are met:
Main break and leak frequency to be no greater than 15 for each 100 miles of utility controlled
pipeline, and
Reducing main break frequency (based upon a rolling, 5-year trend) is an indicator of progress
toward optimization.
arcadis.com 57
ASSESSMENT OF CURRENT PRACTICES AND GAP ANALYSIS TECHNICAL MEMORANDUM
There are also key minimum data requirements recommended by the Partnership for utilities to complete
self-assessment of their distribution system. Some of these requirements include that main break records
should be kept for a running 10-year period, valve and hydrant exercise and repair records should be
maintained, and systems should maintain a distribution system map and asset inventory. Recorded data
should include pipe location, pipe material, pipe size, apparent type of leak or break, soil type, pipe’s
depth, and saturation conditions.
The objectives of this analysis were to compare City practices to the Partnership goals/criteria, try to
identify potential causes of main breaks and opportunities for improvement, and if possible, to identify
impacts of the 2014/2015 switch to the Flint river supply on infrastructure integrity.
Figure 4-15: City Main Break and Distribution System Pressure Data Availability
Data provided for this assessment includes information on main breaks, pressure monitoring, and water
reservoir storage facility operations. Main break data supplied includes a period from January 2008
through February 2016. Data from the City’s SCADA system was provided for several locations from June
2015 through March 2016 (note that West Side Reservoir pressure data prior to December 2015 was not
available). On-line viewable access to pressure information was available from a period starting June
2016 through December 2016, but this information was viewable only (e.g., data could not be
arcadis.com 58
ASSESSMENT OF CURRENT PRACTICES AND GAP ANALYSIS TECHNICAL MEMORANDUM
downloaded from this on-line viewer). As shown in Figure 4-15, there is only a small window (June 2015 –
December 2015) for which both main break and SCADA data overlap.
Because of data gaps, assessment of the City water system should be considered incomplete at this
time. In general, main break data and pressure data do not coexist for a period allowing for sufficient
trending and/or correlative analysis to identify operations impacts on main breaks. A summary of
challenges in assessing water system main breaks due to data gaps include the following:
arcadis.com 59
ASSESSMENT OF CURRENT PRACTICES AND GAP ANALYSIS TECHNICAL MEMORANDUM
Note: Breaks located outside of shaded areas were associated with the nearest adjacent shaded area.
Figure 4-16: Summary of Flint Water System Main Break Locations (2008 – 2015)
arcadis.com 60
ASSESSMENT OF CURRENT PRACTICES AND GAP ANALYSIS TECHNICAL MEMORANDUM
arcadis.com 61
ASSESSMENT OF CURRENT PRACTICES AND GAP ANALYSIS TECHNICAL MEMORANDUM
These data are also summarized below in Table 4-1. This table summarizes the breaks per year, per
geographic location. The difference between what is identified geographically and the total breaks
identified is due to the lack of known location for several breaks. The far-right column lists percent of
known geographic locations for breaks compared to the total list of breaks, indicating that only between
74 to 84% of the breaks can be shown graphically while the remaining number cannot be graphed without
additional effort to pinpoint street addresses.
At this time, the breaks per area seem to show the following characteristics:
On average, the lowest number of breaks occur in the Southeast area of the system (far from the
WTP and prior to water entering reservoir storage),
On average, the greatest number of breaks occur in the Northeast area of the system (nearest to
the WTP), and
Main breaks increased within all areas in 2014 and 2015, as discussed further below.
Main breaks should be evaluated further based upon proper hydraulic partitioning of the system (i.e.,
areas influenced by the WTP, by West Side Reservoir, by Cedar Street Reservoir, etc.). This information
should be done with aid from the City’s hydraulic model. Additionally, analysis of pipe age (installation
date) should be included.
arcadis.com 62
ASSESSMENT OF CURRENT PRACTICES AND GAP ANALYSIS TECHNICAL MEMORANDUM
over time. Figure 4-18 also compares the annual number of breaks to Partnership for Safe Water and
Criteria for Optimized Distribution System goal of ≤ 15 reported breaks and leaks per 100 miles of utility
controlled distribution system mains. Currently, there is inconsistency with regard to the actual number of
pipe miles in the Flint system, ranging somewhere between 580-800 miles. The optimization goal for 800
miles of pipe is 120 breaks per year, compared to 87 breaks per year for 580 miles. In either case, the
City does not currently (nor historically) meet the optimization goal.
Total Breaks (No.)
arcadis.com 63
ASSESSMENT OF CURRENT PRACTICES AND GAP ANALYSIS TECHNICAL MEMORANDUM
90
Monthly Range between Median Main Breaks Flint River Operations
High and Median Low Temperature
82
69 70
60
58
48
45 44
41 40 41
36 35
34 33
30 29
28 27 28
26 2626
23 23 23 23 Monthly
20 21 21
19 19 19 19 1818 Partnership
17 16 16 16 17 17 18 Goal
14 1413 15 14 15 14
12 13 12 13
10 10 10 800 mile system
879 8 9
7 879 8 8 9 8 9 9 9 8
666 6 56 6 6 6 675 5 5 5
7
5 580 mile system
4 3 3 3 4 44
2 2
0
Ja
Ap
Ju
Ja
Ap
Ju
Ja
Ap
Ju
Ja
Ap
Ju
Ja
Ap
Ju
Ja
Ap
Ju
Ja
Ap
Ju
Ja
Ap
Ju
Ja
Ap
Ju
O
ct
ct
ct
ct
ct
ct
ct
ct
ct
n
l
r
r
2008 2009 2010 2011 2012 2013 2014 2015 2016
arcadis.com 64
ASSESSMENT OF CURRENT PRACTICES AND GAP ANALYSIS TECHNICAL MEMORANDUM
This figure clearly shows that the peak number of breaks occur during cold weather months. The coldest
sustained temperatures within the dataset occurred in the winter of 2014 season (average median low of
11 oF sustained for two months). The coldest individual month occurred in February 2015. The warmest
winter occurred in the 2015/2016 season and had one of the lowest rates of winter breaks/month (based
on data available through January 2016).
Figure 4-20 further illustrates how the number of breaks per month peak during January, February, and
March. Number of breaks occurring in each month are shown along with an indication as to what
percentage those breaks are compared to the entire year. The average number of breaks occurring in the
months of January, February, and March ranged from 30 (January) to 38 (February) per month since
2008. Breaks occurring in these months on average account for roughly 50% of total annual breaks. The
lowest average percentage of the total breaks per year occur in May, June, October, and November
(each at 4% of the total breaks per year).
Figure 4-21 compares breaks in 2014 and 2015 to the trend of the breaks per month for the grouping of
all of 2008 through 2013. The 2008-2013 trend shows the typical low number of breaks from April through
November (nearly meeting the Partnership for Safe Water goal), a slight increase in December, and the
peak in breaks occurring in January, February, and March. Winter months in 2014 and 2015 were colder
than usual, likely adding to the increase in breaks during those two winter periods. However, the number
of breaks during the warm months of 2015, specifically June through October, seem to be greater than
the average and therefore possibly due to pressure impacts associated with WTP operation.
arcadis.com 65
ASSESSMENT OF CURRENT PRACTICES AND GAP ANALYSIS TECHNICAL MEMORANDUM
arcadis.com 66
ASSESSMENT OF CURRENT PRACTICES AND GAP ANALYSIS TECHNICAL MEMORANDUM
Figure 4-20: Breaks Per Month Characterized according to their Percentage of the Total Breaks Per Year
arcadis.com 67
ASSESSMENT OF CURRENT PRACTICES AND GAP ANALYSIS TECHNICAL MEMORANDUM
90
2008-2013 Average + St Dev
2014
2015
60
30
Partnership Goal for 800 Mile System
Partnership Goal for 580 Mile System
0
Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec
Month
Figure 4-21: Breaks in 2014 and 2015 Compared to Trend from 2008 – 2013
Figure 4-22 and Figure 4-23 are compilations of several different SCADA monitoring locations along with
a summary of main breaks that occurred in June through August 2015. Specifically, these figures contain
the following information:
Main Breaks
Illustrated on the days they occurred, represented by the colored bubbles at the top of the figures,
Each bubble indicates pipe size of the break (e.g., ≥12 = a break in pipe of diameter ≥12-inches;
8 = a break in pipe of diameter 8-inches),
Pressure
Distribution system entry point pressure was recorded as leaving the WTP,
Cedar Street Reservoir Header Pressure,
Valve Operations
Cedar Street Reservoir Fill Operations (including an opening and a closing of the fill valve) is
indicated by a blue square marker,
When a fill cycle was conducted with an excessively-rapid opening or closing of the fill valve, this
is denoted by changing the color of the marker to red.
arcadis.com 68
ASSESSMENT OF CURRENT PRACTICES AND GAP ANALYSIS TECHNICAL MEMORANDUM
With regard to determining if a fill valve opening or closing was excessively rapid, the goal was to
determine if a rapid operation could be correlated with a main break. Therefore, several different
opening/closing times (e.g., 10 minutes, 5 minutes, 1 minute, shorter than 1 minute, etc.) were evaluated
to determine if they produced trends. Note that the fill valve analysis was conducted on both West Side
and Cedar Street Reservoir, but only Cedar Street Reservoir is shown in the above figure. Attempts to
evaluate header pressure readings from West Side Reservoir were unsuccessful due to the unavailability
of data. Data for this location was only available from December 2015 through March 2016.
During the fill valve assessment, the following was determined:
It was assumed that the more rapid the opening or closing, the more likely a main break would be
associated (if a correlation existed),
There were many openings and closings that were completed in <1 minute for both West Side
and Cedar Street Reservoir during the time period evaluated, and
A resulting valve opening/closing time duration of 45 seconds was chosen to be used to
determine a correlation with main breaks because there were sufficient numbers of occurrences
meeting this criterion yet it reduced the dataset to a low enough number in order to more easily
assess for possible pairing with main breaks.
After assessing the above information, the following results were produced (note this analysis was
conducted only on data collected during the June through August 2015 period):
arcadis.com 69
ASSESSMENT OF CURRENT PRACTICES AND GAP ANALYSIS TECHNICAL MEMORANDUM
Figure 4-22: Distribution System Pressure Plotted with Main Breaks (June – Aug. 2015; Flint WTP in operation over entire period)
arcadis.com 70
ASSESSMENT OF CURRENT PRACTICES AND GAP ANALYSIS TECHNICAL MEMORANDUM
Figure 4-23: Cedar Street Reservoir Operations Plotted with Main Breaks (June – Aug. 2015; Flint WTP in operation over entire period)
arcadis.com 71
ASSESSMENT OF CURRENT PRACTICES AND GAP ANALYSIS TECHNICAL MEMORANDUM
The Cedar Street Reservoir header pressure data shows dramatic, 25-psi pressure drops during every fill
cycle. Communications with City staff indicate that these pressure drops are not actually occurring, and
that the pressure readings are an anomaly associated with the location of the pressure sensor. Obviously,
this type of inaccuracy in recorded pressure measurements prevents analysis of the data and needs to be
fixed.
Findings from the above analysis of pressure and valve operations are inconclusive. The following is a
summary of preliminary results of the data analysis:
Distribution system pressure monitored at the WTP was within acceptable ranges (except for the
90 psi spike and 10 psi dip);
Although the Distribution System Entry Point was not expected to be an adequate location for
assessing system pressure changes against main breaks (it was the only system dataset
available), it was still assessed and there were no pressure fluctuations that could be correlated
with breaks during the time the WTP was in operation;
Main break correlations could not be made with fill events from either individual or combined
reservoirs filling due to limitations in the West Side reservoir data set and anomalies in the Cedar
Street data set; and,
Main break correlations with fill events that included either a rapid valve opening or closing from
either reservoir were not observed, possibly due to issues with lack of time-stamp for main break
events.
Consistency of valve operations practices was also evaluated. Figure 4-24 illustrates four selected
operating records for the fill valve to West Side Reservoir, indicating four different operations strategies.
Each data curve shows the time that elapsed during the opening, filling, and closing periods. These four
curves were chosen to be illustrated here because they show the following differences in operation:
arcadis.com 72
ASSESSMENT OF CURRENT PRACTICES AND GAP ANALYSIS TECHNICAL MEMORANDUM
25
15
10
0
0 100 200 300 400 500 600
Figure 4-24: Selected Operation Summaries for West Side Reservoir Fill Valve
Majority or repairs are conducted under pressure because it is often difficult to find or operate
valves in order to depressurize the main
After the repair is conducted, the main is flushed until chlorine levels are consistent with
background levels
A microbial sample is collected once chlorine levels are satisfactory.
The main is released to service prior to receiving the microbial result. Microbial results are not
documented, but anecdotally, there is rarely, if ever, a positive sample.
City staff indicate that the vast majority breaks would be classified as Type I or Type II, according to the
categorization developed in Water Research Foundation project #4307 – Effective Microbial Control
Strategies for Main Breaks and Depressurization (Kirmeyer et al., 2014) and summarized in the Table 4-2
below. According to this table, microbiological sampling is not required for Type I and Type II breaks.
arcadis.com 73
ASSESSMENT OF CURRENT PRACTICES AND GAP ANALYSIS TECHNICAL MEMORANDUM
4.2.4 Summary
The City water system has numerous main breaks per year, significantly above the number of breaks
recommended by Partnership optimization criteria, mostly occurring in the winter. City infrastructure is
clearly aging and in need of rehabilitation/replacement. City maintenance crews are understaffed, and are
not equipped with planning tools to maintain/replace the water system so that main breaks can be
reduced and adequately tracked for detailed analysis. It is suspected that valve and pump operations are
a contributing factor to main breaks, but clearly temperature is the primary controlling factor.
There is sufficient (possibly excessive) redundancy in the system, so main breaks can be isolated
despite valve problems, while still providing service to customers.
arcadis.com 74
ASSESSMENT OF CURRENT PRACTICES AND GAP ANALYSIS TECHNICAL MEMORANDUM
City staff has been able to make repairs on water mains and service lines, disinfect them in
accordance with best practices, and restore them to service. Recent work to reduce valve
inoperability will help City staff address future main breaks allowing them to quickly isolate the
broken main and minimize the extent to which service is disrupted.
The procedure for documenting main breaks is antiquated and should be updated.
Additional information regarding the approximate time of the break, Type categorization, visual
description (shear break, longitudinal, hole) should be captured.
Main breaks should be evaluated further based upon proper hydraulic partitioning of the system
(i.e., areas influenced by the WTP, by West Side Reservoir, by Cedar Street Reservoir, etc.). This
information should be done with aid from the City’s hydraulic model in order to properly
characterize the number of breaks per mile of pipe per area.
There is a general lack of data, both in the form of various parameters and locations in the
system. Better monitoring of pump operations, valve operations, pressure, and other parameters
is needed. Data that is collected is not communicated to WTP staff.
Increased pressure monitoring through use of portable data loggers is needed through the water
system such that pressure spikes or drops that may cause main breaks can be identified and
causes of pressure fluctuations can be mitigated.
Standard operating procedures for pump start/stops and valve operations are required such that
all staff know how to properly operate this equipment in a consistent, safe manner.
The elevated number of main breaks that occur tend to significantly stress City resources,
causing several staff to be too frequently called away from regular duties.
Staffing is too low to adequately/efficiently respond to and repair leaks that have been identified in
the system (Echologics, 2016).
Annual leak testing is not conducted.
Surge analysis is needed to optimize operations and reduce main break frequency.
There is a lack of system operations planning and staff consistently work in reactive mode.
There is a lack of City vehicles to efficiently respond to and/or complete maintenance.
There is no global positioning system (GPS) coordinates of pipes and valves, and no properly
updated water system map.
arcadis.com 75
ASSESSMENT OF CURRENT PRACTICES AND GAP ANALYSIS TECHNICAL MEMORANDUM
4.3.1 Background
Pressure is a common design metric used in locating, selecting and sizing pipes, tanks, pumps and
control valves to provide desired pressures under normal, and emergency (fire flow) conditions. Under
normal operation of the City’s system, there should be sufficient measurement of pressure such that it
can be properly managed. There should be an understanding of the system’s average operating
pressure, acceptable maximum pressure, acceptable minimum pressure during normal and fire flow
conditions, and the ability to see that pressure remains above 0 psi during main breaks and power
outages. Another direct impact to pressure, which requires SOPs and practices, includes the City’s
practices in pipeline rehabilitation and replacement as well as maintenance and operation of water
storage reservoirs. Proper pressure management can easily offset the cost of increased pressure
management, especially with the result of reduced main breaks, system leakage, and energy
consumption.
Key issues and/or parameters which may directly influence successful pressure management include:
Placement of pressure monitoring instrumentation,
SCADA tracking and alarms for pressure monitoring,
Pipe rehabilitation / replacement,
Main breaks and emergency shut downs,
Reservoir maintenance / cleaning,
Energy management,
Pump station soft start / stop and VFD operation,
Pump station valve operation SOPs (to prevent rapid open / close),
Tracking of customer complaints,
Water loss control, and
Flushing.
Performance indicators for managing pressure should follow the recommendations set forth in the Ten
States Standard and Partnership criteria, as follows:
Ten States Standards Goals:
Normal Demand Periods
o 35 psi average minimum
o Calculated as the monthly average of daily minimum values
Peak Water Demand & Fire Flow Periods
o 20 psi minimum
o Hourly minimum during simultaneous peak demand and fire flow
Emergency Condition Pressure
o During main breaks, power outages, for example
o 0 psi minimum
o Have a completed response plan for preventing possible spread, frequency, duration and
overall extent of contamination due to low pressure events.
arcadis.com 76
ASSESSMENT OF CURRENT PRACTICES AND GAP ANALYSIS TECHNICAL MEMORANDUM
Maximum Pressure
o Not to exceed value set by the City in 95% of measurements
Maximum Pressure Fluctuation
o Not to exceed range set by the City in each pressure zone in 95% of all measurements
Distribution system pressures monitored at the WTP were not sensitive enough to capture any
significant deviations in system pressure due to potential pressure swings that may have either
caused or resulted in a main break (i.e., no pressure swings were correlated with a main break
event);
For reservoir fill valve operations, main break correlations could not be made with fill events from
either individual or both reservoirs filling or if a fill event was associated with either a rapid valve
opening or closing from either reservoir.
arcadis.com 77
ASSESSMENT OF CURRENT PRACTICES AND GAP ANALYSIS TECHNICAL MEMORANDUM
4.3.4 Summary
Available pressure data are not representative of distribution system locations that are to be monitoring
under the Partnership program. Thus, pressure measurement and monitoring within the distribution
system is something that the City needs to undertake to demonstrate optimized conditions. The City
should have pressure monitors installed in enough locations around the system in order to properly
monitor and calculate average pressure. At this time, the only pressure data available is at one point
entering the system and at each reservoir. At the time of writing this report it is understood that there have
been additional portable pressure data loggers employed into the system, but information from these
stations has not yet been made available for review to determine monitoring adequacy. To demonstrate
compliance with Partnership goals, data loggers should be located at the high and low points of the
distribution system or within each pressure zone. The locations of new pressure monitoring sites should
be confirmed as appropriate by the City’s hydraulic model. Monitoring should be accompanied by the
proper SCADA controls and calculations to alert staff when significant deviations occur. Implementing a
structured pressure management program is expected to allow the City to reduce the number of main
breaks each year and reduce other related operating costs such as manpower efforts responding to
breaks.
The City is currently developing a SOP for pressure management at the Cedar Street and West Side
reservoirs. Additional information is provided here to help guide the City in establishing proper pressure
management protocols.
Newly added pressure monitoring stations need to be confirmed by the City’s hydraulic model
and incorporated into the City’s SCADA.
The City needs to set a maximum pressure and pressure fluctuation goal in accordance with the
Partnership approach.
The City needs to perform pressure monitoring calculations based on Partnership criteria in
SCADA and operate the system to achieve pressure goals.
Conduct surge modelling to optimized operations of the reservoir fill valves, pump sequencing,
filter backwash cycles, and elevated tank operations. Determine the need for VFDs, soft starts,
and other surge control devices.
Continue development of SOPs for valve open and close operations, pump start and stop
sequences, etc.
arcadis.com 78
ASSESSMENT OF CURRENT PRACTICES AND GAP ANALYSIS TECHNICAL MEMORANDUM
4.4.1 Background
The Lead and Copper Rule was published by USEPA in 1991 to control lead and copper in drinking
water. Unlike some other drinking water regulations, the LCR does not establish a MCL for either lead or
copper; instead, since lead and copper primarily enter drinking water through the corrosion of plumbing
materials, the LCR establishes a treatment technique to control lead and copper corrosion. As part of this
treatment technique, drinking water systems are required to monitor for lead and copper at customer
taps. If lead exceeds an action level of 15 µg/L at more than 10 percent of customer taps, or if copper
exceeds an action level of 1.3 mg/L at more than 10 percent of customer taps, the utility must take
additional actions including corrosion control.
Since the City began using the Flint River as its sole drinking water source in April 2014 there have been
many instances of high lead levels detected in customers’ residences, which have persisted even after
the City returned to purchasing finished water from DWSD/GLWA in late 2015. To better understand the
current state of internal corrosion control in the Flint water system, available lead data reported by
USEPA and MDEQ have been analyzed. Lead sampling data and other water quality data were collected
by USEPA and MDEQ under different sampling programs. These data provide an indication of the state
of internal corrosion control in the Flint water system over time. In addition to these lead data, water
quality parameter (WQP) data collected at several WQP data collection sites were analyzed to evaluate
how well WQPs relevant to internal corrosion control are currently being controlled.
4.4.2.1 MDEQ
Michigan Department of Environmental Quality conducted two sampling programs to specifically collect
lead data to assess the extent of high lead levels in the Flint distribution system: the Residential and
Sentinel programs. In addition to these sampling programs, the City continued to monitor and record lead
and copper data under the City’s existing LCR compliance program. Thus, there were three datasets from
MDEQ that were analyzed: the Residential, Sentinel and LCR Compliance datasets.
arcadis.com 79
ASSESSMENT OF CURRENT PRACTICES AND GAP ANALYSIS TECHNICAL MEMORANDUM
4.4.2.2 USEPA
USEPA has been extensively involved in data collection in Flint throughout 2016. Data has primarily
been collected under two programs. The first program, referred to here as the Sequential lead sampling
arcadis.com 80
ASSESSMENT OF CURRENT PRACTICES AND GAP ANALYSIS TECHNICAL MEMORANDUM
program, focused on collecting in depth lead profiles at a smaller subset of houses than MDEQ’s Sentinel
program. The second program, conducted to better help understand data from the Sequential lead
sampling program, involves the collection of premise plumbing and service line data to allow for accurate
calculation of premise plumbing and service line volumes. Both data sets are explained more fully below.
4.4.2.2.1 Sequential
The Sequential dataset is comprised of data collected by USEPA at a total of 115 residences across five
sampling rounds between January 28, 2016 and November 15, 2016. Although many of the sampling
sites were included in multiple rounds, many of the sites were not included in every round; instead, over
the five rounds of sampling a total of 279 water samples were collected from 115 sampling sites. Table
4-3 shows summary of all of the sampling rounds conducted under the USEPA Sequential Sampling
Program.
Table 4-3: Summary of USEPA Sequential Sampling Rounds
Unlike the MDEQ sampling programs, the USEPA Sequential sampling was not limited to collecting the
first 1-liter sample through the tap after a stagnation period. Instead, USEPA collected multiple sample
volumes, sequentially, to capture the volume of drinking water that stagnated from the tap all the way
through the premise plumbing and service line back to the distribution main. The specific sampling
protocol is described as follows: Sampling for Rounds 1 through 5, except for Round 2 was conducted by
collecting two 125-mililiter (mL) water samples from the tap following a defined stagnation period (i.e. a
period when no water was used to allow water to rest in contact with the pipes), followed by subsequent
1000-mL water samples. As many samples were collected as needed to capture the volume of water
estimated to be contained within the premise plumbing and service line during the stagnation period. Any
existing filter (either from the faucet or from a whole-house filter system) was removed prior to flushing or
sampling.
Sampling Round 2 was conducted differently than the other rounds. Instead of moving directly to 1-liter
bottles following collection of the first two 125-mL water samples, USEPA used 500-mL water samples to
collect the equivalent volume estimated to be contained within the premise plumbing. After that volume of
water was captured, USEPA returned to using 1-liter sample bottles to collect the remaining volume of
water estimated to be within the service line. Also, Round 2, instead of removing the existing filters
during sampling those filters were bypassed from the faucet. This was done to assess whether lead
sources in specific locations were contributing to the peak lead values observed.
The sampling methods employed in each of the sampling rounds had an impact on the lead values
observed. Because the 125-mL bottles were physically smaller and quick to fill, the water velocity
through the faucet tended to be lower while filling the first two sample bottles. After switching to the larger
arcadis.com 81
ASSESSMENT OF CURRENT PRACTICES AND GAP ANALYSIS TECHNICAL MEMORANDUM
bottles for the subsequent sample volumes, the samplers were able to run the faucets at higher velocities.
This increase in flow rate after the second draw is thought to have mobilized particulate lead that may not
have been removed during the lower flow rates used to fill the first two 125-mL sample bottles.
Therefore, a number of lead peaks observed in the third and fourth sample bottles may be attributable to
particulate lead mobilized by the change in velocity during sampling.
Along those same lines, the faucet flow rates used for sample collection during Round 2 tended to be
lower, because smaller sample bottles were employed during that sample round (500-mL vs. 1,000-mL).
This, in turn, may have reduced the amount of particulate lead mobilized during sampling. In general,
peak lead concentrations observed during Round 2 tend to be lower than those observed in the other
Sequential sampling rounds. It is thought that this can be attributed to the difference in sampling
protocols.
The total number of samples to be collected for any round was based on estimated premise plumbing and
service line volumes, which was based on the lengths and diameters of plumbing materials in the system.
Residences were sampled in one or more rounds of sampling and anywhere between 12 and 40 liters of
water were collected at each home, depending on the configuration of each home’s premise plumbing.
Each of the sample bottles collected during the profile sampling was analyzed for copper, lead, zinc,
magnesium, calcium, sodium, chromium, potassium, manganese, aluminum, nickel, cadmium, and iron.
Another sample, in addition to the profile samples, was collected at each home and analyzed for
additional parameters, such as total phosphorus, sulfate, chloride, fluoride, and total alkalinity, in addition
to the metals listed above. Additionally, the dataset contained information on the sampling sub-location
(e.g. kitchen sink, bathtub, etc.), sample date and time and whether the filter was on or off while sampling.
Information regarding service line materials was not provided directly embedded within the dataset, but
was provided as a separate stand-alone file for each of the 115 locations included in the Sequential
sampling dataset.
4.4.2.2.2 Plumbing
The Plumbing dataset collected by the USEPA provided a detailed summary of the premise plumbing and
service line configuration for 77 of the 115 locations sampled during the Sequential dataset. This dataset
inventories the interior plumbing for each of the 77 residences included in the dataset, with descriptions
for each pipes, fittings, and fixture located between the sample tap and the water main. Data for location,
material, internal diameter, and wetted length are included in the inventory. Based on these data, premise
plumbing and service line volumes were calculated for each of the included residences.
arcadis.com 82
ASSESSMENT OF CURRENT PRACTICES AND GAP ANALYSIS TECHNICAL MEMORANDUM
included in the common database. The MDEQ Residential data collection program, on the other hand,
was open to all water consumers within Flint, and was not restricted solely to high-risk sites. Therefore,
for this analysis, any location within the Residential dataset that was not also included within either the
Sentinel or Sequential datasets was not included in the common database. A total of 4,361 samples
across 801 locations in the Residential dataset were found to be common with the Sentinel and
Sequential datasets. The resulting common database included a total of 8,596 sampling data across 904
locations. The LCR Compliance dataset was not included in the common database because that dataset
did not include complete addresses for the sample locations.
Temperature
Conductivity (micro Siemens (mS))
pH
Total alkalinity (mg/L as CaCO3)
Calcium (mg/L as Ca2+)
Hardness (mg/L as CaCO3)
Turbidity (NTU)
Iron (mg/L)
Chloride (mg/L)
Orthophosphate (reported as mg/L as PO4 but actually analyzed as total P)
Through 2016, data were collected from 8 WQP monitoring sites distributed throughout the distribution
system, as well as at the Cedar Street Reservoir and Pump Station, the West Side Reservoir and Pump
Station, and at the entry point where Flint received DWSD/GLWA water.
arcadis.com 83
ASSESSMENT OF CURRENT PRACTICES AND GAP ANALYSIS TECHNICAL MEMORANDUM
observed during the profile, which was observed in the third sample collected. This sample, which
represented the volume between 250 mL and 1,250 mL in the cumulative volume sampled, had a lead
concentration of 110 µg/L.
This represents a challenge when comparing between databases. Because the Residential and Sentinel
databases only collected one lead value per location, it is not clear if that lead value represents the peak
lead concentration at that location during the sampling event (i.e. if they had continued collecting samples
for a lead profile, it’s possible the peak lead would be higher). However, if the peak value from the
USEPA Sequential database is used for each profile, it may represent a sample volume not captured by
the other datasets.
1,500
1,250
60
40
20
0
1 2 3 4 5 6 7 8 9 10 11 12 13 14
Sample No.
Figure 4-25: Example profile showing lead concentration and sample volumes from USEPA Sequential
dataset
This is best observed by directly comparing the data. Figures B-1 and B-2 in Appendix B present
percentile plots for each of the datasets included in the common database, showing the maximum lead
value observed at each location in the dataset. In other words for the Sentinel curve, each house in the
Sentinel dataset is included but if there are multiple data in the Sentinel dataset for that location, only the
highest lead level at that specific location is plotted. Figure B-1 uses the lead concentrations from the first
sample in each profile, while Figure B-2 was generated using the peak lead concentration from profile in
arcadis.com 84
ASSESSMENT OF CURRENT PRACTICES AND GAP ANALYSIS TECHNICAL MEMORANDUM
the USEPA Sequential dataset. Figure B-1 shows that the USEPA Sequential dataset more closely
matches the other datasets when the first sample lead concentration is used. Therefore, for subsequent
analyses presented here, the lead values used for the USEPA Sequential dataset are based on the lead
concentrations in the first sample of the profile, unless specifically stated otherwise.
While this assumption makes for more accurate comparison between datasets, it should be noted that
Figure B-2 implies the maximum lead values from the first liter samples may not represent the maximum
lead values at each location. The LCR only requires collection of the first liter, so the Residential and
Sentinel sampling programs were in accordance with the existing sampling protocol. However, if the
MDEQ Residential and Sentinel datasets included the collection of profile samples at each location,
similar to the USEPA Sequential dataset, the Residential and Sentinel curves in Figure B-2 may more
closely match the USEPA-Peak curve.
Excluding the unpaired data impacted the lead distribution quite significantly for the winter and spring
data, as more than 80 percent of the data were removed. The removed data tended to be from sites with
lower lead concentrations, either because there was less lead contained in their system plumbing or
because conditions at those locations were less conducive to lead leaching. Removing those sites
effectively increased 90th percentile lead concentrations for Winter and Spring 2016. The lead distribution
percentile plot for paired locations showed higher 90th percentile values as compared to the unpaired
locations plot. The summer and fall data percentile plots did not show considerable change as only 35
percent of the data were excluded from these datasets. Figures B-5 and B-6 in Appendix B illustrate the
arcadis.com 85
ASSESSMENT OF CURRENT PRACTICES AND GAP ANALYSIS TECHNICAL MEMORANDUM
max lead distribution for each location by season for paired data. As these figures show, it appears that
lead levels are improving over time.
arcadis.com 86
ASSESSMENT OF CURRENT PRACTICES AND GAP ANALYSIS TECHNICAL MEMORANDUM
arcadis.com 87
ASSESSMENT OF CURRENT PRACTICES AND GAP ANALYSIS TECHNICAL MEMORANDUM
column. However, for Round 2, three samples were collected within the first liter including sample 1 (125
mL), sample 2 (125 mL), and sample 3 (500 mL). Therefore, for Round 2, the highest lead value
amongst those first three samples collected for that round would be included in the 1st L column in Figure
B-18.
It can be seen that approximately 40 percent of the profile peaks occur in the 1st liter indicating the source
contributing to peak lead values could be either service lines or internal house plumbing. Additionally, this
analysis is helpful in evaluating the number of profile samples to be collected for any future lead profile
sampling.
However, just because the peak lead value occurred within a certain volume doesn’t mean that peak is
high enough to be significant. Several profiles did not find peak lead values in exceedance of single digit
parts per billion. Therefore, the data were re-evaluated considering only those profiles with high lead
values. Figure B-19 shows the comparison of the peak with respect to the high lead values (15 and 50
µg/L) in addition to the associated profile volumes for the peaks. Figure B-20 gives the comparison of the
peaks and all lead values that are higher than 15 and 50 µg/L. It also provides information on the profile
volume associated with each of the peak and high lead values. Out of the 234 profiles that were
collected, ninety-five profiles measured peak lead values in excess of 15 µg/L; of those 95 profiles, 75.8
percent observed the peak lead value within the first 3 liters of water collected. Likewise, forty-two of the
profiles collected measured peak lead values in excess of 50 µg/L; of those 42 profiles, 83.3 percent
observed the peak lead value within the first 3 liters of water collected.
Figure B-21 compares the peak lead values with the 1st draw (sample) values of all of the 234 profiles.
Unlike the preceding two plots, which presented the number of profiles where the peak exceeded a
certain value, Figure B-21 presents the distribution of the magnitude of those peaks. This plot shows that
although peak lead values tend to occur within the first three liters, when looking at the first draw samples
the magnitude of peaks observed is less than the magnitude of peaks that occurred in any sample. The
90th percentile for the first draw sample is 17.4μg/L as compared to the 90th percentile for the peak lead at
79.8 µg/L.
Comparison of peak lead values amongst different sampling rounds (time periods) was also evaluated for
all of the 115 locations. Figure B-22 compares the peak lead values for all five sampling rounds. Round 2
showed the least lead values as compared to all the other rounds. As discussed earlier, the sampling
procedure for Round 2 was significantly different than the other rounds in that the samples were collected
at a lower flow rate starting from the 3rd draw, thus reducing the lead levels. As per the latest sampling
round (Round 5) results, more than 20 percent of the houses have peak lead values greater than 15 µg/L.
Although this is not yet where it needs to be, it does represent a substantial improvement over the first
and third Sequential sampling rounds which approached 50 percent of locations with peak lead values
exceeding 15 µg/L.
Peak lead values for all profiles from the USEPA dataset were compared with the peak lead values from
two other datasets from separate studies. The two datasets were: the national dataset from a study
conducted by EE&T including 34 locations and Del Toral et al. (2013) dataset from an Environmental
Science and Technology article containing 25 locations. Figure B-23 shows the results of this
comparison. It can be seen from the figure that the peak lead values from the Flint USEPA Sequential
dataset are significantly higher than the EE&T national dataset with approximately 50 percent of profiles
greater than 15 µg/L compared to approximately 23 percent of profiles in the EE&T national database.
The percent of profiles with peak lead values higher than 15 µg/L for Flint USEPA Sequential was
substantially lower than the Del Toral et al. (2013) dataset.
arcadis.com 88
ASSESSMENT OF CURRENT PRACTICES AND GAP ANALYSIS TECHNICAL MEMORANDUM
arcadis.com 89
ASSESSMENT OF CURRENT PRACTICES AND GAP ANALYSIS TECHNICAL MEMORANDUM
Prior to July 2016, the orthophosphate feed system used to boost the orthophosphate dose at the Flint
WTP was not flow-paced, which could have increased variance in the orthophosphate feed. To evaluate
the potential impact of flow-pacing the orthophosphate feed, control charts were generated for the data
collected post-July 1, 2016 and compared to the previous control charts generated using all of the 2016
data to see if there were any differences between the differentials between the upper and lower control
limits. If the system was more well-controlled after flow pacing, the differential between the upper and
lower control limits would be expected to decrease. Figure 4-26 shows the findings from this analysis. In
general, flow pacing did not significantly reduce the variance in orthophosphate levels in the distribution
system; however, it may have mitigated the fact that the orthophosphate level in the source water
supplied by DWSD/GLWA was considerably more variable in the second half of 2016.
1.4
Differential between UCL and LCL
1.2
1
(mg/L as PO4)
0.8
0.6
0.4
0.2
0
Figure 4-26: Comparison of differential between upper and lower control limits for orthophosphate control
charts generated using all of 2016 data vs. post-July 1, 2016 data
4.4.4 Summary
The data analyses conducted in this report focused on evaluating the lead results from the different
sampling programs conducted by MDEQ and USEPA. Lead results were evaluated based on seasons,
flushing event, spatial distribution and service line material. For each of these categories, evaluations
were done using unpaired and paired data. The results from these analyzes showed that the unpaired or
raw data can mask/conceal the extent of lead levels at higher-risk sites. Removing unpaired data affected
the percentile distributions, increased the median and 90th percentile lead values. For example, refer to
Figures B-7 and B-8, which show percentile plots of unpaired and paired datasets. The inclusion of
unpaired data makes the distribution of lead prior to and after the May flushing event appear similar, but
removal of the unpaired data clearly shows that most locations saw significant reductions in lead following
the flushing event. While there are many factors that may have contributed to improvement in lead levels
beyond the flushing event, there is a clear signal indicating that lead levels at most locations decreased in
the second half of 2016.
Analysis of the paired data indicates that lead levels appear to be decreasing over time. This is indicated
by Figures B-3, B-5, B-8, B-9, and B-13 through B-17. Figure B-3 shows that the 90th percentile of data
arcadis.com 90
ASSESSMENT OF CURRENT PRACTICES AND GAP ANALYSIS TECHNICAL MEMORANDUM
collected in Fall 2016 is near the lead action level of 15 µg/L, indicating a substantial improvement over
where the system was in early 2016.
Although the most recent data show significant improvements, they are still at a level that would be
concerning in any water system. The goal of compliance with the LCR is not to be at the action level
each quarter; instead, best practice would be to maintain the system with 90th percentile values
substantially below the action level, so that the system can withstand upsets or anomalous spikes without
violation of the LCR. While the specific homes for LCR compliance are showing 90th percentile values
below 15 µg/L, the preponderance of analysis in this report show that overall the 90th percentile is now
right near 15 µg/L but not substantially below. Further, the shape of the lead percentile plots continues to
be concerning. Reviewing the percentile data shows that the curves are relatively “flat” at the high end,
indicating high spikes at locations with the highest lead levels. Again, the high levels are much better, but
still present. It would be better if the curve maintained a similar slope throughout, so that difference
between the 95th percentile and 90th percentile was not substantially different. Improvement is still needed
to bring down those peak lead values at the most impacted locations to better match these goals.
The USEPA Sequential data do suggest that there may be a remedial action beyond the steps that have
already been implemented that could improve lead in many locations. The majority of lead profiles
indicate that the lead peaks are occurring within the premise plumbing. The fact that the peaks are not
coming from the water volumes associated with the service lines suggested that dissolved lead leaching
from lead service lines is not directly contributing to the lead levels observed at most residences. The
observed peaks may instead be related to particulate lead that has deposited in the “nooks and crannies”
associated with premise plumbing.
If there is particulate lead that has deposited within premise plumbing, research performed for the Water
Research Foundation has indicated that improvement in lead levels can be achieved by performing a
whole-house flush to remove particulate lead. Using established protocols, this flushing can increase flow
rates through premise plumbing, and may be able to mobilize particulate lead for removal from the
system.
Relative to treatment goals, it appears that the water quality parameters most relevant to corrosion control
(total alkalinity, pH, orthophosphate dose) are relatively stable within acceptable ranges, particularly for
the second half of 2016. For example, although there were some swings in pH throughout this period, the
Flint system was generally maintained within 0.2 s.u. variance from location to location, and was
generally kept within the pH range in which orthophosphate treatment is effective.
The one exception would be orthophosphate dose. Site D9/D13, which is located near Forest Park in the
north part of the City, had consistently lower orthophosphate concentrations than the rest of the City.
This may be a water age issue, as low chlorine concentrations have been reported in that same area as
well. Throughout 2016, there were at least two periods when the measured orthophosphate
concentration at site D9/D13 was lower than the current orthophosphate target of 3.1 mg/L as PO4.
Outside of site D9/D13, the City has generally met the target orthophosphate minimum of 3.1 mg/L as
PO4 throughout the system. One method of comparing to other water systems would be to review control
charts for the orthophosphate, to see the spread between the upper and lower control limits. These
control limits represent the spread of data over the analysis time period, so if they are closer together they
indicate the water quality is stable, while systems that are not stable would have the control limits spread
further apart. It appears that most of the sites saw a spread of 0.3 to 0.5 mg/L as PO4. This is consistent
with the variance in the water received from DWSD/GLWA, as well as published orthophosphate control
charts from other utilities.
arcadis.com 91
ASSESSMENT OF CURRENT PRACTICES AND GAP ANALYSIS TECHNICAL MEMORANDUM
Moving forward, it would be prudent to evaluate the target minimum orthophosphate dose. The current
target was based on lead-phosphate solubility curves published by USEPA that show that lead-phosphate
solubility is lowest with orthophosphate doses near 1 mg/L as P (equivalent to 3.06 mg/L as PO4). This
dose is relatively high, although not unheard of, for United States (U.S.) drinking water systems (it would
not be considered high in other localities, such as in the United Kingdom). There has been concern that
the high orthophosphate dose could cause “white precipitate” associated with calcium phosphate
precipitation in the distribution system. Although there is no evidence that the current orthophosphate
dose is adversely impacting the system, there may be a desire to reduce orthophosphate doses over time
or modify the WQP limits. Laboratory testing is recommended to evaluate the acceptable range for
orthophosphate without negatively impacting internal corrosion control.
The current lead and WQP monitoring strategy appears to be robust relative to many other water
systems. Due to the extensive lead issues experienced by Flint, the amount of lead sampling conducted
greatly exceeds the samples per monitoring period that would be required by the LCR. The spatial
distribution of WQP monitoring also appears to be sufficient for monitoring WQPs relevant to corrosion
control; however, additional or different monitoring sites may be recommended for other compliance and
water quality surveillance issues as discussed in Section 0 of this report. If additional monitoring
locations are added, it would be prudent to monitor WQPs relevant to corrosion control at those locations
as well.
Frequency of WQP monitoring could be improved. If a WQP is out of compliance with the City’s corrosion
control treatment (i.e., if the orthophosphate dose is below the 3.1 mg/L as PO4), then the system is
considered out of compliance until a sample above the compliance limit is measured. Previously, the City
analyzed WQP data on a weekly basis. Therefore, if a WQP was out of compliance, the system would
remain out of compliance for at least 7 days until the next WQP sample was collected. Since the system
cannot exceed 9 days out of compliance during the monitoring period, more than one out of compliance
sample would cause a violation. Since November 1, 2016 the City has been sampling twice a week,
which will reduce the time the system might potentially be out of compliance. The recommended practice
is to analyze samples on the day they are collected so that the system could be re-sampled the same day
if the result is outside the compliance range to minimize the amount of time the system is considered out
of compliance.
The current lead and WQP monitoring strategy appears to be robust relative to many other water
systems. Due to the extensive lead issues experienced by Flint, the amount of lead sampling conducted
greatly exceeds the samples per monitoring period that would be required by the LCR. In addition to the
10 original WQP monitoring sites, Flint has added an additional 15 sites as part of the RTCR monitoring
program, which could potentially be used for monitoring as well. Figure 4-27 shows the location of the 10
current WQP monitoring sites along with 15 potential WQP monitoring sites (which are currently being
sampled for chlorine and total coliforms). Note that the site labeling has been updated following the
expansion of the RTCR program, and does not correspond with the previous labeling program used prior
to 2017.
arcadis.com 92
ASSESSMENT OF CURRENT PRACTICES AND GAP ANALYSIS TECHNICAL MEMORANDUM
The spatial distribution of WQP monitoring appears to be generally sufficient for monitoring WQPs
relevant to corrosion control. However, it is recommended that D18 be added to the list of WQP
monitoring sites at which WQPs are monitored in addition to chlorine and total coliforms to better
represent the southern portion of the distribution system. Additionally, it would be beneficial to capture
WQP conditions immediately west of S Dort Highway. If possible, as the RTCR program is expanded
from 20 to 25 sites, it would be beneficial to identify a location west of S Dort Highway along E Court St to
be included in both RTCR and WQP monitoring plans. If that is not possible, it is recommended to add
site D15 to the list of WQP monitoring sites.
Frequency of WQP monitoring is sufficient for monitoring of corrosion control parameters. The City has
indicated that they are sampling the 10 original TCR/WQP monitoring sites at least once a week, with
some sites being sampled twice per week (due to the rotating nature of TCR sampling, some sites are
sampled twice per week every other week). It is recommended that site D18 and a future RTCR site west
of S. Dort Highway (or site D15, if this new site cannot be found) be added to the list of WQP monitoring
sites that are sampled at least once per week for the WQPs relevant to corrosion control. This would
result in a sample pool of 12 weekly locations. To reduce the sampling burden, the City could consider
reducing the analysis of WQPs that are not relevant to the City’s corrosion control program to a monthly
basis. Specifically, it is not necessary to monitor for Calcium, Hardness, or Conductivity more than once
per month from the WQP monitoring sites. However, because calcium and conductivity are specifically
arcadis.com 93
ASSESSMENT OF CURRENT PRACTICES AND GAP ANALYSIS TECHNICAL MEMORANDUM
listed in the Administrative Order, the City should obtain permission from the State prior to dropping these
parameters from their weekly WQP monitoring program.
One item that the City should consider is that weekly sampling, by itself, could leave the City vulnerable to
violations of the pH and orthophosphate requirements specified in the Administrative Order. For
example, the City is required to maintain a minimum orthophosphate residual of 3.1 mg/L as PO4
throughout the distribution system. If the measured orthophosphate level is less than 3.1 mg/L, the
system is considered out of compliance, and the City will be considered in violation of the Consent Order
if the system is out of compliance for more than 9 days during a 6-month monitoring period. If one of the
weekly samples was found to be out of compliance and the system were not sampled again until the
following week, the system would be out of compliance for 7 of the 9 days allowed during the period.
Another low sample would drive the system into a violation.
To avoid such a scenario, the City has indicated that all WQP samples are analysed immediately
following sampling and, if a non-compliant sample is found, the City will immediately resample. The City
will continue sampling and, if necessary, make system adjustments until the WQPs meet their required
values.
Lead monitoring programs currently in place exceed those of most systems in the U.S.
Variance in total alkalinity and pH through the distribution system is in line with other systems in
the U.S. Both are currently being maintained within acceptable ranges for the City’s corrosion
control system.
WQP monitoring for an extended list of parameters has been historically conducted twice per
week at 10 locations, which is double the once per week requirement stated in the Administrative
Order. WQP monitoring is currently conducted at least once per week at all 10 locations.
The orthophosphate level at Site D9/D13 (now labelled D19) is consistently lower than the rest of
the Flint system, and has dropped below the 3.1 mg/L as PO4 minimum target at times during
2016.
Lead levels at the higher end of the percentile curve are still excessive. The City does not
currently meet the performance target recommended in the Guidance Manual for Monitoring
Distribution System Water Quality (Kirmeyer et al., 2002) of 100 percent of the samples for lead
below the action level of 15 µg/L. The preponderance of data suggests that the 90th percentile of
lead levels is not currently lower than 15 µg/L. Peak lead levels above the 90th percentile are still
excessively high.
The target minimum orthophosphate level in the distribution system (3.1 mg/L as PO4) is at the
high end of the range of systems in the U.S. Laboratory testing is needed to assess the feasibility
of reducing orthophosphate or widening WQP limits in the Flint system without negatively
impacting corrosion control.
arcadis.com 94
ASSESSMENT OF CURRENT PRACTICES AND GAP ANALYSIS TECHNICAL MEMORANDUM
None identified.
4.5.1.1 Background
The United States SDWA requires public water systems to be responsible for the water quality delivered
to the water customer’s service connection or point of entry. However, they are not responsible for
contaminants and/or pollutants that are added to the drinking water by any circumstances under the
control of the consumer beyond the public water supply point of entry. As such, state and local
governments, and agencies have adopted cross-connection control rules to protect the health, safety, and
welfare of the public water system from potential cross-connections and backflow conditions.
arcadis.com 95
ASSESSMENT OF CURRENT PRACTICES AND GAP ANALYSIS TECHNICAL MEMORANDUM
The existing CCCP appears to align with AWWA M14 – Backflow Prevention and Cross-Connection
Control, Recommended Practices for commercial and industrial customers. This manual is the industry
standard and the City’s program is based on these practices.
None identified.
It appears that the ongoing management and supervision of the cross-connection control program is
negatively impacted by current staffing levels. The City is not able to meet the requirements of cross-
connection inspections and verify the re-certification testing of the current 1,000 backflow protection
devices within the City’s system. According to the City, the number of testable devices has dropped from
1,200 in 2012 to the current number of 1,000 due to the loss of business and industry within the City.
Records are currently maintained within a database developed by the City. While the City has indicated
that the database has served them well, there are commercially available software programs that are
specifically developed for the management and recordkeeping of a cross-connection control program.
The City should investigate the various programs and determine if there are any that meet their needs.
The Michigan SDWA (Part 14 of Act 399) prohibits cross-connections and requires public utilities to have
a formal program to control them. This Act also does not differentiate between residential and other
customer types. Based on conversations with City Staff, they have not implemented cross-connection
control on the residential level. AWWA M14 provides specific recommendations should the City opt to
implement a residential cross-connection control program. Aspects of a residential program include
several key components:
• Performing a cross-connection control survey to identify the degree of hazards, whether existing
backflow prevention exists at the property, hydraulic conditions of an actual or potential cross-
connection, and if thermal expansion provisions may be required after the backflow preventer is
installed. Most residential properties fall under a low hazard category which can be adequately
protected with a double check valve assembly.
• Developing an installation schedule, which typically recommends installing the backflow preventer
during a meter read or when the meter is being replaced.
• Performing public outreach through stakeholder meetings or other communications so that the
residential customers understand the importance of backflow protection and the need for a
thermal expansion tank within the property’s plumbing system once the backflow preventer is
installed.
arcadis.com 96
ASSESSMENT OF CURRENT PRACTICES AND GAP ANALYSIS TECHNICAL MEMORANDUM
• Performing periodic inspections of each property to reassess risk. Water customers that pose a
high hazard or have a high potential for backflow to occur should be reassessed more frequently.
It is unclear if the City has a formal backflow incident response plan that includes procedures for tracking
and investigating a backflow incident during and after the event. This plan should be implemented if an
investigation of a low pressure or water quality complaint indicates that a backflow incident has or may
have occurred. AWWA M14 provides specific recommendations for this plan and the City should
consider developing a response plan if it doesn’t exist.
4.5.2.1 Background
Customer complaint tracking can be used to identify abnormal water quality condition resulting from
maintenance activities, water age issues or even a potential contamination event. A utility’s customer
complaint system should include timely identification, investigation, and response to abnormal water
quality complaints. This includes the necessary processes, procedures, and management systems to
collect, compile, and analyze complaints.
According to the Partnership program, optimized distribution systems should have few complaints. The
best performing systems have less than 2.5 Technical Quality Complaints per 1,000 customer accounts
annually. Technical quality complaints include complaints associated with water quality, taste, odor,
appearance, pressure, main breaks, and disruptions of water service. They are directly related to the core
services of the utility. Assuming the City serves 31,520 customer accounts (Rowe, 2016b), optimization
would equate to fewer than 80 technical water quality complaints annually.
arcadis.com 97
ASSESSMENT OF CURRENT PRACTICES AND GAP ANALYSIS TECHNICAL MEMORANDUM
Detailed analyses are conducted on water quality samples collected from complaint locations.
The analyses provide insight into metals, aesthetic, and microbial conditions.
According to discussions with the City, records are currently maintained that track water quality
complaints. These records have not been reviewed by the Project Team.
Assuming records adequately capture the nature, locations, and response taken to each
complaint, the data should be analyzed to develop baseline conditions that can be tracked and
trended over time. Once a baseline has been established, the baseline should be periodically
reviewed and updated. As summarized in Criteria for Optimized Distribution Systems (Friedman
et al., 2010), information that should be captured includes:
o Type (classification), selected from a sheet with clear descriptors.
o Date
o Location
o Caller
o Responder/response (dispatch date, time, personnel involved)
o Response follow up including description of remedial actions taken
o Call responders should be trained to recognize and respond to situations where there is
potential for serious health risks and/or property damage.
o Cross connections should be investigated as potential causes of customer-
observed/reported water quality irregularities
Complaints should be entered into GIS or a CMMS (once implemented) for better
tracking/analysis
Compare current complaint rates to industry benchmark of no more than 2.5 technical water
quality complaints per 1,000 customers.
SOP for customer complaint response and tracking. Include a Communications Plan between the
WQL and WSC.
4.5.3 Flushing
4.5.3.1 Background
Distribution system flushing is an important practice for maintaining distribution system water quality.
Flushing can be used to clean water mains of sediment, impurities and biofilm, or to increase turnover to
arcadis.com 98
ASSESSMENT OF CURRENT PRACTICES AND GAP ANALYSIS TECHNICAL MEMORANDUM
reduce water age, disinfection byproduct formation, and to increase chlorine residuals. Flushing programs
can be system-wide or can focus on problem areas.
There are two general types of flushing – conventional (CF) and unidirectional (UDF) – and each has a
specific application purpose. The correct flushing technique(s) should be selected and implemented in
response to the desired water quality objectives. As summarized in Figure 4-28, conventional flushing can
achieve water turnover to temporarily reduce water age, but does not generally achieve actual main
cleaning. In fact, imposing high flow rates using a conventional approach can stir up metals, sediments,
and biofilm which can be spread further into the system since valve isolation is not used to control the
direction of flow, and smaller diameter upstream mains will experience higher scouring velocities.
Conversely, strict UDF requires significant planning resources and diligent valve sequencing to maximize
velocities through sequential stretches of pipe. UDF can be quite effective for main cleaning in many
circumstances, but is too labor-intensive if water age reduction is the ultimate goal. Pipe material,
condition, disposal options, and finished water chemistry stability must also be considered when selecting
a flushing approach.
arcadis.com 99
ASSESSMENT OF CURRENT PRACTICES AND GAP ANALYSIS TECHNICAL MEMORANDUM
for implementation in the spring of 2017 as discussed below. Ultimately, SOPs for system flushing will be
developed for the City of Flint in support of a long-term flushing program.
The need for conventional and dead-end flushing to encourage water turnover has increased due to
declining occupancy of residential and commercial areas over time and resulting increases in water age.
In response to the source changeover event, the City worked closely with USEPA Region 5 to implement
low velocity bulk water turnover flushing with the goal of increasing chlorine residual without stirring up
and further damaging destabilized pipe scales. The City’s current flushing program consists of use of:
The City should continue its conventional flushing program once weather allows. This program consists of
use of auto and manual flushers. Now that the distribution system has time to re-stabilize, it is probable
that a higher hydrant flow rate can be utilized to reduce labor time and more rapidly improve water quality
conditions. Step flow rate trials beginning at 20 gpm, and increasing to 50, 100, and 200 gpm should be
conducted, while measuring chlorine residual, turbidity, and iron to assess response. A detailed SOP will
be prepared for implementation in the spring of 2017. Until this trial is complete, the City should continue
flushing at 20 gpm as needed to maintain adequate chlorine levels throughout the distribution system.
The City’s conventional and dead-end flushing practices currently used to manage water age
meet and exceed industry standards.
arcadis.com 100
ASSESSMENT OF CURRENT PRACTICES AND GAP ANALYSIS TECHNICAL MEMORANDUM
o Specific flow rates have been targeted that are known to avoid sediment and biofilm
disruption and spreading of contaminants.
o Numerous programmable auto-flushers are used and can be rotated around the City as
needed to meet water quality objectives.
o A hydrant sampling device has been created and is used to control flow rates and collect
sanitary water quality samples.
As discussed above, it is possible that flowrates during conventional and auto-flushing can now
be increased. A step-wise approach to increasing flow rates and measuring the water quality
response should be conducted to select the optimal flow rate(s).
The City should assess the effectiveness of UDF for reducing pipe wall chlorine demands and
removing accumulated sediment from the distribution system.
o The City should conduct a UDF pilot study on approximately 2 miles of water main.
Training will be provided on asset mapping, optimal loop layout, water disposal, and
monitoring to document water quality impacts.
o If successful, the City should develop a more comprehensive UDF program to remove
destabilized materials from the distribution system in a controlled manner, thereby
reducing discolored water complaints and reducing chlorine demand caused by
accumulated sediments and biofilm on pipe walls. The hydraulic model will be needed to
optimize flushing loops.
o Estimated level of effort for developing and implementing a UDF program will be
prepared as part of the Resource Analysis and Needs Assessment.
4.5.4.1 Background
Water systems often maintain a hydraulic model to predict performance of the system and to solve a wide
variety of design, operational and water quality problems. For example, a hydraulic model can predict
pressures and flow rates through the system for comparison to design standards, or can track flow
through a system to determine water age and water quality areas of concern. The USEPA recommended
use of a calibrated hydraulic model meeting certain minimum requirements to determine long term
compliance monitoring locations for the Stage 2 DBPR in 2006 through a System Specific Study (USEPA
Stage 2 DBPR; Final Rule. Fed. Reg. 71:2:388). Per the Rule, the following minimum criteria must be
met for a hydraulic model:
arcadis.com 101
ASSESSMENT OF CURRENT PRACTICES AND GAP ANALYSIS TECHNICAL MEMORANDUM
Model must be an Extended-Period Simulation (EPS) and must simulate 24-hour variation in
demand and show a consistently repeating diurnal pattern of residence time.
Model must include:
o A minimum of 75 percent of pipe volume and 50 percent of pipe length
o All pressure zones
o All 12-inch diameter and larger pipes
o All diameter pipes which are hydraulically significant
o All storage facilities with typical operations represented
o All active pumping stations and all active controls
Specific modeling standards vary based on the purpose of the model and available data. However,
modeling standards specific to each system are recommended to be maintained and followed (AWWA,
2012).
The model has been setup to perform real-time modeling, although it is not directly connected to the
City’s SCADA system at this time. The real-time modeling method to achieve calibration for a specific
arcadis.com 102
ASSESSMENT OF CURRENT PRACTICES AND GAP ANALYSIS TECHNICAL MEMORANDUM
period of time is above expectations. Real-time modeling is not widely adopted and it is still early in the
industry to be considered successful as a future industry standard.
The model needs to be further developed to include scenarios for conditions other than a specific period
(i.e., October 2016). Data analysis is recommended to determine the annual average day and annual
maximum day demands and operational conditions.
Model calibration should be performed with data that stresses the hydraulic capacity of the system.
Calibration during peak demand conditions such as during fire flow testing is necessary to determine if
the model can represent the system under peak conditions. Currently, the model is calibrated to match
typical daily conditions with no specific testing being performed as part of calibration.
Demand allocation and representation within the model is unintuitive and does not align with industry
standards. Each node should be loaded with average demand from the customers nearest to it, and any
non-revenue water should be equally allocated among the model nodes. A diurnal pattern should be
included to simulate 24-hour repeatable variation in system demands. The current pattern is not a
repeatable 24-hour pattern and acts as a multiplier to simulated demands.
The model facilities should be confirmed and better represented within the distribution system model.
Many pumps are represented with a single head and flow point rather than the complete pump curve.
When a single point is utilized in the model, the software arbitrarily adjusts the flow above or below the
specified value as head conditions change. The pump curves should be included and data reviewed at
pumping stations to determine if existing, aged pumps are operating along the original design curve or if a
modified curve (often a result of impeller wear) is required to match pumping station operational data.
The model is currently being expanded by USEPA to predict free chlorine concentrations throughout the
system. However, initial efforts suggest a jar test bulk decay coefficient did not adequately represent
system conditions. Additional work is needed to ensure that the simulated model results match the
chlorine concentrations measured throughout the distribution system by both the City of Flint staff and the
USEPA.
The City has not developed a model maintenance plan, or regular schedule for updating the model
components and performing periodic calibration.
Additionally, City staff have not been trained on the model. Training should include basic model usage,
model capabilities, modeling goals, location of input/output data, instructions on how to export model
results for viewing, analysis of various fictitious standard model simulations, and demonstration of
potential model updates and calibration.
A surge model is needed to assess and prevent pressure transients caused by operational activities, as
discussed in other sections of this report.
arcadis.com 103
ASSESSMENT OF CURRENT PRACTICES AND GAP ANALYSIS TECHNICAL MEMORANDUM
4.5.5.1 Background
Valves and hydrants are essential components to a distribution system, but are often neglected as they
are typically only needed during emergencies, such as fires or main breaks, or during periodic
maintenance activities, such as flushing. As such, these assets tend to be overlooked or given a lower
priority when considering the many other components that require maintenance, rehabilitation or
replacement. The Partnership program recommends that utilities establish and implement a well-defined
valve exercising, rehabilitation and replacement program and a hydrant inspection, maintenance and
replacement program. These programs should include an asset inventory, condition assessment plan,
work management process, planning process, and communication plan.
Additionally, proper design and installation of these assets is necessary for achieving desired water
pressure and flow control within the distribution system and to allow for shutdown of sections for
emergency repair and planned maintenance. Various guidance is available and includes design
specifications for locations, spacing and sizing of hydrants and valves as well and proper installation
procedures.
Since 2015, the City has made significant investment progress on the valve assessment and exercising
program. A total of 8,228 appurtenances (76%) were field-located and assessed, with 5,438 valves
exercised successfully and confirmed to be in good/fair condition. Of these valves, approximately 1,880
valves were initially found to be inoperable/frozen, but staff was able to rehabilitate into working condition
(thus improving valve operability from 44% to 70%). Out of the total 2,000 valves recommended for
follow-up (repair, replacement or data discrepancy identified), evidence was found confirming City did
complete the replacement of 85 valves and 70 hydrants in 2016. This scenario provides an example of
the commitment of City resources to ensure the flow of information from the office to the field. It also
arcadis.com 104
ASSESSMENT OF CURRENT PRACTICES AND GAP ANALYSIS TECHNICAL MEMORANDUM
demonstrates that current business processes can allow the completed work to be reported up to
departmental management.
In addition, evidence has been found confirming the fire hydrants within the City of Flint have been able to
meet the needs of emergency services. Per the 2013 City of Flint Water System Sanitary Survey, “the
majority of the system is meeting fire flow standards.” The City of Flint’s last Insurance Services Office
Community Grading Schedule evaluation was completed in 2010, when it received a community rating of
5/10. No documented evidence found within the 2016 Fire Department Operational Report indicating any
issues related to ability of fire fighters to locate hydrants or issues related to inoperability. This suggests
that the water distribution system and hydrants are currently meeting the emergency needs for the
residents within the City of Flint.
The City should continue efforts related to obtaining GPS coordinates for all appurtenances (at minimum
sub-meter horizontal accuracy). Per the WFR Study 4369 UIM Article 4/15/14, “Knowing the exact
location of each valve, whether the position of the valve was open or closed and having performed
regular operability maintenance of the valves, makes it more likely that the utility’s crew can quickly
isolate the area of the break, and thereby, minimize the amount of water discharged and damage to the
surrounding properties…”. Out of the estimated total appurtenances, approximately 1,780 valves (24%)
and 1,874 hydrants (52%) have not been field-verified and coordinates are unknown.
The City should also consider enhancing the GIS database schema to include other additional
information. Most importantly, the current valve database does not properly identify the valve type (such
as blow-off compared to gate/air valves) or function (in-line main isolation vs off-line). This is most critical
information to be leveraged from the Wach database. Attributes recommended by AWWA which are
currently missing include:
Valve Attributes:
o Type (gate, butterfly, cone, check, air valve, blow-off, etc.)
o Function (in-line main, hydrant isolation, service line, bypass, division, pressure reducing,
etc.)
o Direction to close (close right, close left)
o Position (open, closed, partially open)
o Rated working pressure
o Torque required to turn the valve
o Frequency
o Criticality
o Serial Numbers (for valves 12-in and larger)
o Special Coatings
Hydrant Attributes:
o Date of installation
o Size
o Type
o Inspection Date
o Condition
o Historical repairs made
arcadis.com 105
ASSESSMENT OF CURRENT PRACTICES AND GAP ANALYSIS TECHNICAL MEMORANDUM
o Direction of open
o Number of turns
Overall, a comprehensive data model review is recommended to consolidate information into a central
location, review gaps, and duplications. Multiple data sources were identified (such as paper, excel, GIS)
tracking various valves and hydrant details, however no evidence has been found identifying a plan to
migrate all asset-related information into a single data repository. Specifically, the Hydrant Maintenance
Record Form, Fire Flow Testing Form, and 1908 Darlings Replacement Schedule, should at minimum be
integrated/uploaded on a regular basis (such as monthly). Most importantly, a plan should be established
and executed immediately to leverage the newly developed Wachs valve/hydrant database and reconcile
this information with historical paper location cards, excel spreadsheets, and GIS maps. This plan
should also consider modifying the current asset identification numbering system (which currently uses
Map Grid/Page Numbers) to a numbering system more consistent with industry standards
(comprehensive fixed asset hierarchy with no duplicate numbers) and consistent use throughout all work
activities. In addition, the information currently being tracked during a watermain break/leak does not
include a field related to the “location” of the break (i.e. identify watermain breaks which occurred at valve
or appurtenance).
Historically, the City’s Water Distribution operators have relied on paper maps and location cards as the
primary source of information. With the delivery of the Wachs GIS database, the City now is challenged
with integrating this data source into the daily routine of their field and office staff. Additional information
is needed to confirm that appurtenances are locatable in the field (i.e., field staff can find the right valve in
a reasonable period of time even in areas impacted by weather events). Additional strengthening may be
needed to ensure asset information is accessible to the specific person who is going to locate, access,
and operate the asset (including historical maintenance records and reliability information, currently in
paper and excel). This also includes updating the Water Main Maps, ensuring important information is
visible (such as pipe diameter, appurtenance locations/ID numbers, and critical water users). It is
recommended that Arcadis shadow a water operator for a day to review current workflow and confirm the
appropriate data sources are being leveraged to field staff and appurtenances are locatable in the field.
As stated in the 2016 Capital Improvement Plan, “Making progress on the budget deficit has required
difficult decisions and sacrifices, including dedicating minimal resources to capital improvements beyond
emergency repairs.” While this statement may hold true to most City-wide projects, this approach should
not be applicable related to the O&M of the water distribution system. Specifically, the Valve Exercising
and the Hydrant Replacement Program is currently prioritized in the CIP as level 2 (Important) and should
be re-evaluated to consider the feasibility to moving to a level 1 (Urgent). Both the valves and hydrants
have been identified as a main priority and meets the current CIP definition of Priority 1 (“Urgent, high-
priority projects that are required to comply with a Federal or state requirement, or projects that would
address an emergency or remedy a condition dangerous to public health, welfare, and safety… and other
projects vital to the economic stability of the City”).
It is also recommended that the City consider departmental organization and other options to optimize
resources. Many similar sized agencies have been able to successfully balance reactive/unplanned and
preventive/planned work by dedicating separate crews to each task. The preventive maintenance related
to fire hydrants and valves are often coordinated efforts performed by the same crew. As the owner of its
public water system, the City of Flint has responsibility for compliance with the SDWA under Act 399.
arcadis.com 106
ASSESSMENT OF CURRENT PRACTICES AND GAP ANALYSIS TECHNICAL MEMORANDUM
These responsibilities include ensuring proper design, construction, and operations and maintenance.
Since this has been an ongoing struggle for the City of Flint, organizational change should be considered.
Per AWWA Standard G200 (Section 4.2.5-6), “The utility shall have a documented Valve and Hydrant
Exercising Program … and shall include a goal for the number of transmission valves to be exercised
annually... and… set goals to reduce the percent of inoperable valves and replacement.” Evidence has
been found confirming a documented plan for valve exercising and hydrant replacement does exist. The
2015 Capital Improvement Plan identifies a goal to “operate approximately 1,430 valves throughout the
system on an annual basis” and “replace hydrants throughout the City on 20-25 year rotation”. While this
plan is a good first step, two activities are not accounted for: valve repair/replacement (R&R) and annual
hydrant assessment. A plan to “repair approximately 200 valves and 60 hydrants a year” was proposed
as part of the 2016 Drinking Water Revolving Fund Draft Project Plan, but was not approved. Thus, a
documented commitment towards these efforts is still outstanding.
Evidence was found of an annual hydrant assessment and fire flow testing (Excel), however no
documented frequency/schedule was found, nor documented plan for winter preparation. Related to
valves, A plan is recommended to ensure the City staff is able to pick up where Wachs Services left off
and continue annual valve assessment/exercising for the remining 2,635 appurtenances (24% of City
system) which were not successfully exercised in 2015. When comparing this to the plan identified in the
2015 CIP, proposed changes include: 16-in and larger to be exercised/assessed once per year (not 2-yr
cycle) and the rest of valves on a three to five-year cycle (not 10-yr cycle). In addition, gate valves (once
the valve type is known) should be exercised at minimum every 5 years. With these updated frequencies,
the annual valve exercising budgets estimated in the 2015 should be adjusted to 2,572 valve
maintenance per year (not 1,430). In addition, new language is recommended which states “Documented
repairs and replacements also to be completed during the course of that year” to ensure timely work order
completion. No evidence found confirming a plan is in place to pre-plan the sequencing during normal
operations and hydraulic model coordination efforts. In addition, any areas flagged for future
abandonment (based on cost-benefit ratio or obsolesce) should also be communicated and the
preventive maintenance plan updated. In summary, the proposed preventive maintenance and R&R
annual goals are presented in Table 4-5.
Table 4-5: Proposed Preventative Maintenance and Rehabilitation and Replacement Goals
1
The AWWA Standard G200 (Section 4.2.5) indicates the utility shall have a documented Valve
Exercising Program which follows M44 and the manufacturer’s recommended procedures.
arcadis.com 107
ASSESSMENT OF CURRENT PRACTICES AND GAP ANALYSIS TECHNICAL MEMORANDUM
The risk criteria used in previous studies to identify critical pipeline segments should be re-evaluated and
applied to appurtenances. The current CIP identifies critical valves solely based on size. In order to
make this consistent with AWWA or ISO 550001 standards, additional scoring criteria (related to
consequence of failure) should also take into account triple-bottom line approach (such as impact to
critical water users, proximity to environmentally sensitive areas, and cost of replacement/depth, etc.).
The International Standards Organization (ISO) 55001 for Asset Management identified “Performance
Evaluation” as a critical component of any Asset Management System. “The processes for monitoring
performance should address … the setting of performance metrics, including qualitative and quantitative
measurements (financial and non-financial) that are appropriate to the needs of the organization;
reactive measures of performance to monitor failures, incidents, non-conformities (including near misses
and false alarms) and other historical evidence of deficient asset management system and activity
performance; recording the data and results of monitoring and measurement, sufficient to facilitate
subsequent corrective action analysis…”. No evidence has been found of documented current or future
Level of Service (LOS) goals and the priority of each. In addition, no evidence has been found of a plan
or tracking mechanism to regularly monitor/publish/audit LOS goals. Examples of this include:
2
Per AWWA Manual M17, “all hydrants should be inspected regularly, at least once a year, to ensure
their satisfactory operation (most manufacturers recommend twice per year).” In addition, “it is good
practice to conduct flow tests on all parts of the distribution system approximately every 10 years to
identify the service areas affected by significant changes in the distribution system. An accurate and
digital record should be kept of each flow test so it is readily available.”
arcadis.com 108
ASSESSMENT OF CURRENT PRACTICES AND GAP ANALYSIS TECHNICAL MEMORANDUM
When this information is tracked and reported annually, it can be compared to other similar sized water
utilities to measure progress over time. Please see the 2013 AWWA Benchmarking Report for more
information on best-in-class key performance indicators. Performance measurement systems beyond
AWWA’s Utility Benchmarking Program include the Kaplan and Norton balanced scorecard and the
GASB performance measurement framework. These tools offer additional approaches for organizing
performance measurement.
According to the March 2016 Flint Water Advisory Task Force Final Report, “Flint’s lack of reinvestment in
its water distribution system contributed to the drinking water crisis and ability to respond to water quality
problems.” “Most of the department’s Standard Operating Guidelines have neither been reviewed nor
updated for more than a decade and some were last reviewed in the 1990s. The Flint Fire Department’s
standard operating guidelines require that each guideline is to be reviewed at least once every 24 months
or when a change in policy is made…“. New SOPs to be developed include those shown in Table 4-6.
SOP Task
Procurement
Pre-planning and Coordination
Valve/Hydrant Installation and Replacement Procedures Inspection Before Installation
Installation and Replacement
Testing and/or Inspection after Installation
Valve/Hydrant Preventive Maintenance Program and Specific to Valve Type, Manufacturer, Size, etc.
Procedures
Shutdown
Customer Notifications
Valve Emergency and Corrective Maintenance Inventory and location of spare/replacement
Procedures parts
Contact information for potential contractors to
assist
Work Order Documentation Protocols (identifying cause of break was due to valves)
Community Fire Risk Assessment Fire Flow Testing and occupancy risk
No evidence has been found confirming manufacturer manuals are available or used to
establish/verify preventive maintenance requirements.
No evidence found of hydrant/valve coordination/training with other departments/agencies (fire
department, traffic accident reporting, street cleaning, sewer cleaning, building construction,
recreation, contractors, etc.)
No evidence was found confirming the public is notified of a potential impact to service during
planned or unplanned maintenance
No evidence was found on how hydrant testing results are incorporated into the hydraulic model
calibration
arcadis.com 109
ASSESSMENT OF CURRENT PRACTICES AND GAP ANALYSIS TECHNICAL MEMORANDUM
4.5.6.1 Background
Conditions in distribution systems can change frequently and, in some cases, drastically for a number of
reasons, including changes in water demand, changes in source water quality and treatment, main
breaks, and maintenance activities. These changes can impact everything from distribution system
pressure and flow to water quality. Managing these potential impacts to maintain the desired pressure,
flow and water quality requires data to understand what is happening in the distribution system. Routine
sampling and visual inspection can be effective tools, but do not provide real-time information to respond
to an event until, in many cases, it is too late. Real-time, online monitoring can be used to identify a
potential problem before it occurs and minimize water quality issues or service disruptions for customers.
Online water quality monitoring (OWQM) can be used as a potential indicator of drinking water
contamination for a variety of contaminants including pesticides and pathogens. The Partnership
program provides recommendations on the number of OWQM stations and parameters monitored both at
the point of entry (POE) and within the distribution system based on system size. These guidelines are
largely based on contamination warning system research and pilot studies performed by the USEPA,
which also includes guidelines on selecting distribution system OWQM locations, the types of OWQM
instrumentation, and event detection systems.
The Partnership program also recommends implementation of a continuous monitoring system to monitor
flows and pressures at tanks, pump stations and other key points in the distribution system to identify
potential water loss events, which is discussed in Sections 4.5.10 and 4.5.12.
Control Station No. 2: This station includes two Hach CL17 free chlorine analyzers, one to
measure the chlorine residual as the water is received from GLWA and one downstream of the
chlorine boosting.
Cedar Street Reservoir: This station includes two Hach CL17 free chlorine analyzers, one
upstream and one downstream of chlorine boosting.
West Side Reservoir: This station includes one Hach CL17 free chlorine analyzer, located
upstream of chlorine boosting, and a new Hach CLF10 free chlorine analyzer, located
downstream of chlorine boosting.
These are all used for process control purposes only. Data are stored at a local programmable logic
controller (PLC) and are communicated via cellular to the SCADA system.
The City is currently considering installation of additional Hach OWQM instrumentation, given their
familiarity with this manufacturer, at the POE and at six locations throughout the distribution system.
Hach manufactures a distribution system monitoring panel (DSMP), which provides real-time analysis of
chlorine residual, turbidity, pH, conductivity, pressure and temperature – all critical parameters to
maintaining distribution water quality. In addition, an optional on-line phosphate analyzer can also be
provided in the DSMP. In addition to the DSMP equipment, Hach will also provide maintenance,
calibration and field support services.
arcadis.com 110
ASSESSMENT OF CURRENT PRACTICES AND GAP ANALYSIS TECHNICAL MEMORANDUM
The City is currently staff limited and engaged in a number of rehabilitation and optimization efforts at
both the Flint Water Treatment Plant and in the Flint distribution system. The DSMP and associated
service contracts will enable the City to continue to monitor the improvements within the distribution
system and respond to potential water quality upsets in a timely manner, as well as understand what
might have caused those upsets in the event a water quality event occurs.
It is recommended that the City proceed with purchasing of six Hach DSMP units and associated service
agreements for calibration, maintenance and field support as was previously recommended by others.
The location of those units will be determined after assessing the City’s water distribution system
hydraulic model currently being modified by the USEPA.
Chlorine residual analyzers have been installed upstream and downstream of chemical addition points to
provide dosing feedback.
None identified.
For systems with populations greater than 50,000, the Partnership recommends, at a minimum, the
following security-related online monitoring:
POE monitoring chlorine residual, pH, turbidity temperature, pressure and conductivity
Three distribution sites monitoring chlorine residual, pH, turbidity temperature and conductivity,
Five distribution system pressure monitoring sites
Total organic carbon (TOC) monitoring at the largest POE and at one point within the system.
Event detection system.
With the exception of TOC monitoring, the Hach DSMP would exceed these requirements. However,
event detection, either via a software or by using automated alarm levels for key parameters, would also
be required.
4.5.7.1 Background
Pipeline installation, rehabilitation, and replacement practices have a long-term effect on water
distribution system infrastructure condition and performance. Proper pipeline design and installation
combined with ongoing rehabilitation and replacement are important practices for extending the useful life
arcadis.com 111
ASSESSMENT OF CURRENT PRACTICES AND GAP ANALYSIS TECHNICAL MEMORANDUM
of the system, minimizing leaks and breaks, and achieving the desired water quality and quantity at
customers’ taps. Industry standards, such as Ten States Standards and AWWA C600, C603, C605, for
the proper design and installation of pipelines, can maximize asset life, minimize leaks and breaks, and
deliver desired water quantity and quality.
Disinfection of new mains and placement into service appears to be conducted in accordance with
standard practice.
According to AWWA Standard G200, utilities shall have a formal, standardized design procedure that
provides for a comprehensive review of all construction projects. Also, utilities shall prepare as-built
drawings of all installed facilities and shall maintain records associated with inspection, design, and
construction of all new and retrofitted facilities. The City has developed technical specifications and
standard construction details for new watermain installations and replacements. However, there is no
established schedule for reviewing and/or updating these documents.
arcadis.com 112
ASSESSMENT OF CURRENT PRACTICES AND GAP ANALYSIS TECHNICAL MEMORANDUM
It is recommended that the existing technical specifications and construction details be reviewed and
updated to current standards. Any updates should be based on the Ten States Standards (2012 or latest
version) and AWWA with specific criteria developed for the City of Flint.
As per AWWA Standard G200, the City should develop a formal risk-based approach as the foundation of
their watermain replacement/rehabilitation program. Comprehensive and advanced analytical tools are
available within GIS and hydraulic models that can be applied to identify the most critical individual pipe
segments based on age, breaks, water quality concerns, hydraulic data, customer type, number of
service connections, lead services, maintenance costs, and impacts to downstream customers. Ideally,
this program is developed by the City but at a minimum, if it is prepared by an outside
consultant/contractor, the City should be directly involved to assist in identifying and prioritizing required
investments based on a formal business analysis and addressing the most critical infrastructure in a
systematic way.
AWWA / Industry standards recommend replacing a minimum of 1% of watermains per year with a goal of
maintaining the age of all piping with the system at less than 100 years. To replace 1% of watermain per
year would require the City to finance nearly 6 miles of replacement projects per year. Working towards a
goal of all watermains being less than 100 years old by the year 2030 would require the replacement of
over 28 miles per year from 2017 to 2030. The financial hardships of the City may make either goal
difficult to achieve.
To develop a comprehensive approach to prioritizing the watermain replacements, the City should
develop a program for assessing the physical condition of the pipes. This will allow for a more focused
prioritization of the capital expenditures.
None identified.
4.5.8.1 Background
The performance of pumping stations is integral to maintaining benchmark pressures and flows
throughout the distribution system. Adherence to SOPs for pumping stations can prevent hydraulic and
even structural damage to distribution system piping and valves. According to the Partnership, in
optimized distribution systems, time spent on preventative maintenance should outweigh time spent on
corrective repairs by a ratio of four to one (AWWA, 2011).
Maintaining accurate records on pump operability, redundancy, and working capacity in a comprehensive
asset management plan is crucial to ensuring pressure and flow stability, even during unexpected events
such as main breaks or equipment malfunction. The aforementioned factors can be summarized in a
pumping facility assessment that includes the following:
Pump asset management – criteria for replacement and rehabilitation and inspection results
arcadis.com 113
ASSESSMENT OF CURRENT PRACTICES AND GAP ANALYSIS TECHNICAL MEMORANDUM
Pump redundancy – backup supply to adequately serve the maximum day demand at a
continuous pressure of 20 psi
Pump maintenance – ratio of planned maintenance to total maintenance
arcadis.com 114
ASSESSMENT OF CURRENT PRACTICES AND GAP ANALYSIS TECHNICAL MEMORANDUM
pressurizes the system, contributing to water main breaks. The smaller pump is nearing
the end of its useful life and is difficult to keep operating consistently.
Dort Reservoir Pump Station (PS No. 4)
o This station consists of five pumps but two are considered inoperable. Capacities are
provided below:
o Pump 1: 20 MGD, out of service
o Pump 2: 25 MGD, out of service
o Pump 7: 20 MGD, available for operation
o Pump 8: 20 MGD, available for operation
o Pump 9: 6 MGD, available for operation
o This station is primarily used to provide water supply to the system from Dort Reservoir
during peak demand or emergency conditions.
o According to city personnel some work on the two pumps that are out of service has
taken place but it is unclear if they are currently operational.
None identified.
Install pump control valves at West Side Pump Station to reduce the potential for pressure surges during
startup and shutdown of the pumps.
AWWA Standard G200 recommends the development of written SOP’s for the operation of each pumping
facility with operating logs maintained for variables such as suction and discharge pressures, flow rate,
pump hour meter readings, and chlorine residual. The data should be recorded on the WTP’s SCADA
servers and be available for historical reference.
AWWA Standard G200 also recommends that detailed maintenance procedures be developed for each
pump station defining the frequency, procedures, and maintenance of records for each piece of
mechanical and electrical equipment. It is our understanding that the City has developed SOP’s and
detailed maintenance procedures for some activities. These should be expanded upon and with
additional items for consideration listed below:
arcadis.com 115
ASSESSMENT OF CURRENT PRACTICES AND GAP ANALYSIS TECHNICAL MEMORANDUM
failures. Typical items include rehabilitation of the pump rotating assembly which involves
replacement of bearings, wear rings, shaft sleeves, packing, or mechanical seals. The impeller
should be inspected for premature wear and rebalanced if necessary. This is also an opportune
time to have the motor bearings replaced. Based on the number of pumps in the distribution
system, this would require rehabilitating 3 to 5 pumps per year.
None identified.
Conduct a needs assessment at each pump station to identify short and long-term capacity requirements
as well as other capital needs. With the downward trend in water consumption within the City, smaller
capacity pumps may be necessary.
4.5.9.1 Background
Security and Emergency Management policies and practices are essential to ensuring the safe
functioning and overall integrity of the distribution system. Security infrastructure and practices actively
protect and decrease the vulnerability of assets and systems to threats such as sabotage or terrorism that
could compromise the infrastructure, the ability to maintain service in the distribution system, or the water
quality. Despite best security practices, the distribution system cannot be made perfectly secure, so
Emergency Management procedures and practices are essential to ensuring that appropriate responses
are taken in case of an emergency and to facilitate quick recovery after an emergency.
After the passage of the Bioterrorism Act in 2002, all utilities serving more than 3,300 people were
required to perform a security vulnerability assessment and prepare an ERP. Updates of the vulnerability
assessment or the ERP are not required, although updates to both are recommended according to
various industry standards, such as those outlined in Table 3-2.
Operators are not always out in teams when visiting dangerous locations at night;
arcadis.com 116
ASSESSMENT OF CURRENT PRACTICES AND GAP ANALYSIS TECHNICAL MEMORANDUM
As recommended in the Ten States Standards (GLUMRB, 2012), the City meets industry standards
requiring all exterior entrances to lock, fencing around facilities (with the notable exception of the Cedar
Street Reservoir and Pump Station, which is not entirely enclosed in fencing), intrusion alarms, and some
security camera monitoring.
While the City does have intrusion alarms, minor strengthening is recommended in the area of access
control and intrusion detection per guidance in the AWWA Standard G430-14 Security Practices for
Operation and Management (AWWA, 2015). Specific guidance includes: establishing and maintaining
physical control of access to identified critical assets, a means of restricting authorization for access, a
protocol for employees that have had a relevant change in status, and testing. Additionally, annual
inspections of identified critical assets are recommended.
Minor hardening of the Cedar Street Pump Station and Reservoir security is recommended through
completely enclosing the facility in fencing, per guidance in the Ten States Standards and AWWA
Standard G430-14, especially given the history of vandalism at the pump station. Additionally, the fence
should be maintained clear of plant growth, which significantly decreases the efficacy of the fence.
The ERP requires major strengthening. The existing ERP does not meet FEMA standards (FEMA, 2010),
(USDHS, 2013), (USDHS, 2008) and should be updated upon large changes in operations. Specific steps
include forming a planning team, understanding the threats and risks, determining the operational goals,
and developing a plan.
Regular exercise of the ERP also requires major strengthening and is recommended in the Partnership
for Safe Water Self-Assessment Guide under ‘Security, Emergency Management’ (AWWA, 2010), AWWA
Standard G430-14, and the AWWA Standard G440-11 Emergency Preparedness Practices (AWWA,
2011). Specific guidance includes the following: regular review and revision of the plan and contact lists,
arcadis.com 117
ASSESSMENT OF CURRENT PRACTICES AND GAP ANALYSIS TECHNICAL MEMORANDUM
frequent exercise of emergency response, evaluation and reporting of emergency response exercises,
and development and updates of SOPs.
The development and promotion of a security culture is an area that appears to require major
strengthening. This includes commitment from senior leadership and promotion of security awareness, as
recommended in AWWA Standards G430-14 and G440-11. Other areas important to fostering a security
culture include: defined security roles and expectations for employees, dedication of resources to
maintaining security, and identification of security priorities.
The development and strengthening of communications and partnerships is also an area that appears to
require major strengthening to enable Flint to best respond in emergency situations, as recommended in
AWWA Standards G430-14 and G440-11. Strong partnerships and communications will also support the
creation of a Unified Command Structure in the case of large-scale incident response operations. As part
of the preparedness process specific guidance includes a focus on both internal communications with
employees as well as external communications and partnerships with response organizations, customers,
and regulatory agencies. It is also recommended that the City develop mutual aid and assistance
partnerships.
arcadis.com 118
ASSESSMENT OF CURRENT PRACTICES AND GAP ANALYSIS TECHNICAL MEMORANDUM
4.5.10.1 Background
The design, operation, and maintenance of distribution system storage facilities can have a substantial
impact on water quality throughout the distribution system. Tank design including baffling and internal
walls will often determine the level of mixing achievable in the storage facility without the aid of
mechanical mixers. However, the use of effective standard operation procedures including tank turnover
plans can avoid chlorine and water quality degradation. A comprehensive plan for storage facility
operations should include:
• Tank turnover plan including optimal operating water level
• Inspections schedule - typically annually
• Safety and security
Routine inspections and maintenance of tanks can prevent structural failures leading to contamination
events. AWWA standards D102, D103, and D104 cover coatings, paintings, and cathodic protection of
water tanks and AWWA Manual M42 describes inspection events to prevent structural failures. Storage
facility age relative to expected useful life is an important metric to track in a comprehensive asset
management plan to forecast operational issues and stay ahead of major failures. Understanding the
construction materials used for facilities and having a plan in place for rehabilitation and replacement of
tanks near the end of their expected useful life is key to minimizing impacts to the water system
arcadis.com 119
ASSESSMENT OF CURRENT PRACTICES AND GAP ANALYSIS TECHNICAL MEMORANDUM
As discussed elsewhere in this report, inlet control valves to Cedar Street and West Side Reservoirs open
and close too quickly which can create pressure surges in the system resulting in watermain breaks in the
distribution system.
Recent modifications to improve turnover, conduct monitoring with depth to verify conditions, and to
improve controls of boosting operations are positive operational steps that meet industry standards.
Replace inlet valves to West Side and Cedar Street Reservoirs with globe body style altitude valves
having back pressure sustaining pilots. This will maintain an acceptable pressure within the distribution
system during the filling operation at each reservoir. The opening and closing rate of these valves can be
adjusted to minimize the potential for pressure surges.
Per AWWA Standard G200 and Manual of Practice M42, establish formal procedures for the inspections
of each storage facility. These inspections should focus on structural, safety, security, coatings system,
and sanitary conditions. It is recommended that the following timeframes be established for the
inspections:
arcadis.com 120
ASSESSMENT OF CURRENT PRACTICES AND GAP ANALYSIS TECHNICAL MEMORANDUM
o Roles and responsibilities have to be clearly defined throughout the process including
isolating the tank from the system, draining, cleaning, inspecting, disinfecting,
bacteriological testing, and returning the storage facility to service. Utilities typically self-
perform the isolation, draining, and returning to service. Cleaning, disinfecting, and
bacteriological testing may be performed by either party depending on the capabilities,
available capacity of the utility to complete this work, and willingness to accept
responsibility from a liability standpoint. The inspection should be completed by the
contractor.
o SOPs should be prepared by the City for isolating the storage facility, draining and
dechlorination procedures, system operations while the storage facility is out of service,
disinfection, and returning the facility to service.
o Disinfection procedures should be developed according to AWWA C652.
o The City should also develop standard specifications and project requirements for the
comprehensive inspections to provide consistency of the actual inspections and reporting
each time the work is performed.
Detailed hydraulic analysis of storage facility capacity and optimal operations, considering decreased
system demand, need for adequate capacity during main breaks, and planned operational changes
associated with switch to KWA and operation of Flint WTP.
4.5.11.1 Background
Minimizing water age, also referred to as retention time, is an important aspect of maintaining adequate
water quality within a water distribution system. While water quality delivered to a customer is affected by
many complex chemical, physical, and hydraulic factors, water age is often a key aspect of many water
quality problems, as shown in Table 4-7.
Table 4-7:Water Quality Problems Affected by Retention Time (Source: Brandt, 2004)
arcadis.com 121
ASSESSMENT OF CURRENT PRACTICES AND GAP ANALYSIS TECHNICAL MEMORANDUM
An average system water age can be determined by simply comparing the volume of water in the
distribution system pipe network and storage facilities with the total volume of water supplied into the
system. Areas of high water age within the system are often the focus of water age management.
Calibrated water distribution system models can estimate water age throughout the system and identify
areas of concern.
Various techniques exist to manage water age within a system (Brandt, 2004). These techniques include:
arcadis.com 122
ASSESSMENT OF CURRENT PRACTICES AND GAP ANALYSIS TECHNICAL MEMORANDUM
arcadis.com 123
ASSESSMENT OF CURRENT PRACTICES AND GAP ANALYSIS TECHNICAL MEMORANDUM
Optimization of storage facility operations – Reservoirs are currently operated in a deep cycle
pattern to fully draw down over multiple days. Reservoir operation needs to be better understood
to determine the impacts on water age, pressures, and water quality.
Manual flushing in areas of concern. Manual flushing practices must be strategically located to
turnover water in the highest age areas.
Storage Needs Assessment – A review of overall system water demands, storage needs and
capacity. With declining demands, total storage capacity must be periodically reviewed to
manage increasing system water age.
Water system demand projections – identify methods to increase demand in areas of high water
age, project future water age based on customer demand changes (residential, commercial,
industrial), and develop a plan for system operation as demand decline which maintains adequate
water age.
Unidirectional Flushing Program (Refer to Section 4.5.3)
Tracer evaluations to confirm hydraulic model predictions of water age
4.5.12.1 Background
Best practice for quantification and management of water loss in water distribution systems is defined in
AWWA Manual of Water Supply Practice M36 Water Audits and Loss Control Programs, Fourth Edition
(2016). This latest edition enhances the content and methodology published in prior editions and reflects
the updated AWWA Policy Statement on Metering and Accountability (last revised June 8, 2014), which
recommends that:
“…every water utility accurately meter all water taken into its system and all water distributed
from its system at its customers’ points of service. Meters should be read at sufficiently
frequent intervals appropriate to support the utility’s understanding of volume of production,
rate structures and to provide accurate bills and feedback to its customers”, and that
“…utilities conduct annual water audits to ensure accountability and effective use of metering
information.”
Water audits evaluate the effectiveness of metering and meter reading systems, as well as billing,
accounting, and loss control programs. Accurate metering, water auditing and effective loss control ensure
an equitable recovery of revenue based on level of service and wise use of available water resources.
The State of Michigan’s requirements for water loss control, in the broader context of maximizing the
efficiency of water resource utilization, are embodied in Part 327 Great Lakes Preservation, of the Natural
Resources and Environmental Protection Act, 1994 PA 451, as amended (NREPA). In addition to looking
arcadis.com 124
ASSESSMENT OF CURRENT PRACTICES AND GAP ANALYSIS TECHNICAL MEMORANDUM
more favorably at applications for permits to increase water supply by systems that demonstrate effective
water loss control, incentives for water systems to effectively manage their water losses are provided
through the Drinking Water State Revolving Fund’s program for funding (with principal forgiveness) for
projects involving water loss reduction and prevention.
The MDEQ worked with the AWWA Water Efficiency and Conservation Committee to develop sector-
based conservation measures for community water supplies as required by Part 327. These conservation
measures support the following principles and practices: efficient use of supply sources, appropriate
facility rehabilitation or replacement, leak detection and repair, accurate monitoring of consumption and
billing based on metered usage, as well as more traditional conservation-oriented activities. The AWWA’s
recommended water conservation practices are outlined in Guidelines for Generally-Accepted Water
Management Practices for the Public Water Supply (MI-AWWA, 2008).
Regarding the oversight of Michigan water utilities’ water loss control performance, as part of the sanitary
survey process for water utilities, the MDEQ conducts a review of the total water pumped compared to the
amount of water for which the water utility has billed. “Water unaccounted for” represents the water loss
through system leaks and peripheral consumptive use processes, which are routinely in the 5 to 10
percent range. The MDEQ advises the community to address water losses when unaccounted for water
reaches 15 percent of the total volume pumped. However, it should be noted that the term “unaccounted
for water” is not consistent with the best practice methodology described in AWWA Manual M36. Also, the
M36 methodology recommends moving away from percentage-based performance indicators of water
loss in favor of volumetric-based performance indicators. The national AWWA Water Loss Control
Committee recently assembled a Task Force to evaluate candidate replacements for percentage-based
indicators, with work to be completed in roughly 2 to 3 years. The AWWA Free Water Audit Software
(FWAS) calculates both types of indicators. When combined with monetary valuations of the components
of water loss volume (apparent and real) as defined in the AWWA Water Balance, a more defensible
basis will be established for developing a cost-effective water loss control plan that recognizes system-
specific characteristics, including operational, water resource availability, and financial considerations.
arcadis.com 125
ASSESSMENT OF CURRENT PRACTICES AND GAP ANALYSIS TECHNICAL MEMORANDUM
For the most part, Flint does meter water delivered and sold to customers and tracks main breaks and
leaks. However, there are several functional areas related to water loss management that can be
improved. Key to this effort, as shown in the above figure, is consistently producing and collecting valid
data needed to perform an annual water audit, and after improving data validity, quantifying and placing a
monetary value on the volume of real and apparent water loss components, and identifying areas of
opportunity to achieve economic levels of water loss performance, after establishing appropriate
performance targets. Monitoring of performance going forward using key performance indicators
described in the AWWA M36 Manual will be essential for maintaining cost-effective water loss control.
As noted in the Echologics report, Flint’s Data Validity Score calculated with the aid of the AWWA FWAS
was 48 out of 100, indicating that while some activities, such as leak detection and repair, and pipe
replacement that will likely produce relatively quick returns on investment can go forward, further work is
needed to establish a sound basis for a more complete Water Loss Control Plan, minimizing the chances
that unnecessary and uneconomical activities will be implemented.
None identified.
None identified.
Perform annual water audits in accordance with the AWWA M36 Methodology.
arcadis.com 126
ASSESSMENT OF CURRENT PRACTICES AND GAP ANALYSIS TECHNICAL MEMORANDUM
Develop a Water Loss Control Plan that aligns with the Water Department’s organizational goals and
objectives, and adjust, as appropriate, going forward. The following steps are recommended as part of
the plan development:
Perform an annual water audit, using AWWA FWAS, to determine volume and value of non-
revenue water
By applying the Validity Grading Matrix within the FWAS, determine the validity of underlying
components of water audit inputs and improve Data Validity Score to the level needed to
establish performance targets
Determine present expenditures for water loss management-related activities
Supply meters
o Perform an annual meter accuracy determination of all distribution system entry point
flow meters, including volumetric test and calibration of associated instrumentation
o Apply reported errors in the annual water audit
Read-to-Bill System
o Verify the presence of all customer service connections (active and inactive in
accounts database
o Verify the accuracy of meter readings
o Verify the accuracy of billed metered consumption
o Evaluate options for advanced meter reading/advanced metering infrastructure
technologies (AMR/AMI) and implement cost-effective technology
Authorized Consumption
o Quantify all types of authorized consumption (billed metered, billed unmetered,
unbilled metered, unbilled unmetered)
o Align all authorized consumption with SCADA-reported flow through distribution
system entry point supply meters, to eliminate “lag time” associated with meter
reading cycles
Apparent Loss Control
o Residential water meters – determine the accuracy of representative sample of
existing and newly purchased meters on ongoing basis and maintain operation within
applicable AWWA accuracy standards through program of R&R, including review of
meter typing and sizing
o Commercial water meters – determine the accuracy of representative samples of
existing and newly purchased meters on ongoing basis and maintain operation within
applicable AWWA accuracy standards through program of R&R, including review of
meter typing and sizing
o Unauthorized consumption – Evaluate/quantify sources of unauthorized consumption
and implement actions to minimize
o Systematic data handling error – Quantify volumes associated with errors in
converting meter readings into billable consumption. Modify read-to-bill system
procedures as appropriate.
Real Loss Control
o Quantify real loss component volumes
o Evaluate active leakage control (leak detection)
o Evaluate leak/break data to determine root causes of leakage
Improve consistency and content of incident reports to provide actionable
information for effective leakage management
arcadis.com 127
ASSESSMENT OF CURRENT PRACTICES AND GAP ANALYSIS TECHNICAL MEMORANDUM
Establish an internal Water Loss (Non-Revenue Water (NRW)) Management Team to develop Flint’s
Water Loss Control Plan and support effective utility management. The Team will be tasked with
determining and achieving economic levels of water loss.
Consider participating in the AWWA Water Loss Control Committee Water Audit Data Initiative to share
performance data and further the development of guidance for effective water loss control program
planning.
4.5.13.1 Background
Several SDWA regulations require monitoring within the distribution system to demonstrate compliance.
Those applicable to the City of Flint and regulated parameters include:
arcadis.com 128
ASSESSMENT OF CURRENT PRACTICES AND GAP ANALYSIS TECHNICAL MEMORANDUM
o Heterotrophic bacteria
arcadis.com 129
ASSESSMENT OF CURRENT PRACTICES AND GAP ANALYSIS TECHNICAL MEMORANDUM
Develop baseline data to help City staff better understand prevailing conditions within the
distribution system.
Provide a tool to allow water quality data to be evaluated and assessed and to potentially provide
an early-warning of water quality upsets.
Provide a source of up-to-date water quality information to be considered by City staff for CIP
planning, operations, and maintenance activities.
Recommended parameters and associated rationale for inclusion in the City of Flint’s Surveillance
Monitoring Plan include:
Free Chlorine: An adequate chlorine residual (achieving the goal of 0.2-0.5 mg/L in 95% of
monthly samples) is fundamental for maintaining microbial water quality, controlling metals
release, and meeting State and Federal drinking water regulations (e.g. RTCR, SWTR).
Heterotrophic Plate Count Bacteria: A more general indicator of the extent of microbiological
activity compared to total coliform. R2A agar is a more sensitive, low nutrient agar that provides
a better indicator of heterotrophic growth compared to Standard Plate Count (SPC) agar.
Turbidity: Is an indicator of the extent of solids in the system, amount of sediment entrained in
flow, and potential scale instability and/or sloughing. Should also be used to terminate flushing
operations.
pH and Temperature: pH is a governing variable for much of water chemistry and corrosion
control. The rate of all chemical and microbiological interactions is temperature dependent and,
as such, temperature is an important factor when considering conditions effecting microbial
growth and metals release.
Iron: Is an indicator of the effect of pipe material and the stability of corrosion scales on water
quality.
Apparent Color: Is an indicator of the amount of visible color in a water sample. It does not
provide an indication of the source of color, and the sample is not filtered to remove particulates.
If well correlated with iron, it is a more rapid technique for field application to track aesthetic
characteristics of the water.
The City currently monitors for all of the parameters listed above (aside from apparent color) as part of
various regulatory compliance programs. For example, the City performs heterotrophic plate counts using
SimPlate plates and media supplied by Idexx. Analysis is performed on one of three rounds of routine
coliform samples collected each week. A plant supply sample is collected and analyzed for HPC with
each sample set. HPCs in the plant supply and routine coliform samples are typically less than 10 colony-
forming units per millilter (CFU/mL) and are rarely above 100 CFU/mL. Under the SWTR, the City is
required to collect a sample and conduct a heterotrophic plate count (HPC) whenever more than 5% of
chlorine residual samples are non-detectable. Per MDEQ regulations for public water systems, water in
the distribution system that has a heterotrophic plate count (HPC) less than or equal to 500 CFU/mL is
considered equivalent to having a detectable disinfectant residual. While the City is using an approved
arcadis.com 130
ASSESSMENT OF CURRENT PRACTICES AND GAP ANALYSIS TECHNICAL MEMORANDUM
laboratory method for HPCs under the Surface Water Treatment Rule, industry best practices have found
that use of R2A agar is more appropriate for culturing HPCs. It is estimated to provide 10-100X higher
results, providing more valuable insight into microbial conditions, problem areas, and effectiveness of
mitigation strategies.
Although surveillance water quality parameters summarized above are routinely collected by the City,
they are not entered into electronic spreadsheets that can be readily used for tracking, correlating, and
trending. This limits the usefulness of the information for characterizing system conditions and prioritizing
problem areas for remediation activities. Additionally, there are opportunities for streamlining the number
of sample locations to maximize the amount and type of information collected from specific locations,
allowing for pairing whenever possible. Streamlining will also reduce complexity and labor requirements,
will improve consistency of results, and will facilitate data tracking and trending. Development of a
streamlined surveillance monitoring program will be conducted as part of a future project task.
The City’s 2016 Quality Assurance Program document was also reviewed for the purposes of
understanding City capabilities and practices. The need for modifications to the QAP and development of
additional SOPs will be addressed in future tasks.
RTCR Plan and chlorine surveillance monitoring plans meet industry standards
EWQP monitoring exceeds industry standards in terms of monitoring frequency and parameters
included.
arcadis.com 131
ASSESSMENT OF CURRENT PRACTICES AND GAP ANALYSIS TECHNICAL MEMORANDUM
4.6.1 Administration
arcadis.com 132
ASSESSMENT OF CURRENT PRACTICES AND GAP ANALYSIS TECHNICAL MEMORANDUM
4.6.1.2 Funding
Financial resources are required to pursue distribution system optimization. Adequate rate structures
provide the funding for obtaining professional operating, maintenance, and laboratory staff; training of
these personnel; purchasing laboratory supplies; and maintaining equipment and existing facilities
necessary to provide high quality water and uninterrupted service.
Despite its relatively high water rates, it is well understood that the Water Department is underfunded. No
capital improvement plan is funded for the water distribution system, and all maintenance is funded in a
reactionary manner. In prior years, the Water Department has requested moderate levels of funding for
proactive distribution system rehabilitation and replacement, but those requests have not made it into the
City’s budget. In addition, all purchases no matter how minor, including cleaning supplies, toilet paper,
etc. must be approved by the City’s Purchasing Department. That department, like much of the City, is
understaffed (only 2 employees) and it takes an inordinate amount of time to make minor purchases
within the Water Department.
The AWWA Benchmarking Survey found that the median annual system renewal rate is 3 percent of the
total present worth of the system. The range of annual renewal percentage from the survey was:
25th percentile – 1%
Median – 3%
75th percentile – 11%
The City is currently budgeting zero percent for system renewal. An appropriate level of renewal
investment will be developed during preparation of the final Distribution System Optimization Plan and will
include a prioritized list of improvements. It is likely to be in the 5 percent range, if not higher due to
current system condition and criticality of major distribution system assets. In addition, the Water
Department should be afforded an appropriate level of funding for routine facility purchases (cleaning
supplies, toilet paper, etc.).
4.6.1.3 Staffing
Under the Partnership, utility staffing must be sufficient to continuously maintain system performance
goals. This is especially important for distribution system operation, where professional operating staff is
required to monitor system function at all times and to make adjustments to achieve performance goals.
For a system operating continuously, staffing levels should be adequate to provide around-the-clock
coverage; including allowances for staff absences (like vacations, sick leave, and training).
An organization chart for the City of Flint Utilities Department is presented in Figure 4-31. The Water
Department is divided into two main departments, the WSC and the WTP. The WSC is responsible for
distribution system operations and maintenance, specifically piping, valves, and hydrants, and the WTP is
responsible for water quality and operations and maintenance of the Flint WTP as well as Cedar Street
and Westside Reservoirs and Pumping Stations for pumping treated water at prescribed pressures into
the City of Flint’s water distribution system. A summary of budgeted, filled, and vacant positions for each
of those departments is provided in Table 4-8.
arcadis.com 133
ASSESSMENT OF CURRENT PRACTICES AND GAP ANALYSIS TECHNICAL MEMORANDUM
Account Clerk 1 1 0
Building Maintainer 1 0 1
Deputy Supervisor 1 1 0
TOTAL 39 27 12
Electrician 1 1 0
TOTAL 24 20 4
The Resource Analysis and Needs Assessment task, which follows this gap analysis, includes staff
interviews and development of a competency-based training program. Recommendations regarding
workforce roles, job descriptions, and staffing levels will be made as a part of that assessment. That said,
a few observations can be made regarding the current WSC (i.e., distribution system O&M) staffing
levels. As can be seen in Table 4-8, nearly one-third of the water distribution positions (12/39) are
vacant, with the majority of those vacancies being in the “operator” and “operator trainee” categories. In
fact, ten of the twelve vacancies are in those categories. Based on interviews with WSC supervisory and
operations staff, most of the maintenance activities performed are reactive in nature, rather than
arcadis.com 134
ASSESSMENT OF CURRENT PRACTICES AND GAP ANALYSIS TECHNICAL MEMORANDUM
preventative, and prevent staff from completing non-repair related work, such as recordkeeping, in a
timely manner.
Training and staff retention are also challenging. No formal distribution system operator training program
currently exists. Training requirements and the associated program will be developed following the
Resource Analysis and Needs Assessment. It was also noted during interviews that many operator
trainees and operators leave for better paying positions at other local utilities after gaining experience at
the City.
Comparing performance results to other utilities and historical records can reveal trends and support
performance improvement efforts. The values shown below summarize several AWWA Benchmarking
Performance Indicators (AWWA, 2013) related to staffing and training. Staffing related benchmarking
measures included in the benchmarking survey include, customer accounts per employee, MGD
delivered per employee, and training hours per employee. A comparison of survey results to current
staffing and training for Flint is shown in Table 4-9. These results should not be viewed as goals but
rather useful comparisons for utilities funding optimization efforts.
1) Flint calculations based 31,520 customer accounts (Rowe, 2016b), 13 MGD annual average day, and Water
Service Center Distribution staffing numbers in Table 4-6.
Upon completion of the staff interviews and determination of the most appropriate staff mix to meet the
current and future needs of the water system, job descriptions should be prepared which describe the
qualifications, roles and responsibilities for each employee in the Water Department. A distribution
system operator training program should be developed and implemented. Water Department staff
salaries should be evaluated relative to comparable positions with other utilities locally and adjustments
should be made to better retain employees – who will now not only have outstanding field experience but
will be well trained. To address the vacant staff positions (when the appropriate staffing is determined),
the City should reach out to the community to find and employ qualified operators and operator trainees.
arcadis.com 135
ASSESSMENT OF CURRENT PRACTICES AND GAP ANALYSIS TECHNICAL MEMORANDUM
arcadis.com 136
ASSESSMENT OF CURRENT PRACTICES AND GAP ANALYSIS TECHNICAL MEMORANDUM
A defined risk methodology that complies with MDEQ’s guidance for evaluating risk,
A discussion of current and future LOS and associated costs and consequences of achieving
these LOS, and
Impacts to O&M budgets and water rates.
In addition to the items noted above, the Water Department lacks a computerized maintenance
management system (CMMS). The City does have a GIS database of its water infrastructure; however,
that database has not been updated recently. Similarly, although specific assets such as hydrants and
valves are tracked in spreadsheets as mentioned in previous sections, the City has not incorporated that
information into an integrated, comprehensive asset management program for the water system.
The City hired Echologics to conduct leak detection and prepare a non-revenue water assessment.
Echologics conducted leak detection on 564 miles of pipe and estimates the total pipe length in the
system to be just under 800 miles. This is counter to the City’s understanding – which is that there are
approximately 600 miles of pipe in the system – as well as the recent hydraulic model prepared by
USEPA which includes 579 miles of pipe. Obviously, a discrepancy exists. An accurate asset inventory
and CMMS would answer this question once and for all.
Similarly, the City hired Wachs Water to conduct a distribution system valve assessment. Wachs
evaluated 8,228 valves and determined approximately 35 percent of those valves to be in poor or
inoperable condition. A CMMS could store this valve inventory and be used to track and schedule valve
maintenance as well as the results of the condition assessment.
A CMMS or Enterprise Asset Management (EAM) system is needed. That system should be populated
with an accurate, up-to-date water treatment plant and distribution system asset inventory. A condition
assessment should be conducted for every water system asset, taking advantage of previous work by
Wachs and Echologics, and an estimate of remaining effective useful life (EUL) should be made for each
asset. Rehabilitation or replacement of those assets should be scheduled into a well-coordinated capital
improvements plan, and the City should budget annually for system rehabilitation and replacement.
arcadis.com 137
ASSESSMENT OF CURRENT PRACTICES AND GAP ANALYSIS TECHNICAL MEMORANDUM
The final Distribution System Optimization Plan will include a comprehensive review of the 2016 Asset
Management Plan and gap analysis followed by the development of an asset management plan that
complies with MDEQ’s format and requirements, using the 2016 report as a starting point. This plan will
include development of a condition assessment and criticality determination methodology, assessment of
distribution system asset condition, prioritization of R&R projects, and a forecast of future operating and
capital funding needs. In addition, the plan will include identification of EAM needs and recommendations
to a Water Department-wide (distribution and treatment) asset management program to assist the City
with long-term planning and management of the complete water system.
arcadis.com 138
ASSESSMENT OF CURRENT PRACTICES AND GAP ANALYSIS TECHNICAL MEMORANDUM
5.1 Summary
5.1.1 Approach
The City of Flint’s existing practices were evaluated using the Partnership for Safe Water optimization
method as a guide for enhancement of system performance based upon improving water quality,
physical, and hydraulic characteristics of the system. The standards developed therein were used to
identify gaps between the City’s practices and industry benchmarks. Using the self-assessment
approach, the distribution system was evaluated based on three major performance variables:
Disinfection,
Main breaks, and
Pressure.
These three performance variables are representative of water quality, physical, and hydraulic integrity of
the system. As such, analysis of various practices and goals associated with each variable leads to and
overall optimization status of the distribution system. While achieving full optimization under the
Partnership for Safe Water method is difficult, the benchmarks and standards in the method provide a
pathway for continuous improvement.
In addition to metrics tied to the three major performance variables, the Arcadis team performed a
detailed evaluation of corrosion control practices. Additionally, a high-level evaluation of administrative
areas including asset management, information technology, and staffing was conducted. These
administrative areas will be covered in more detail in the next phase of the Distribution Optimization Plan.
5.1.2 Gaps
Through the analysis against the Partnership optimization method and other industry standards, gaps in
performance were identified for each of the categories evaluated. Table 5-1 through Table 5-4 provide a
summary of the areas that meet or exceed industry standards, need minor strengthening, need major
strengthening, or do not have applicable programs. Additional detail for each of these areas is provided
in Section 4.0.
arcadis.com 139
ASSESSMENT OF CURRENT PRACTICES AND GAP ANALYSIS TECHNICAL MEMORANDUM
arcadis.com 140
ASSESSMENT OF CURRENT PRACTICES AND GAP ANALYSIS TECHNICAL MEMORANDUM
arcadis.com 141
ASSESSMENT OF CURRENT PRACTICES AND GAP ANALYSIS TECHNICAL MEMORANDUM
arcadis.com 142
ASSESSMENT OF CURRENT PRACTICES AND GAP ANALYSIS TECHNICAL MEMORANDUM
arcadis.com 143
ASSESSMENT OF CURRENT PRACTICES AND GAP ANALYSIS TECHNICAL MEMORANDUM
arcadis.com 144
ASSESSMENT OF CURRENT PRACTICES AND GAP ANALYSIS TECHNICAL MEMORANDUM
arcadis.com 145
ASSESSMENT OF CURRENT PRACTICES AND GAP ANALYSIS TECHNICAL MEMORANDUM
arcadis.com 146
ASSESSMENT OF CURRENT PRACTICES AND GAP ANALYSIS TECHNICAL MEMORANDUM
arcadis.com 147
ASSESSMENT OF CURRENT PRACTICES AND GAP ANALYSIS TECHNICAL MEMORANDUM
arcadis.com 148
ASSESSMENT OF CURRENT PRACTICES AND GAP ANALYSIS TECHNICAL MEMORANDUM
Continue to revise and expand the current RTCR monitoring program beyond the City’s current
20 sites to 25 sites.
Implement surge control at reservoirs and pump stations, and conduct analysis of future WTP
operations on distribution system pressures.
Use of hydraulic model to evaluate water age and opportunities for minimizing.
arcadis.com 149
ASSESSMENT OF CURRENT PRACTICES AND GAP ANALYSIS TECHNICAL MEMORANDUM
Siting and installation of distribution system pressure data loggers to verify pressure control.
Purchase and install online distribution system water quality monitoring panels. Prior to
installation, use the hydraulic model to confirm optimal locations for installation.
Develop and implement a distribution system operator training program.
Develop and implement a unidirectional flushing program.
Develop SOPs for routine maintenance activities and for those activities that impact water quality
(flushing, chlorine residual maintenance, etc.).
Increase funding for main replacement activities.
Conduct a local utility salary survey and adjust operator salaries as necessary to be competitive
in the local market.
Develop a hiring plan to fill vacant positions within the WSC Distribution Department.
Provide “whole house” flushing guidance to residents and encourage them to flush their homes
regularly until water quality within their home is restored.
Remaining gaps will be grouped, where possible, and combined with information obtained during the
upcoming Resource Analysis and Needs Assessment to develop a customized set of improvements for
the City’s distribution system. The improvements will be prioritized using the selected criteria and
weightings, which will be determined at the Resource Analysis and Needs Assessment Workshop, and
presented in the final Distribution System Optimization Plan.
arcadis.com 150
ASSESSMENT OF CURRENT PRACTICES AND GAP ANALYSIS TECHNICAL MEMORANDUM
6.0 REFERENCES
Alexander, M. T. & Dugan, A. G. (n. d.). Area-Wide Optimization Program’s Approach to Maintaining
Distribution System Water Quality. Retrieved from:
http://www.dep.pa.gov/PublicParticipation/AdvisoryCommittees/WaterAdvisory/TAC/Documents/AWOP_A
pproach_for_Water_Quality_Presentation%20.pdf
American Water Works Association (AWWA). (2001). M19 Emergency Planning for Water Utilities (4th
ed.). Denver, CO: AWWA
American Water Works Association (AWWA). (2006a). M17 Fire Hydrants: Installation, Field Testing and
Maintenance (4th ed.). Denver, CO: AWWA
American Water Works Association (AWWA). (2006b). M33 Flowmeters in Water Supply (2nd ed.).
Denver, CO: AWWA
American Water Works Association (AWWA). (2006c). M44 Distribution Valves: Selection, Installation,
Field Testing and Maintenance (2nd ed.). Denver, CO: AWWA
American Water Works Association (AWWA). (2010). J100-10 Risk and Resilience Management of Water
and Wastewater Systems. AWWA.
American Water Works Association (AWWA). (2011a). C652-11 Disinfection of Water Storage Facilities.
AWWA
American Water Works Association (AWWA). (2011b). G440-11 Emergency Preparedness Practices.
AWWA.
American Water Works Association (AWWA). (2011c). Partnership for Safe Water Self-Assessment
Guide for Distribution System Optimization. Denver, CO: AWWA.
American Water Works Association (AWWA). (2012a). M6 Water Meters – Selection, Installation, Testing
and Maintenance (5th ed.). Denver, CO: AWWA
American Water Works Association (AWWA). (2012b). M32 Computer Modeling of Water Distribution
Systems (3rd ed.). Denver, CO: AWWA.
American Water Works Association (AWWA). (2013a). Benchmarking – Performance Indicators for Water
and Wastewater: 2013 Survey Data and Analyses Report. Denver, CO: AWWA.
American Water Works Association (AWWA). (2013b). M42 Steel Water Storage Tanks, Revised Edition.
Denver, CO: AWWA.
American Water Works Association (AWWA). (2014a). G430-14 Security Practices for Operation and
Management. AWWA.
arcadis.com 151
ASSESSMENT OF CURRENT PRACTICES AND GAP ANALYSIS TECHNICAL MEMORANDUM
American Water Works Association (AWWA). (2014b). M22 Sizing Water Service Lines and Meters (3rd
ed.). Denver, CO: AWWA.
American Water Works Association (AWWA). (2014c). Policy Statement on Metering & Accountability.
Retrieved from: http://www.awwa.org/about-us/policy-statements/policy-statement/articleid/206/metering-
and-accountability.aspx
American Water Works Association (AWWA). (2015a). C700-15 Cold-Water Meters—Displacement Type,
Metal Alloy Main Case. AWWA.
American Water Works Association (AWWA). (2015b). C701-15 Cold-Water Meters–Turbine Type, for
Customer Service. AWWA.
American Water Works Association (AWWA). (2015c). C702-15 Cold Water Meters—Compound Type.
AWWA.
American Water Works Association (AWWA). (2015d). C703-15 Cold-Water Meters—Fire-Service Type.
AWWA.
American Water Works Association (AWWA). (2015e). C704-15 Propeller-Type Meters for Waterworks
Applications. AWWA.
American Water Works Association (AWWA). (2015f). C710-15 Cold-Water Meters—Displacement Type,
Plastic Main Case. AWWA.
American Water Works Association (AWWA). (2015g). C712-15 Cold-Water Meters, Singlejet Type.
AWWA.
American Water Works Association (AWWA). (2015i). C750-16 Transit-Time Flowmeters in Full Closed
Conduits. AWWA.
American Water Works Association (AWWA). (2015j). G200-15 Distribution Systems Operation and
Management. AWWA.
American Water Works Association (AWWA). (2015k). M14 Backflow Prevention and Cross-Connection
Control: Recommended Practices (4th ed.). Denver, CO: AWWA.
American Water Works Association (AWWA). (2016b). M36 Water Audits and Loss Control Programs (4th
ed.). Denver, CO: AWWA.
arcadis.com 152
ASSESSMENT OF CURRENT PRACTICES AND GAP ANALYSIS TECHNICAL MEMORANDUM
American Water Works Association (AWWA). (In Publication). M68 Water Quality in Distribution Systems.
American Water Works Association (AWWA) & Economic and Engineering Services, Inc. (2002, August
15). Finished Water Storage Facilities.
Brandt, M., Clement, J., Powell, J., Casey, R., Holt, D., Harris, N., & Ta, C. T. (2005). Managing
Distribution Retention Time to Improve Water Quality. Denver, CO: American Water Works Association
Research Foundation.
Del Toral, M. A., Porter, A., & Schock, M. R. (2013). Detection and Evaluation of Elevated Lead Release
from Service Lines: A Field Study. Environmental Science & Technology, 47(16), 9300-9307.
Federal Emergency Management Agency (FEMA). (2010). Developing, and Maintaining Emergency
Operations Plans, Comprehensive Preparedness Guide 101, Version 2.0.
Flint Advisory Task Force. (2016, March). Flint Water Advisory Task Force Final Report.
Friedman, M., Kirmeyer, G., Pierson, G., Harrison, S., Martel, K., Sandvig, A.& Hanson, A. (2005).
Development of Distribution System Water Quality Optimization Plans. Denver, CO: Water Research
Foundation.
Friedman, M. J., Hill, A. S., Reiber, S. H., Valentine, R. L., Larsen, G., Young, A., … & Peng, C. Y.
(2010a). Assessment of Inorganics Accumulation in Drinking Water System Scales and Sediments.
Denver, CO: Water Research Foundation.
Friedman, M., Kirmeyer, G., Lemieux, J., LeChevallier, M., Seidl, S., & Routt, J. (2010b). Criteria for
Optimized Distribution Systems. Denver, CO: Water Research Foundation.
GLUMRB (Great Lakes – Upper Mississippi River Board of State and Provincial Public Health and
Environmental Managers). (2012). Recommended Standards for Waterworks. Albany, NY: Health
Research Inc.
HDR Engineering, Inc., Environmics, Inc., & United States Environmental Protection Agency (USEPA).
(2006). Inorganic Contaminant Accumulation in Potable Water Distribution Systems.
International Society of Automation (ISA). (2015). ISA99 Security Guidelines and User Resources for
Industrial Automation and Control Systems (4th ed.).
Kirmeyer, G. J., Friedman, M., Martel, K., & Sandvig, A. (2002). Guidance Manual for Monitoring
Distribution System Water Quality. Denver, CO: Water Research Foundation.
Kirmeyer, G. J., Thomure, T. M., Rahman, R., Marie, J. L., LeChevallier, M. W., Yang, J., … & Schneider,
O. (2014). Effective Microbial Control Strategies for Main Breaks and Depressurization. Denver, CO:
Water Research Foundation.
arcadis.com 153
ASSESSMENT OF CURRENT PRACTICES AND GAP ANALYSIS TECHNICAL MEMORANDUM
Michigan American Water Works Association (MI-AWWA). (2008). Guidelines for Generally Accepted
Water Management Practices for the Public Water Supply. Lansing, MI: MI-AWWA.
National Institute of Standards and Technology (NIST). (2015). Special Publication 800-82 Guide to
Industrial Control System Security.
Rossman, L. A., Clark, R. M., & Grayman, W. M. (1994). Modeling Chlorine Residuals in Drinking-Water
Distribution Systems. Journal of Environmental Engineering, 120(4), 803-820.
Rowe Professional Services Company (Rowe). (2016a). Drinking Water Revolving Fund – Draft Project
Plan.
Sleeping Giant Environmental Company (Sleeping Giant). (2016). Flint Water Treatment Plant Evaluation.
Smith, C. (Ed.). (2005). Water Distribution System Assessment Workbook. Denver, CO: AWWA.
Sturm, R., Gasner, K., Wilson, T., Preston, S., & Dickinson, M. A. (2014). Real Loss Component Analysis:
A Tool for Economic Water Loss Control. Denver, CO: Water Research Foundation.
United States Department of Homeland Security (USDHS). (2008). National Incident Management
System.
United States Department of Homeland Security (USDHS). (2013). Homeland Security Exercise and
Evaluation Program.
United States Environmental Protection Agency (USEPA). (2001). Potential Contamination Due to Cross-
Connections and Backflow and the Associated Health Risks. Washington, DC: USEPA.
United States Environmental Protection Agency (USEPA). (2006). Stage 2 Disinfectants and Disinfection
Byproducts Rule (DBPR). Online. http://water.epa.gov/lawsregs/rulesregs/sdwa/stage2/index.cfm.
Vasconcelos, J. J., Boulos, P. F., Grayman, W. M., Kiene, L., Wable, O., Biswas, P., … & Goodrich, J. A.
(1996). Characterization and Modeling of Chlorine Decay in Distribution Systems. Denver, CO: American
Water Works Association Research Foundation & AWWA.
Walski, T. M., Chase, D. V., Savic, D. A., Grayman, W., Beckwith, S., & Koelle, E. (2003). Advanced
Water Distribution Modeling and Management. Civil and Environmental Engineering and Engineering
Mechanics Faculty Publications. Paper 18.
arcadis.com 154
APPENDIX A
Chlorine Residual Management Program
CITY OF FLINT
CHLORINE RESIDUAL MANAGEMENT PROGRAM - OVERVIEW
BACKGROUND
Maintenance of an adequate disinfectant residual throughout the distribution system is paramount to
protecting public health. In addition to providing microbial control, disinfectant residuals provide oxidizing
conditions to help stabilize pipe scales, and can serve as an indicator of distribution system integrity.
Therefore, a key aspect of distribution system water quality management and optimization is to identify
appropriate disinfectant residual level(s) and strategies for monitoring and maintaining them on an on-
going basis.
OBJECTIVE
The objective of this document is describe the process that will be used to develop a Chlorine Residual
Management Program for the City of Flint (Flint). The AWWA Partnership for Safe Water (Partnership)
uses disinfectant residual as one of three key indicators of distribution system optimization. Numeric
targets that follow Partnership and Criteria for Optimized Distribution Systems (Water Research
Foundation and AWWA, 2010) principles will be developed as an outcome of this process. It is clearly
understood that chlorine residual maintenance is complex process with many influencing variables, and
therefore an adaptive management process is needed that is responsive to changing conditions over time,
customer acceptance, and other competing issues.
APPROACH
The Arcadis Project Team will be working with representatives from Flint, USEPA, and MDEQ to establish
near-term and long-term chlorine residual goals within the distribution system. Near-term goals will be
implemented until data analysis indicates that the distribution system has recovered from the
destabilization event. After demonstration of distribution stabilization through data analysis, water
quality modeling, and additional surveillance monitoring (if needed), long-term goals will be identified
and implemented. Additional implementation issues such as disinfection byproduct formation, customer
acceptance, and chemical costs will also be considered. The anticipated steps for developing the Chlorine
Residual Management Plan are outlined below.
1) Look at spatial and seasonal trends, before, during after destabilization event from the following
locations:
a. Flint TCR Sites – Chlorine residual, measured at 10 Locations, 100 samples per month.
Also HPC results measured at 10 locations on a weekly basis
b. Flint DBP Sites – DBPs measured at 15 Locations, sampled quarterly
c. Flint WQP Sites – pH, temperature, conductivity, alkalinity, calcium, hardness, turbidity,
iron, chloride, and phosphate measured at 10 Locations, sampled weekly since
November 2015
d. USEPA Sites – Chlorine residual measured at 24 Locations, sampled weekly/biweekly
2) Develop summary statistics by location, season, and operational conditions
1
City of Flint – Chlorine Residual Management Program Overview
Draft 11-8-16
2. Evaluate Factors Impacting Residual Maintenance
1) Near-Term Goal
a. Numeric goal – under development; however, until a goal and the appropriate dose to
achieve this goal has been identified, it is recommended that Flint maintain current
chlorine levels leaving the WTP and storage facilities where boosting is practiced.
b. Number of locations – 34 through December 2016. Number of locations beginning Q1
2017 to be determined based on RTCR sample site selection and the need for additional
surveillance sites.
c. Percent meeting – under development
2) Long-Term – To Be Determined
a. Numeric goal
b. Number of locations
c. Percent meeting
3) Constraints for On-going Review and Consideration
a. Disinfection by-products
b. Customer acceptance
c. Chemical costs
d. Other
2
City of Flint – Chlorine Residual Management Program Overview
Draft 11-8-16
4. Develop Monitoring Plan to Demonstrate On-Going Status
3
City of Flint – Chlorine Residual Management Program Overview
Draft 11-8-16
APPENDIX B
Corrosion Control Figures
APPENDIX B: ASSESSMENT OF CURRENT PRACTICES
CORROSION CONTROL FIGURES AND GAP ANALYSIS TECHNICAL MEMRANDUM
90%
80%
70%
60%
50%
40%
30%
20%
10%
0%
0.1 1 10 100 1000 10000
Pb (µg/L)
Figure B-1: Distribution of maximum lead values at each location, by source data set, using
EPA 1st sample lead values
B1
APPENDIX B: ASSESSMENT OF CURRENT PRACTICES
CORROSION CONTROL FIGURES AND GAP ANALYSIS TECHNICAL MEMRANDUM
90%
80%
70%
60%
50%
40%
30%
20%
10%
0%
0.1 1 10 100 1000 10000
Pb (µg/L)
Figure B-2: Distribution of maximum lead values at each location, by source data set, using
EPA peak lead values
B2
APPENDIX B: ASSESSMENT OF CURRENT PRACTICES
CORROSION CONTROL FIGURES AND GAP ANALYSIS TECHNICAL MEMRANDUM
90%
80%
70%
60%
50%
40%
30%
20%
10%
0%
0.1 1 10 100 1000 10000
Pb (µg/L)
Figure B-3: Seasonal distribution of maximum lead levels at each location, using EPA 1st
sample lead values
B3
APPENDIX B: ASSESSMENT OF CURRENT PRACTICES
CORROSION CONTROL FIGURES AND GAP ANALYSIS TECHNICAL MEMRANDUM
90%
80%
70%
% of Locations
60%
50%
40%
30%
20%
10%
0%
Pb < 15 µg/L 15 µg/L ≤ Pb ≤ 50 µg/L Pb > 50 µg/L
Figure B-4: Binned distribution of locational maximum lead values, by season, using EPA 1st
sample lead values
B4
APPENDIX B: ASSESSMENT OF CURRENT PRACTICES
CORROSION CONTROL FIGURES AND GAP ANALYSIS TECHNICAL MEMRANDUM
90%
80%
70%
60%
50%
40%
30%
20%
10%
0%
0.1 1 10 100 1000 10000
Pb (µg/L)
Figure B-5: Distribution of maximum lead values at each location with paired data in all four
seasons, using EPA 1st sample lead values
B5
APPENDIX B: ASSESSMENT OF CURRENT PRACTICES
CORROSION CONTROL FIGURES AND GAP ANALYSIS TECHNICAL MEMRANDUM
90%
80%
70%
% of Locations
60%
50%
40%
30%
20%
10%
0%
Pb < 15 µg/L 15 µg/L ≤ Pb ≤ 50 µg/L Pb > 50 µg/L
Figure B-6: Binned distribution of maximum lead values at locations with paired data in all four
season, by season, using EPA 1st sample lead values
B6
APPENDIX B: ASSESSMENT OF CURRENT PRACTICES
CORROSION CONTROL FIGURES AND GAP ANALYSIS TECHNICAL MEMRANDUM
90%
80%
70%
60%
50%
40%
30%
20%
10%
0%
0.1 1 10 100 1000 10000
Pb (µg/L)
Figure B-7: Distribution of maximum lead at each location sampled in 2016, using EPA 1st
sample lead values
B7
APPENDIX B: ASSESSMENT OF CURRENT PRACTICES
CORROSION CONTROL FIGURES AND GAP ANALYSIS TECHNICAL MEMRANDUM
90%
80%
70%
60%
50%
40%
30%
20%
10%
0%
0.1 1 10 100 1000 10000
Pb (µg/L)
Figure B-8: Distribution of maximum lead values at each location with paired data before/after
May 15, 2016, using EPA 1st Sample Lead Values
B8
APPENDIX B: ASSESSMENT OF CURRENT PRACTICES
CORROSION CONTROL FIGURES AND GAP ANALYSIS TECHNICAL MEMRANDUM
100%
90%
80% After > Before
Lead increased after
70% Flushing
60%
Before >
50%
After 40%
Lead
decreased 30%
after Flushing
20%
10%
0%
-200% -100% 0% 100% 200% 300%
Change in Pb Level after May Flush Event
Figure B-9: Distribution of percent change in lead levels after the May flush event at each
location with paired data sampled both before and after May 15, 2016, using EPA
1st sample lead values
B9
APPENDIX B: ASSESSMENT OF CURRENT PRACTICES
CORROSION CONTROL FIGURES AND GAP ANALYSIS TECHNICAL MEMRANDUM
100%
90%
80%
70%
60%
50%
40%
30%
20%
10%
0%
0.1 1 10 100 1000 10000
Pb (µg/L)
Figure B-10: Spatial distribution of maximum lead levels at each location by zip code, using
EPA 1st sample lead values
B10
APPENDIX B: ASSESSMENT OF CURRENT PRACTICES
CORROSION CONTROL FIGURES AND GAP ANALYSIS TECHNICAL MEMRANDUM
Figure B-11: 90th percentile of maximum lead values at locations sampled in each zip code in
2016, using EPA 1st sample values
B11
APPENDIX B: ASSESSMENT OF CURRENT PRACTICES
CORROSION CONTROL FIGURES AND GAP ANALYSIS TECHNICAL MEMRANDUM
90%
80%
70%
60%
50%
40%
30%
20%
10%
0%
0 1 10 100 1,000 10,000
Pb (µg/L)
Figure B-12: Distribution of locational maximum lead values or homes served by copper, lead,
or galvanized service lines sampled in 2016, using EPA peak lead values
B12
APPENDIX B: ASSESSMENT OF CURRENT PRACTICES
CORROSION CONTROL FIGURES AND GAP ANALYSIS TECHNICAL MEMRANDUM
90%
80%
70%
60%
50%
40%
30%
20%
10%
0%
0 1 10 100 1,000 10,000
Pb (µg/L)
Figure B-13: Distribution of locational maximum lead values or homes served by copper, lead,
or galvanized service lines sampled in Winter 2016, using EPA peak lead values
B13
APPENDIX B: ASSESSMENT OF CURRENT PRACTICES
CORROSION CONTROL FIGURES AND GAP ANALYSIS TECHNICAL MEMRANDUM
90%
80%
70%
60%
50%
40%
30%
20%
10%
0%
0 1 10 100 1,000 10,000
Pb (µg/L)
Figure B-14: Distribution of locational maximum lead values or homes served by copper, lead,
or galvanized service lines sampled in Spring 2016, using EPA peak lead values
B14
APPENDIX B: ASSESSMENT OF CURRENT PRACTICES
CORROSION CONTROL FIGURES AND GAP ANALYSIS TECHNICAL MEMRANDUM
90%
80%
70%
60%
50%
40%
30%
20%
10%
0%
0 1 10 100 1,000 10,000
Pb (µg/L)
Figure B-15: Distribution of locational maximum lead values or homes served by copper, lead,
or galvanized service lines sampled in Summer 2016, using EPA peak lead values
B15
APPENDIX B: ASSESSMENT OF CURRENT PRACTICES
CORROSION CONTROL FIGURES AND GAP ANALYSIS TECHNICAL MEMRANDUM
90%
80%
70%
60%
50%
40%
30%
20%
10%
0%
0 1 10 100 1,000
Pb (µg/L)
Figure B-16: Distribution of locational maximum lead values or homes served by copper, lead,
or galvanized service lines sampled in Fall 2016, using EPA peak lead values
B16
APPENDIX B: ASSESSMENT OF CURRENT PRACTICES
CORROSION CONTROL FIGURES AND GAP ANALYSIS TECHNICAL MEMRANDUM
90%
80%
70%
60%
50%
40%
30%
20%
10%
0%
1 10 100 1000
Pb (µg/L)
Figure B-17: Distribution of paired locational maximum lead values or homes served by lead or
galvanized service lines sampled in all four seasons of 2016, using EPA peak lead
values
B17
APPENDIX B: ASSESSMENT OF CURRENT PRACTICES
CORROSION CONTROL FIGURES AND GAP ANALYSIS TECHNICAL MEMRANDUM
45%
39.1%
Percent of All Profile Peaks
40%
35%
30%
25%
20% 17.2%
15%
10.4%
10% 7.2%
4.7% 4.3% 5.0%
5% 2.9%
2.2% 2.2% 1.4% 1.1% 2.5%
0%
1st L 2nd L 3rd L 4th L 5th L 6th L 7th L 8th L 9th L 10th 11th 12th after
L L L 12th
L
Profile Volume
Figure B-18: Distribution of sample volumes where peak lead levels were observed during EPA
Sequential sampling
B18
APPENDIX B: ASSESSMENT OF CURRENT PRACTICES
CORROSION CONTROL FIGURES AND GAP ANALYSIS TECHNICAL MEMRANDUM
30%
25%
20%
15%
10%
5%
0%
B19
APPENDIX B: ASSESSMENT OF CURRENT PRACTICES
CORROSION CONTROL FIGURES AND GAP ANALYSIS TECHNICAL MEMRANDUM
30%
25%
20%
15%
10%
5%
0%
B20
APPENDIX B: ASSESSMENT OF CURRENT PRACTICES
CORROSION CONTROL FIGURES AND GAP ANALYSIS TECHNICAL MEMRANDUM
100%
90%
80%
70%
60%
50%
40%
30%
20%
10%
0%
1 10 100 1000 10000
Lead (µg/L)
Figure B-21: Comparison of the distributions of 1st sample lead values vs. peak lead values
observed during EPA Sequential sampling
B21
APPENDIX B: ASSESSMENT OF CURRENT PRACTICES
CORROSION CONTROL FIGURES AND GAP ANALYSIS TECHNICAL MEMRANDUM
90%
80%
70%
60%
50%
40%
30%
20%
10%
0%
1 10 100 1000 10000
Peak Lead in a Given Profile (µg/L)
Figure B-22: Distribution of peak lead observed during EPA Sequential sampling, by Sequential
sampling round
B22
APPENDIX B: ASSESSMENT OF CURRENT PRACTICES
CORROSION CONTROL FIGURES AND GAP ANALYSIS TECHNICAL MEMRANDUM
Figure B-23: Percentage of locations in the EPA Sequential dataset with “high lead level” peaks,
compared to datasets from other locations
B23
APPENDIX B: ASSESSMENT OF CURRENT PRACTICES
CORROSION CONTROL FIGURES AND GAP ANALYSIS TECHNICAL MEMRANDUM
Figure B-24: Distribution of locations in the EPA Sequential dataset where the peak lead levels
occurred inside or outside of the premise plumbing
B24
APPENDIX B: ASSESSMENT OF CURRENT PRACTICES
CORROSION CONTROL FIGURES AND GAP ANALYSIS TECHNICAL MEMRANDUM
B25
APPENDIX B: ASSESSMENT OF CURRENT PRACTICES
CORROSION CONTROL FIGURES AND GAP ANALYSIS TECHNICAL MEMRANDUM
76
Total Alkalinity - D6
74
72
70
68
66
1/3/2016
1/31/2016
2/28/2016
3/27/2016
4/24/2016
5/22/2016
6/19/2016
7/17/2016
8/14/2016
9/11/2016
10/9/2016
11/6/2016
12/4/2016
Figure B-26: Example control chart of total alkalinity from WQP monitoring site D6
B26
APPENDIX B: ASSESSMENT OF CURRENT PRACTICES
CORROSION CONTROL FIGURES AND GAP ANALYSIS TECHNICAL MEMRANDUM
2016 Total Alkalinity
D1/2 D3/D4 D5 D6 D7/D8 D9/D13 D10 D11 R1 R2 S1
76
75
74
Total Alkalinity (mg/L as CaCO3)
73
72
71
70
69
68
67
66
12/12/2015 1/31/2016 3/21/2016 5/10/2016 6/29/2016 8/18/2016 10/7/2016 11/26/2016 1/15/2017
Date
Figure B-27: Control chart group medians for total alkalinity at the Flint WQP monitoring sites in
2016
B27
APPENDIX B: ASSESSMENT OF CURRENT PRACTICES
CORROSION CONTROL FIGURES AND GAP ANALYSIS TECHNICAL MEMRANDUM
2016 pH
D1/2 D3/D4 D5 D6 D7/D8 D9/D13 D10 D11 R1 R2 S1
7.8
7.7
7.6
7.5
pH
7.4
7.3
7.2
7.1
7.0
12/12/2015 1/31/2016 3/21/2016 5/10/2016 6/29/2016 8/18/2016 10/7/2016 11/26/2016 1/15/2017
Date
Figure B-28: Control chart group medians for pH at the Flint WQP monitoring sites in 2016
B28
APPENDIX B: ASSESSMENT OF CURRENT PRACTICES
CORROSION CONTROL FIGURES AND GAP ANALYSIS TECHNICAL MEMRANDUM
2016 Orthophophate
D1/2 D3/D4 D5 D6 D7/D8 D9/D13 D10 D11 R1 R2 S1
4.0
3.5
Measured Orthophosphate (mg/L PO4)
3.0
2.5
2.0
1.5
1.0
12/12/2015 1/31/2016 3/21/2016 5/10/2016 6/29/2016 8/18/2016 10/7/2016 11/26/2016 1/15/2017
Date
Figure B-29: Control chart group medians for orthophosphate at the Flint WQP monitoring sites
in 2016
B29
APPENDIX C
Water Distribution Model Audit
WATER DISTRIBUTION MODEL AUDIT
City of Flint, Michigan
Model received from USEPA/Citilogics. Model represents actual conditions during a specific
one month date range. Minimal data utilized for latest calibration. Excessive use of controls to
replicate actual conditions rather than typical operations.
SIMULATION SETTINGS
☒ Hydraulics ☒ Extended Period ☐ Steady State
☐ Water Quality ☒ Time-Based Controls
☐ Energy ☒ Rule-Based Controls
Page:
1/2
WATER DISTRIBUTION MODEL AUDIT
City of Flint, Michigan
CONDITION ASSESSMENT
☐ Does not complete simulation
Comments
Comments
☒ Other concerns
Comments
Many of the pumps are simulated as a single head and flow point and not with
a pump curve.
- This introduces significant error in model simulation results.
The excessive detail within logical controls applied in the model force a
specific operating scheme which can be considered over-calibrated. This
impairs the model’s ability to reasonable simulate various operations and
various conditions other than the specific modeled period when adjustments
are made.
The water demand allocated with the model (and the diurnal demand curve)
are not practical and need to be corrected. A diurnal pattern must average 1.0
and in this model the diurnal pattern (time step multipliers to average demand
average) is 8,368.48.
arcadis.com Page:
2/2
APPENDIX D
Interim Final Revised Total Coliform Rule Monitoring Plan
City of Flint
Final RTCR Coliform Monitoring Plan
February 1, 2017
Table of Contents
Table of Contents
A. Introduction and Objectives ................................................................................................................. 2
B. Coliform Monitoring Requirements ...................................................................................................... 2
C. System Information............................................................................................................................... 3
D. Key Contacts.......................................................................................................................................... 4
E. Sample Locations and Schedule ............................................................................................................ 4
Repeat Sample Sites.............................................................................................................................. 7
F. Sampling and Reporting SOP ................................................................................................................ 7
Monthly Monitoring.............................................................................................................................. 7
Sample Instructions .............................................................................................................................. 7
Procedure When Sample Result Is POSITIVE ........................................................................................ 8
G. Coliform Violations................................................................................................................................ 9
E. coli MCL Violations ............................................................................................................................ 9
Other Types of Violations...................................................................................................................... 9
H. Assessments ........................................................................................................................................ 10
Level 1 Assessment ............................................................................................................................. 10
Level 2 Assessment ............................................................................................................................. 10
City of Flint
Final RTCR Monitoring Plan
February 2017 1
A. Introduction and Objectives
This Coliform Monitoring Plan (Plan) has been developed for the City of Flint (City) to meet requirements
of the Revised Total Coliform Rule (RTCR) and R 325.10704c of Michigan’s rules for public water
supplies.
Drinking water systems conduct total coliform monitoring with the objective of evaluating microbial
water quality to ensure water provided to consumers is free of disease-causing organisms. While
drinking water systems can’t evaluate the microbial quality of all water in all locations of their
distribution system in a cost-effective manner, a coliform monitoring program allows systems to
evaluate water quality in many locations throughout their system on a monthly basis.
● Update the City’s coliform monitoring plan to comply with recent requirements of the RTCR.
● Ensure representative routine coliform sampling in the City’s system by expanding the number
of routine samples collected.
● Provide Standard Operating Procedures (SOPs) for coliform sampling and regulatory reporting.
● Provide emergency response and public notification procedures in the event of a total coliform
or E. coli positive sample.
● Provide information for the completion of Level 1 and Level 2 system assessments
● Enhance water quality surveillance.
According to the 2015 American Community Survey (ACS-2015), the City serves a population estimated
to be 98,310. SDWIS records have recently been updated to reflect this population. According to
requirements in R 325.10704g, the City is required to collect at least 100 coliform samples throughout
their distribution system each month. Samples are analyzed by the City’s certified laboratory. The City
may, with DEQ approval, decrease sample collection to 90 samples per month if the population drops
below 96,000. The City will be required to increase sample collection to 120 per month if the population
served by the City increases to more than 130,000.
Coliform samples are analyzed for the presence or absence of total coliform. A satisfactory test
indicates no coliform are present. An unsatisfactory test is positive for the presence of coliform
bacteria. Further testing is conducted on unsatisfactory samples to determine if E. coli is present.
Repeat samples are required if the routine sample is unsatisfactory. Three repeat samples are required
City of Flint
Final RTCR Monitoring Plan
February 2017 2
at the following locations:
C. System Information
Flint purchases treated surface water (finished water) from the Great Lakes Water Authority to serve a
population of approximately 98,310 (per 2015 census). Finished water enters the system at the water
treatment plant and is continuously monitored for free chlorine residual. Chlorine, caustic soda (as
needed), and orthophosphate are supplemented at this location (CSII). The water is then distributed to
the City’s system via one transmission line. The City has four reservoirs for a total of 54 MG of storage
capacity. The City is comprised of a single pressure zone, although there is a small pumped pressure
district in the southwest region of the City. The water treatment plant and storage reservoirs (sample
sites 5 and 11) are shown on the attached map. Approximately 90% of the City’s distribution system is
comprised of unlined cast iron pipe.
City of Flint
Final RTCR Monitoring Plan
February 2017 3
D. Key Contacts
As of December 2016, the following are the key contacts from the City, MDEQ, County, and local news
media for RTCR monitoring and emergency response activities.
The sample sites are presented geographically on the attached RTCR Sample Site Map. Also shown are
five additional surveillance monitoring sites for which chlorine residual will be monitored on a weekly
basis for informational purposes. Thus, when sites are accessible, a total 35 chlorine measurements will
be taken from 25 unique locations each week. Table 3 lists the City’s existing and new routine coliform
sampling locations, chlorine residual surveillance monitoring locations, sampling schedule, and repeat
sample site locations (for coliform sites only). Sampling will occur on the days indicated in the third
column when possible and adjustments will be made on an as-needed basis for holidays and irregular
months.
1
*Instances, including but not limited to holidays, declining participation, and site closings, may reduce the number
of weekly samples collected.
City of Flint
Final RTCR Monitoring Plan
February 2017 5
North Flint Automotive Th(e/o)2
5018 Clio Rd Tu
9 TCR3 5005 Cloverlawn Rd 4825 Clio Rd
Rite-Aid Th(e/o)2
4090 Clio Road Tu
10 New 4117 Clio Rd 4006 Clio Rd
Auto Zone Th(e/o)2
1416 Dupont St Tu
11 TCR3 1360 Dupont St 1430 Dupont St
West Side Reservoir Th(e/o)2
2503 Garland Street Tu
12 New 605 Garland 401 Garland
Hoffman's Deco Deli & Café Th(e/o)2
3538 Richfield Rd W Quarterly
13 3246 Richfield Rd 3702 Richfield Rd
Grandma Recipes F (e/o)2 (DBP)5
Former
3802 Davison Rd W Arby’s TCR
14 3718 Davison Rd 3835 Davison Rd
Admiral F (e/o)2 site
downstream3
2132 Davison Road W
15 New 2320 Davison Rd 2100 Davison Rd
Luigi's Restaurant F (e/o)2
2838 E. Court St W
16 TCR3 905 S. Dort Hwy 2845 E. Court St
Rite-Aid F (e/o)2
3302 S. Dort Hwy W
17 TCR3 3124 S. Dort Hwy 3316 S. Dort Hwy
Liquor Palace F (e/o)2
3717 Fenton Rd W Quarterly
18 3708 Fenton Rd 3621 Fenton Rd
Rite-Aid F (e/o)2 (DBP)5
3216 MLK Blvd Th
19 TCR3 3110 MLK Ave 3317 MLK Ave
Salem Housing F(e/o)2
1525 MLK Th EPA Chlorine
20 1402 MLK 1602 Oren Ave
Fire House #3 F (e/o)2 Monitoring4
510 Leta Avenue
21 Don's Market M New Not applicable Not applicable
CHLORINE ONLY
1002 W. Home Ave Not applicable Not applicable
EPA Chlorine
22 Hasselbring senior center Tu
Monitoring4
CHLORINE ONLY
4612 Western Rd Not applicable Not applicable
EPA Chlorine
23 Sam’s Rollingwood Market W
Monitoring4
CHLORINE ONLY
3109 Kleinpell St Not applicable Not applicable
EPA Chlorine
24 Genesee Community Health Th
Monitoring4
CHLORINE ONLY
3402 Western Rd Not applicable Not applicable
EPA Chlorine
25 Fire House #5 F
Monitoring4
CHLORINE ONLY
1
When a routine sample is positive for total coliform or E.coli, collect samples from repeat sites in the distribution system. Supplies
that purchase their source water must notify their supplier of water within 24 hours of a positive routine sample result. Surface water
supplies are not required to sample their source water.
City of Flint
Final RTCR Monitoring Plan
February 2017 6
2
Samples will be collected on an every-other week basis.
3
Sample site was previously used for the Total Coliform Rule monitoring plan.
4
Sample site was previously used during EPA chlorine monitoring.
5
Sample site is currently used for quarterly expanded water quality parameter (EWQP) and disinfection byproduct monitoring (DBP).
Sample Instructions
Sample Containers – Sample containers used for microbiological examination are collected in
plastic or glass bottles that have been cleansed, carefully rinsed, and sterilized. For drinking water
samples containing chlorine, the sample bottles should contain a dechlorinating agent. The
dechlorinating is added prior to sterilization.
Sample Collection Procedure – Overview: When collecting a sample leave ample air space in the
bottle to allow for mixing before examination. When using Idexx 120mL sample bottles, fill bottle to
100mL line. Flush and disinfect sample port and use aseptic techniques to avoid sample contamination.
The sample should be representative of the water being tested. If the sample is taken from a
distribution system tap, select a tap that is water from a service pipe directly connected with the water
main.
Detailed Instructions:
City of Flint
Final RTCR Monitoring Plan
February 2017 7
4) When sampling from a mixing faucet, run hot water for 2 to 3 minutes, then cold water until
cold.
5) Flush the tap until you measure a temperature change, then record the chlorine residual level.
Measure and record the chlorine residual at the same time and place as every routine and
repeat sample collected.
6) Reduce water flow to permit filling of sample bottle without splashing, and to prevent over
filling.
7) Remove cap from bottle and hold cap with the inner surface facing downward. Do not set cap
down. Dust-like particles in the bottle are a preservative; do not empty out or rinse out this
preservative.
8) Fill the bottle to 100 mL line. Avoid contact with sample tap or other surfaces. Do not overfill or
underfill.
9) Recap the sample bottle before turning off the water.
10) Complete the laboratory chain of custody form and attach it to the sample bottle.
2) Notify the DEQ District Office within 24 hours to learn what further action is required when
greater than 5.0 percent are positive. Follow up action includes a formalized assessment of the
water supply.
3) Notify the DEQ District Office by the end of the day if any sample result is positive for E.coli.
4) For an E. coli MCL Violation, public notification is required within 24 hours. Due to the public
health risk, a boil-water advisory will typically be issued in response to an E. coli MCL violation.
An E. coli MCL violation will trigger a Level 2 Assessment, which is to be completed by DEQ.
City of Flint
Final RTCR Monitoring Plan
February 2017 8
City of Flint E. coli Response Plan
G. Coliform Violations
Under the RTCR, the non-acute MCL violation (when greater than 5% of monthly samples are positive
for total coliform) has been removed. Instead, a Level 1 Assessment must be performed. An MCL
violation is issued when the E. coli MCL is not met. In this case a Level 2 Assessment must be performed.
Information on Assessments is provided below.
H. Assessments
A key component of the RTCR is to require a system assessment when triggered by results from coliform
monitoring. The two types of assessments are described below.
Level 1 Assessment
A Level 1 Assessment is triggered when more than 5% of monthly samples are TC+ or when the system
fails to collect any required repeat samples following a TC+. The assessment is an evaluation intended to
identify possible presence of sanitary defects, defects in DS coliform monitoring practices, and (when
possible) the likely reason that the supply triggered the assessment. The Level 1 Assessment is
conducted by the system’s owner and/or operator. Information gathered in the assessment is described
on the MDEQ Level 1 Assessment Form for Community Water Supplies (included in Appendix A) and
submitted to MDEQ for review within 30 days of learning that the assessment was triggered. The form
must identify any sanitary defects found during the investigation and provide a list of corrective actions
that were completed during the investigation and/or a proposed time table for completing corrective
actions.
Level 2 Assessment
A Level 2 Assessment is triggered by an E. Coli MCL violation or when two Level 1 Assessments are
triggered within a rolling 12-month period. A Level 2 Assessment is a more detailed evaluation than a
Level 1 Assessment and is conducted by the State. MDEQ will send a team of staff members to Flint to
complete the assessment. The team will conduct the assessment as soon as practicable and generally
within one week of being notified of the violation.
City of Flint
Final RTCR Monitoring Plan
February 2017 10
Table 4. Comparison of Level 1 and Level 2 Assessment Requirements
City of Flint
Final RTCR Monitoring Plan
February 2017 11
City of Flint
RTCR Sampling Sites
13 Grandma's Recipes
10 Auto Zone
1 University Market
12 Hoffman's Deco Deli & Café
16 Rite-Aid Pharmacy
4 Rite-Aid Pharmacy
18 Rite-Aid Pharmacy
21 Don's Market
Legend
" Water Treatment Plant
Sampling Sites Esri, HERE, DeLorme, MapmyIndia, © OpenStreetMap
RTCR Sites contributors, and the GIS user community
This form must be completed and submitted to the appropriate DEQ District Office as soon as possible, but no later than
30 days after the supply triggered the assessment. It should be completed by the Operator In Charge, Water Supply
Owner, or a knowledgeable representative of the water system.
1. General Information
CWS Name: WSSN:
Trigger Event: Greater Than 5% Total Coliform Positives or Failure to Collect All Repeat Samples
2. Bacteriological Sample Summary (Include all results associated with monitoring period, add additional pages if necessary)
Purpose (Routine,
Repeat, Result (ND, TC+,
Date & Time Location Triggered, EC+, invalid, Collected By Laboratory
Construction, interference)
Repair)
3. Assessment Questions: Answer each question in Subsections A - G either Yes, No or Not Applicable (NA). Review and evaluate each
question for potential causes of contamination. If the answer to any of these questions is unknown, leave blank and indicate on a separate sheet
what actions will be taken to determine the necessary information.
Yes No NA
Were the samples collected in accordance with the Sample Site Plan?
Was the location and condition of the sample tap sanitary?
Were proper sample collection procedures followed?
Were the samples submitted to the lab in a timely & acceptable manner?
B. Source – Wells (if wells are not used check here and go to subsection C) Answer
Yes No NA
Do the wells have a proper well cap, sanitary seal and vent screens?
Have the wells/pumps undergone any recent repairs or maintenance activities?
Is the exposed portion of the casing (including electrical conduit) in good condition?
Is the area near the well cap/casing free of insects, bugs, brush and vegetation?
Is there standing water or other unsanitary conditions near the wells?
Any signs of vandalism to wells or forced entry into well houses?
E. Storage (if no water storage tank check here and go to subsection F) Answer
Yes No NA
Are there any holes, leaks or other structural problems?
Are access hatches and manhole openings tightly covered and secured?
Are all vents and overflow pipes screened?
For hydropneumatic tanks, is the tank waterlogged?
Any signs of vandalism or unauthorized access to storage facilities?
Have the tank(s) been recently drained, cleaned or inspected?
5. Corrective Actions Taken or to be Taken for any Issues Identified in Part 3: Use this space to describe corrective actions already taken and
date(s) completed; or a proposed timetable for corrective actions not yet completed. Attach additional page(s) if needed.
6. Certification: I hereby certify that the information contained herein is true, accurate and complete to the best of my knowledge and information.
-----------------------------------------------------------------------------------------------------------------------------------------
This form must be completed as soon as possible, but no later than 30 days after the supply triggered the assessment. It must
be completed by DEQ - Office of Drinking Water & Municipal Assistance staff.
1. General Information
2. Bacteriological Sample Summary (Include all results associated with monitoring period, add additional pages if necessary)
Purpose (Routine,
Result (ND, TC+, EC+,
Date & Time Location Repeat, Triggered, Collected By Laboratory
invalid, interference)
Construction, Repair)
3. Assessment Questions: Answer each question in Subsections A - H either Yes, No or Not Applicable (NA). Review and evaluate each
question for potential causes of contamination. If the answer to any of these questions is unknown, leave blank and indicate on a separate
sheet what actions will be taken to determine the necessary information, including any supplemental information that needs to be provided
by the water supply.
Answer
A. Sample Site Selection and Sample Collection
Yes No NA
Were the samples collected in accordance with the Sample Site Plan?
For positive samples, were the taps used in appropriate condition for collection?
For positive samples, were the taps used on a regular basis?
Did someone other than a regular sample collector collect the samples?
Were proper sample collection procedures followed? (tap flushed, aerator removed, cap properly handled,
clean and sealed sample bottles used, bottles not rinsed, etc.)
Were the samples kept cool and delivered to the lab within 30 hours of collection?
Have there been any recent plumbing changes or construction at the site?
Any identified cross connections near the sample tap or premise plumbing?
Is there any Point of Entry (POE) treatment units after the service line connection or in the premise?
Is there any Point of Use (POU) treatment units at the sample tap(s) location?
Answer
B. Source – Wells (if wells are not used check here and go to subsection C)
Yes No NA
Do the wells have approved and secured well caps or sanitary seals?
Are the well caps or sanitary seals vented and screened?
Is the top of the well head at least 12-inches above grade?
Answer
C. Source – Surface Water (if surface water is not used check here and go to subsection D)
Yes No NA
Is the intake screened and in good condition?
Any signs of vandalism or unauthorized access to source facilities?
Does the raw water quality data indicate changes to the source water?
Are there any obvious sources of contamination in the source?
Have there been any sewer or chemicals spills, or other disturbances in the area of the source?
Any signs of Algal blooms?
Was there any heavy precipitation, rapid snowmelt or flooding recently?
Any signs of drought or low water levels in the source?
Has source water turnover occurred?
D. Well House or other Low or High Service Pump House (if there are no well/pump houses, Answer
check here and go to subsection E) Yes No NA
Are there unsanitary conditions?
Any openings where animals may enter?
Are there signs of animal activity?
Are air/vacuum relief valves properly screened and air gapped?
Are any vents/reliefs associated with control valves air gapped and not subject to flooding?
Any cross-connections (piping in drains, chemical feed, irrigation, fire suppression)?
Is the pump-to-waste piping capped and air gapped?
Is the well/pump house subject to flooding?
Is the well/pump house used for any other purposes such as storage or maintenance activities?
Is there evidence of unauthorized entry?
Answer
E. Treatment (if no treatment check here and go to subsection F)
Yes No NA
Have there been additions or modifications to any treatment process?
Have there been interruptions in any chemical feed, treatment unit or process?
Have there been any recent maintenance or repair of treatment equipment?
Are all treatment devices and processes operational and properly maintained?
Any signs of vandalism or unauthorized access to treatment equipment or facilities?
Are there any signs that the chemicals being fed have been contaminated (discoloration, unusual odors,
suspended particles, etc.)?
If chlorine is used, was there a detectable residual at the sample sites where the positive samples
occurred?
If chlorine is used, is a residual currently being detected at the plant tap and within the distribution
system?
Were there any instances where C*T was not properly maintained?
Does water quality data indicate inadequate or inappropriate treatment of water?
Answer
F. Storage (if no water storage tank check here and go to subsection G)
Yes No NA
Are there any holes, leaks, cracks or other structural problems that could be a source of contamination?
Are access hatches and manhole openings tightly covered, and secured?
Do the access hatches/openings have a tightly fitted, rim overlapped cover and non-deteriorated gasket?
Are all vents and overflow pipes properly screened?
Are the vents turned downward with an adequate air gap at the termination point?
Do the overflow pipes have at least a 12-inch air gap at the outlet?
Do overflow pipes and downspouts drain water away from the structure?
For hydropneumatic storage, is the tank maintaining adequate minimum pressure?
For hydropneumatic storage, is the tank waterlogged?
Are the storage facilities secured to prevent unauthorized access?
Any signs of vandalism or unauthorized access visible?
Is there evidence of bird activity on the storage tank roof (nests, droppings, feathers, etc.)?
Is any portion of the storage facilities buried or installed below grade?
Has there been any tank maintenance or recent work?
Any recent inspections indicating sanitary deficiencies or recommended repairs?
If the tank has been inspected or removed from service recently, was it properly disinfected and sampled?
If chlorine is used, is there a detectable residual in or leaving the tank(s)?
Answer
G. Distribution System
Yes No NA
Is there any evidence the system experienced low (< 20 psi) or negative pressure?
Have there been any water main breaks, repairs, or new main installations?
Have there been any firefighting, system flushing or other high demand events recently?
Have there been any distribution system booster pump issues, repairs or new installations?
Have there been other construction activities like hydrant or valve replacement that could have introduced
contamination into the system?
Are there hydrants or blow-offs with unplugged weep/drain holes located in areas of high water table or
poorly draining soils?
Are there any dead-ends that are not flushed on a regular basis?
Are there any air relief valves located in vaults where the vent terminates below grade or are not properly
air gapped above grade?
Is the supply actively performing cross connection control inspections, including regular testing of all
testable backflow preventers including those at residential accounts?
Is there any evidence of intentional contamination in the distribution system?
Are there any control or altitude valves subject to flooding?
Answer
H. Operation and Maintenance (O & M)
Yes No NA
Any changes in procedures or staff effecting O & M activities?
Is maintenance of all facilities and equipment being performed per appropriate schedule?
Any recent interruptions to electrical power?
Have there been any automation/control system interruptions recently?
DEQ Environmental Assistance Center www.michigan.gov/deq
Telephone: 1-800-662-9278 Page 3 of 4 EQP 5827 (Rev. 3/2016)
Answer
H. Operation and Maintenance (O & M)
Yes No NA
Any complaints from customers related to water quality or low pressure?
Have there been any illnesses reported or suspected of being waterborne?
Any other issues that could have contributed to bacteriological contamination?
4. Issue Description: For any answer in Part 3 that is in a shaded box, use this space to describe the event and provide additional
information on potential causes of contamination identified during the assessment. Include corresponding dates with your findings. Attach
additional pages if needed. Include dates of sample collection, water main breaks, maintenance activities, etc. with your findings.
5. Corrective Actions Taken or to be Taken for any Issues Identified in Part 3: Use this space to describe corrective actions already
taken and date completed; and/or a proposed timetable for corrective actions not yet completed. Attach additional pages if needed.
6. Certification: I hereby certify that the information contained herein is true, accurate and complete to the best of my knowledge and
information. Must be DEQ - ODWMA staff.
-----------------------------------------------------------------------------------------------------------------------------------------
7. DEQ District Supervisor Review: This section is to be completed by DEQ District Supervisor.
Supervisors Name: Date Reviewed:
Date Received: Within 30 days of trigger: Yes No
Likely Reason for Positive Samples Identified:
Assessment Complete: Yes No
Yes No
Corrective Actions Completed: Proposed Schedule Acceptable:
Yes No NA Yes No NA
Assessment Level Reset: Yes No
Comments:
www.arcadis.com
ADDENDUM 1
Update to Disinfectant Residual and Disinfection Byproducts
MEMO
Date:
April 6, 2018
Subject:
This memorandum appends the Disinfectant Residual and Disinfection Byproducts analysis (Section 4.1)
presented in the Assessment of Current Practices and Gap Analysis Technical Memorandum, March 2017
prepared for the City of Flint (Flint) as part of the Flint Drinking Water Distribution System Optimization
Project. The data analysis presented in the original memo has been updated with data from the latest
sampling events. Additionally, as part of this updated analysis, a Chlorine Residual Management (CRM)
Tool was developed to support the Flint in managing and tracking chlorine residual and disinfection
byproducts (DBP) data. A description of the tool is provided herein.
Daily chlorine residual data collected at Flint’s distribution system point of entry (POE) from
January – September 2017.
arcadis.com Page:
1/6
MEMO
Daily chlorine residual data collected at 25 distribution system locations from January –
September 20171.
Quarterly DBP and accompanying chlorine residual data collected at eight distribution system
locations from October 2016 – June 2017.
These data were analyzed using the new CRM Tool, and results are discussed in the following section.
Objective
The CRM Tool was developed to support Flint in managing and tracking chlorine residual data collected at
the POE to Flint’s distribution system and within the distribution system. The tool provides statistical and
graphical analysis of the monthly data collected from all sites as well as for each individual site. This tool
also tracks DBP sampling and performs DBP compliance calculations. Instructions for entering data into
the CRM Tool and updating analyses are included in the “Instructions” tab of the CRM Tool spreadsheet.
An overview of the spreadsheet is provided in below. It should be noted that the CRM Tool was
transferred to Flint in April of 2017. A tutorial was provided to review data entry and interpretation of
results. Incorporation of results from the CRM Tool and control chart tools provided separately are
discussed in the SOP 441: Maintaining Distribution System Chlorine Residual.
Daily chlorine residual data collected at Flint’s distribution system POE (from Monthly Operating
Reports),
Daily chlorine residual samples collected from 25 distribution system locations (from Monthly
Operating Reports)2, and
Quarterly DBP and accompanying chlorine residual data collected at eight distribution system
locations.
Monthly statistics for all sites – 5th, 10th, 50th, 90th, and 95th percentiles are calculated and
presented in tabular and graphical formats for each month. These data are compared with the
chlorine residual goals established by the Partnership for Safe Water (≥ 0.2 mg/L in 95% of
monthly samples collected at all sites) and established by Flint (≥ 0.5 mg/L in 95% of monthly
1 Distribution sites include 20 Revised Total Coliform Rule (RTCR) monitoring and 5 “chlorine only” surveillance sites as of April
2017. Note that typically seven sites are sampled per day, and sites are rotated each day so that each site is sampled at least four
times per month.
2 Distribution sites include 20 Revised Total Coliform Rule (RTCR) monitoring and 5 “chlorine only” surveillance sites as of April
2017. The CRM Tool should be updated to reflect the approved 23 RTCR sites as of February 2018 and once the final 25 sites
have been approved by the State. Note that typically seven sites are sampled per day, and sites are rotated each day so that each
site is sampled at least four times per month.
arcadis.com Page:
2/6
MEMO
samples collected at all sites). Figure 1 below presents monthly percentiles for residuals collected
between January and September 2017 at all 26 locations. With the exception of September, Flint
has met both the Partnership and Flint internal chlorine residual goals. Slightly higher residuals
have been targeted leaving the plant during the summer to account for the additional residual loss
due to higher water temperatures.
Individual site data and statistics – The CRM Tool calculates percentiles for each monitoring site
and provides comparison with the chlorine goals described above. These analyses provide a
review of site-specific conditions to allow for localized or regional improvements where needed.
Figure 2A presents an overview of daily chlorine residuals collected at all sites and Figure 2B
presents percentiles for each site from January – September 2017. As seen from the figures, all
sites, except Site #19 (Salem Housing) have met the Partnership and Flint internal chlorine
residual goals. Lower residuals are common at Site #19 (as compared to other sites) due to low
demands. Localized strategies (i.e., installation of an autoflusher) could be explored to help to
improve residuals in this area particularly during the summer months.
Additionally, the CRM Tool calculates the locational running annual average (LRAA) for total
trihalomethanes (TTHM) and five haloacetic acids (HAA5) samples collected at each DBP site (Figures 3
and 4 below) to compare with the maximum contaminant levels (MCLs) for these parameters. As seen
from the figures, LRAAs for both TTHM and HAA5 have been well below their respective MCLs. Graphical
representations of all DBP sample results and associated chlorine residual data collected at each DBP
locations are also created (Figures 5 and 6). As expected, sites with lower residuals (i.e., higher chlorine
residual loss) correlate with higher TTHM concentrations as shown in Figure 5. HAA5 values, however,
appear to be relatively consistent regardless of chlorine residuals as shown in Figure 6.
Figure 1. Monthly chlorine residual data compared to goals from 26 sites (January - September 2017)
arcadis.com Page:
3/6
MEMO
Figure 2A. Overview of daily chlorine residual data from all sites (January - September 2017)
Figure 2B. Daily chlorine residual data – percentiles for each monitoring location from all sites (January -
September 2017)
arcadis.com Page:
4/6
MEMO
Figure 3. TTHM LRAA for all DBP sites (Q4 2016 – Q2 2017)
Figure 4. HAA5 LRAA for all DBP sites (Q4 2016 – Q2 2017)
arcadis.com Page:
5/6
MEMO
Figure 5. Comparison of TTHM to Chlorine Residual for all DBP sites (Q4 2016 – Q2 2017)
Figure 6. Comparison of HAA5 to Chlorine Residual for all DBP sites (Q4 2016 – Q2 2017)
arcadis.com Page:
6/6
ADDENDUM 2
UPDATE TO INTERNAL CORROSION CONTROL
MEMO
Subject:
This memorandum appends the Internal Corrosion Control analysis (Section 4.4) presented in the
Assessment of Current Practices and Gap Analysis Technical Memorandum, March 2017 prepared for the
City of Flint, MI as part of the Flint Drinking Water Distribution System Optimization Project. The data
analysis presented in the original memo has been updated with data from the latest sampling events.
Residential Dataset
The MDEQ Residential dataset is comprised of lead and copper data from 1-L samples collected by
residential customers who then submitted the samples to MDEQ for analysis. At the time of the original
arcadis.com Page:
1/7
MEMO
analysis, the residential dataset contained 24,251 samples across 14,193 locations that were collected
and analyzed between September 3, 2015 and September 15, 2016. Between September 15, 2016 and
December 14, 2017, six new sample locations and 966 samples were added across existing and new
sample locations. As of December 14, 2017, the residential dataset consisted of 25,217 samples
collected over 14,199 locations.
Sentinel Dataset
The Sentinel program was initiated by MDEQ on February 16, 2016 and consisted of sampling sites that
were established through coordination between MDEQ and Flint residents, and were reviewed by USEPA
prior to sampling. The original analysis considered data from 10 total rounds of sampling conducted under
the Sentinel program; this included five rounds under the original Sentinel sampling program, and five
rounds under an Extended Sentinel sampling program, covering a period between February 16, 2016 and
September 27, 2016. That data set included 3,988 lead data collected over 840 total sites. The original
Sentinel sampling rounds averaged 630 sites, while on average the Extended Sentinel sampling rounds
included around 168 sites.
Since the original analysis, five additional rounds of Extended Sentinel sampling have occurred.
Extended Sentinel Sampling Rounds 6 through 10 averaged 138 sites per round, adding 689 lead data.
The last sampling round, Extended Sentinel Sampling Round 10, was conducted in July 2017. As of July
21, 2017, the Sentinel dataset consisted of 4,677 lead data over 840 sample sites.
Sequential Dataset
The Sequential dataset is comprised of data collected by USEPA at a total of 115 residences across five
sampling rounds between January 28, 2016 and November 15, 2016. Although many of the sampling sites
were included in multiple rounds, many of the sites were not included in every round; instead, over the five
rounds of sampling a total of 279 water samples were collected from 115 sampling sites. No data were
added to the Sequential dataset following the original analysis.
Common Database
The original analysis was conducted using the MDEQ Residential, MDEQ Sentinel and USEPA Sequential
datasets. Given that each dataset used a separate population of sample collection sites (some of which
were in multiple data sets) and that each data set used different sample location codes, it was necessary
to condense the three datasets into one common database to allow for comparison between sampling
sites. This required that the datasets be correlated by location, so that sites included in two or more of the
sampling programs would be considered as one common location. In order to focus on high-risk sites
(i.e., sites known to or thought to have lead service lines or lead-containing premise plumbing), all of the
locations from the USEPA Sequential and MDEQ Sentinel datasets were included in the common
database. Any data from the Residential dataset that was collected from one of the locations contained in
the Sequential or Sentinel datasets was also included in the common database.
The original analysis was based on the common database, which at the time included a total of 8,596
sampling data across 904 locations. The updated common database added an additional 1,656 lead data
from the existing 904 sites, giving a total of 10,252 lead data across those same 904 locations.
arcadis.com Page:
2/7
MEMO
Lead Levels
The original study analyzed the available lead data in a number of ways, including:
arcadis.com Page:
3/7
MEMO
As was described during the original analysis, comparison of paired locations provides a more accurate
depiction of how lead levels are trending because it compares the same locations over time. The unpaired
datasets also tend to contain more sites with lower lead concentrations because those sites tend to be
excluded from subsequent sampling. Figures Figure 1 and Figure 2 present the lead levels at each of the
82 locations with paired data in all seven seasons (Winter 2016 through Summer 2017). As these figures
show, it appears that lead levels have improved over time and stabilized with ~93 percent of the sample
sites testing below the 15 µg/L action level for lead. However, the small portion of sites at the highest
percentiles are still seeing lead levels significantly above 15 µg/L, with the maximum lead levels in each
season at or above 100 µg/L.
90%
80%
70%
60%
50%
40%
30%
20%
10%
0%
0.1 1 10 100 1000 10000
Pb (µg/L)
Figure 1: Distribution of maximum lead values at each location with paired data by season, using USEPA 1st
sample lead values
arcadis.com Page:
4/7
MEMO
100%
94% 93%
89%
90%
80%
80%
73%
70%
60% 57%
% of Samples
50% 45%
40%
34%
30%
24% 24%
21%
18%
20%
15%
10% 6% 6% 6% 5% 5%
2%
0% 1%
0%
Pb < 15 µg/L 15 µg/L ≤ Pb ≤ 50 µg/L Pb > 50 µg/L
Figure 2: Binned distribution of maximum lead values at locations with paired data by season, using USEPA
1st sample lead values
Analysis of the weekly WQP data has continued since the original analysis. Control chart tools have been
developed for the City to assess how well parameters such as pH, orthophosphate residual, and chlorine
residual are being controlled. The control chart tools allow for comparison of these parameters between
sites, over any date range available.
The most concerned WQP-related issue noted during the original analysis was that the orthophosphate
level at Site D9/D13 (now labelled D19) was consistently lower than the rest of the sites in the distribution
system. The orthophosphate residual at WQP Site D19 dropped below the 3.1 mg/L as PO 4 minimum
target at times during 2016. Figure 3 presents the updated control chart for this site through 2017. As
Figure 3 indicates, the City successfully increased the orthophosphate residual at WQP Site D19 since the
original analysis, and has maintained a consistent residual throughout 2017.
Several other locations observed orthophosphate residuals below the 3.1 mg/L as PO 4 minimum in 2017;
however, unlike 2016, those low values were only one-week occurrences and orthophosphate residuals
were back above the requirement minimum in subsequent samples.
arcadis.com Page:
5/7
MEMO
4.0
Orthophosphate (mg/L PO4) - WQP Site 19
3.9
3.8
3.7
3.6
3.5
3.4
3.3
3.2
3.1
3.0
May-16
May-17
Feb-16
Mar-16
Feb-17
Mar-17
Apr-16
Oct-16
Apr-17
Oct-17
Jul-16
Jul-17
Jan-16
Jun-16
Jan-17
Jun-17
Aug-16
Sep-16
Nov-16
Dec-16
Aug-17
Sep-17
Nov-17
Dec-17
Figure 3: Control chart of orthophosphate residual at WQP Site D19 (2016-2017)
arcadis.com Page:
6/7
MEMO
Lead Concentration for Location with Paired Data, by Round (94 locations)
Round 1 Round 2 Round 3
Round 4 Round 5 Round 6
100%
90%
Percent of Observations Less Than Value
80%
70%
60%
50%
40%
30%
20%
10%
0%
1 10 100 1,000 10,000
Pb (μg/L)
Figure 4: CLEAR dataset lead concentrations for locations with paired data, by round (94 locations)
arcadis.com Page:
7/7
Arcadis U.S., Inc.
Suite 500
www.arcadis.com