Appendix A: Assessment of Current Practices and Gap Analysis Technical Memorandum

APPENDIX A
ASSESSMENT OF CURRENT PRACTICES AND GAP ANALYSIS

TECHNICAL MEMORANDUM
City of Flint, MI
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:
Arcadis of Michigan, LLC

28550 Cabot Drive
Suite 500
Novi
Michigan 48377
Tel 248 994 2240
Fax 248 994 2241
Our Ref.:
20616001.0000
Date:
March 2017
This document is intended only for the use of

the individual or entity for which it was
prepared and may contain information that is
privileged, confidential and exempt from
disclosure under applicable law. Any
dissemination, distribution or copying of this
document is strictly prohibited.
arcadis.com
ASSESSMENT OF CURRENT PRACTICES AND GAP ANALYSIS TECHNICAL MEMORANDUM
CONTENTS
Acronyms and Abbreviations ........................................................................................................................ vi
Executive Summary....................................................................................................................................... 1
System Overview .................................................................................................................................... 1

Approach ................................................................................................................................................. 2
Gaps ........................................................................................................................................................ 9
Immediate Recommendations and Next Steps .................................................................................... 19
1.0 Introduction ........................................................................................................................................... 21
1.1 Project Background....................................................................................................................... 21

1.2 Report Objectives ......................................................................................................................... 23
1.3 Report Organization ...................................................................................................................... 23
2.0 Water System Overview ....................................................................................................................... 24
2.1 Water Treatment Plant .................................................................................................................. 24
2.1.1 Supply ................................................................................................................................ 24

2.1.2 Treatment .......................................................................................................................... 24
2.1.3 Finished Water Storage and Pumping .............................................................................. 25
2.2 Water Distribution System ............................................................................................................ 25
2.2.1 Distribution System Storage .............................................................................................. 25

2.2.2 Pumping Stations .............................................................................................................. 26
2.2.3 Piping Network .................................................................................................................. 26
2.2.4 System Demands .............................................................................................................. 26
3.0 Methodology.......................................................................................................................................... 28
3.1 Industry Best Practices ................................................................................................................. 28

3.2 Gap Analysis ................................................................................................................................. 36
3.2.1 Coordinate with Regulatory Agencies ............................................................................... 37

3.2.2 Review Available Information ............................................................................................ 39
3.2.3 Conduct Staff Interviews and Site Visits ........................................................................... 42
3.2.4 Identify Gaps ..................................................................................................................... 42
4.0 Results .................................................................................................................................................. 44
4.1 Disinfectant Residual and Disinfection Byproducts ...................................................................... 44
arcadis.com i
4.1.1 Background ....................................................................................................................... 44

4.1.2 Data Sources ..................................................................................................................... 45
4.1.3 Data Analysis..................................................................................................................... 46
4.1.4 Summary ........................................................................................................................... 56
4.2 Main Breaks and Impacts of Pressure .......................................................................................... 57
4.2.1 Background ....................................................................................................................... 57

4.2.2 Data Sources ..................................................................................................................... 58
4.2.3 Data Analysis..................................................................................................................... 59
4.2.4 Summary ........................................................................................................................... 74
4.3 Pressure Management ................................................................................................................. 76
4.3.1 Background ....................................................................................................................... 76

4.3.2 Data Sources ..................................................................................................................... 77
4.3.3 Data Analysis..................................................................................................................... 77
4.3.4 Summary ........................................................................................................................... 78
4.4 Internal Corrosion Control ............................................................................................................. 79
4.4.1 Background ....................................................................................................................... 79

4.4.2 Data Sources ..................................................................................................................... 79
4.4.3 Data Analysis..................................................................................................................... 83
4.4.4 Summary ........................................................................................................................... 90
4.5 Distribution System Practices ....................................................................................................... 95
4.5.1 Cross-Connection Control ................................................................................................. 95

4.5.2 Customer Complaint Tracking and Response .................................................................. 97
4.5.3 Flushing ............................................................................................................................. 98
4.5.4 Hydraulic Modeling .......................................................................................................... 101
4.5.5 Installation, Operation and Maintenance of Valves and Hydrants .................................. 104
4.5.6 Online Monitoring ............................................................................................................ 110
4.5.7 Pipeline Installation, Rehabilitation and Replacement .................................................... 111
4.5.8 Pump Station Design, Operation and Maintenance ........................................................ 113
4.5.9 Security and Emergency Management ........................................................................... 116
4.5.10 Storage Facility Design, Operation and Maintenance ..................................................... 119
4.5.11 Water Age Management ................................................................................................. 121
arcadis.com ii
4.5.12 Water Loss Control .......................................................................................................... 124

4.5.13 Water Quality Sampling and Response .......................................................................... 128
4.6 Supporting Utility Programs ........................................................................................................ 132
4.6.1 Administration .................................................................................................................. 132

4.6.2 Asset Management.......................................................................................................... 137
5.0 Summary and Immediate Recommendations..................................................................................... 139
5.1 Summary..................................................................................................................................... 139
5.1.1 Approach ......................................................................................................................... 139

5.1.2 Gaps ................................................................................................................................ 139
5.2 Immediate Recommendations .................................................................................................... 149
6.0 References .......................................................................................................................................... 151
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
Table 4-8: Water Department Staffing Summary ...................................................................................... 134

Table 4-9:Comparison of Survey Results to Current Flint Staffing and Training ...................................... 135
Table 5-1: Areas that Meet or Exceed Industry Standards ....................................................................... 140
Table 5-2: Areas that Need Minor Strengthening ...................................................................................... 141
Table 5-3: Areas that Need Major Strengthening ...................................................................................... 144
Table 5-4: Areas Where No Program Exists ............................................................................................. 147
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
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
ACRONYMS AND ABBREVIATIONS

AL Action Level
AMR/AMI Automated Meter Reading/Advanced Metering Infrastructure
Arcadis Arcadis of Michigan, LLC
AWOP Area-Wide Optimization Program
AWWA American Water Works Association
BLL Blood Lead Level
CCCP Cross-Connection Control Program
CCPP Calcium Carbonate Precipitation Potential
CF Conventional Flushing
CFU/mL Colony-Forming Units per Milliliter
CIP Capital Improvement Plan
City City of Flint
CMMS Computerized Maintenance Management System
DBP Disinfection Byproduct
DBPR Disinfectants and Disinfection Byproduct Rule
DI Deionized
DSMP Distribution System Monitoring Panel
DWRF Drinking Water Revolving Fund
DWSD Detroit Water and Sewerage Department
E. coli Escherichia coli
EAM Enterprise Asset Management
EPS Extended Period Simulation
ERP Emergency Response Plan
EUL Effective Useful Life
EWQP Extended Water Quality Parameter
FEMA Federal Emergency Management Agency
ft. Foot
ft/s Feet per second
FWAS Free Water Audit Software
arcadis.com vi
GAC Granular Activated Carbon

GIS Geographic Information System
GLWA Great Lakes Water Authority
gpm Gallons per Minute
gpm/ft2 Gallons Per Minute per Square Foot
GPS Global Positioning System
HAA5 Haloacetic Acids Five
HPC Heterotrophic Plate Count
IAP Independent Advisory Panel
in. Inch
ISA International Society of Automation
ISO International Standards Organization
KWA Karegnondi Water Authority
LCR Lead and Copper Rule
LOS Level of Service
LRAA Locational Running Annual Average
LSL Lead Service Line
MCL Maximum Contaminant Level
MDEQ Michigan Department of Environmental Quality
MDHHS Michigan Department of Health and Human Services
MG Million Gallon
MGD Million Gallons per Day
mg/L Milligrams per liter
mL Milliliter
mS MicroSiemens
NIST National Institute of Standards and Technology
NOM Natural Organic Matter
NREPA Natural Resource & Environmental Protection Act
NRW Non-Revenue Water
NTU Nephelometric Turbidity Units
O&M Operations and Maintenance
arcadis.com vii
OWQM Online Water Quality Monitoring

Partnership Partnership for Safe Water
PLC Programmable Logic Controller
POE Point of Entry
POU Point of Use
psi Pounds per Square Inch
QAP Quality Assurance Program
R&R Rehabilitation and Replacement
RAA Running Annual Average
RTCR Revised Total Coliform Rule
SCADA Supervisory Control and Data Acquisition
SDWA Safe Drinking Water Act
SOP Standard Operating Procedure
SPC Standard Plate Count
SWTR Surface Water Treatment Rule
TCR Total Coliform Rule
THM Trihalomethane
TOC Total Organic Carbon
TTHM Total Trihalomethanes
UDF Unidirectional Flushing
µg/L Micrograms per Liter
U.S. United States
USEPA United States Environmental Protection Agency
VFD Variable Frequency Drive
WQL Water Quality Laboratory
WQP Water Quality Parameter
WRF Water Research Foundation
WSC Water Service Center
WTP Water Treatment Plant
arcadis.com viii
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.
Table ES-1: Distribution System Pipe Characteristics
Item Description
 52% as 6-in pipe (~300 miles)

Pipe Diameter
 16% of remaining pipe ranges from 2-in to 72-in (~98 miles)
arcadis.com 1
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:
 Disinfection residual (representative of distribution system water quality integrity),

 Main breaks (representative of distribution system physical integrity), and
 Pressure (representative of distribution system hydraulic integrity).
In addition to metrics tied to the three major performance variables, the Arcadis team evaluated other
factors (administrative, design, maintenance, and operations) that can influence, or can be influenced by,
these integrity performance indicators. With the exception of the administrative topics and asset
management (which will be assessed in greater detail as part of future work on this project), results for
each optimization topic were categorized into one of the four qualitative designations shown in Figure ES-
1. This categorization will help to identify potential early action items and to prioritize elements for
implementation, which will occur as part of the development of the overall Optimization Plan.
arcadis.com 2
Areas that Meet

Areas that Need Areas that Need Areas where No
or Exceed
Minor Major Program or
Industry
Strengthening Strengthening Practice Exists
Standards
Figure ES-1: Gap Analysis Categories
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
Category Description Source

Performance Indicators
Friedman et al. (2010b). Criteria for
≥0.2 mg/L and ≤4.0 mg/L in 95% of samples Optimized Distribution Systems
collected on a monthly basis from locations AWWA. (2011c). Partnership for Safe
above and beyond TCR sites Water Self-Assessment Guide for
Chlorine Residual
Distribution System Optimization
Alexander & Dugan. (n. d.) Area-Wide
≥0.2 mg/L at all locations in the distribution Optimization Program’s Approach to
system at all times Maintaining Distribution System Water
Quality
≤15 reported breaks and leaks per 100 miles
Optimized Distribution Systems
of utility controlled distribution system piping
per year1 AWWA. (2011c). Partnership for Safe
Main Breaks Water Self-Assessment Guide for
Rolling 5-year reduction trend
Establish an annual goal and system for AWWA (2015j). G200-15 Distribution
recording and reporting breaks Systems Operation and Management
20 psi (minimum including peak day and fire
flow) from 99.5% of readings taken from AWWA. (2011c). Partnership for Safe
lowest pressure locations Water Self-Assessment Guide for
Maximum pressures and fluctuations meet Distribution System Optimization
Pressure
utility-specified goals in 95% of readings
35 psi (minimum normal working)
60-80 psi (normal working) GLUMRB. (2012). Ten States Standards
100 psi (maximum)
arcadis.com 3

Improvement Variables
Establish and enforce cross connection and AWWA. (2011c). Partnership for Safe
backflow prevention program Water Self-Assessment Guide for
Install backflow prevention, where required, Distribution System Optimization
and test at least annually AWWA (2015j). G200-15 Distribution
Cross-Connection Systems Operation and Management
Develop and maintain a backflow incident
Control
response plan AWWA. (2015k). M14 Recommended
Practice for Backflow Prevention and
Cross Connection Control
USEPA. (2001). Potential Contamination
Maintain pressure >20 psi (all points in
Due to Cross-Connections and Backflow
distribution system)
and the Associated Health Risks
Establish a program to track and respond to
customer complaints Friedman et al. (2010b). Criteria for
Customer Complaint Maintain up-to-date contact information for Optimized Distribution Systems
Tracking and critical or special needs customers (e.g., AWWA. (2011c). Partnership for Safe
Response health care providers, schools, and regulators) Water Self-Assessment Guide for
<2.5 Technical Quality Complaints per 1,000 Distribution System Optimization
customer accounts annually
Each TTHM sample ≤ 80 µg/L
Each haloacetic acids (HAA5) sample ≤ 60 AWWA. (2011c). Partnership for Safe
µg/L Water Self-Assessment Guide for
Employ methods to maintain or reduce Distribution System Optimization
disinfection byproducts (DBPs)
Each TTHM locational running annual
average (LRAA) ≤ 80 µg/L
Disinfection USEPA. (2006). Stage 2 Disinfectants
Byproduct Control Each HAA5 LRAA ≤ 60 µg/L
and Disinfection Byproducts Rule (DBPR)
Bromate running annual average (RAA) ≤ 10
µg/L (at point of entry only)
Individual sites: Quarterly maximum TTHM
LRAA ≤ 70 µg/L and HAA5 LRAA ≤ 50 µg/L Alexander & Dugan. (n. d.) Area-Wide
Optimization Program’s Approach to
Long-term: Maximum TTHM LRAA ≤ 60 µg/L Maintaining Distribution System Water
and HAA5 LRAA ≤ 40 µg/L (based on 11 Quality
quarters)
Develop a systematic and localized flushing
plan and metrics to monitor flushing
AWWA. (2011c). Partnership for Safe
Develop a communication plan to notify
Water Self-Assessment Guide for
Flushing stakeholders of planned flushing
Friedman et al. (2005). Development of
Flush to 2.0 Nephelometric Turbidity Units
Distribution System Water Quality
(NTU) turbidity at hydrants
Optimization Plans.
arcadis.com 4

HDR Engineering, Inc. et al. (2006).
Flush velocity >3.5 ft/s to avoid particle
Inorganic Contaminant Accumulation in
deposition
Distribution Systems
Avoid damaging or dislodging tubercles from
AWWA. (In Publication). M68 Water
unlined cast iron pipes
Quality in Distribution Systems
Equip dead ends with flushing devices
GLUMRB. (2012). Ten States Standards
targeting a velocity > 2.5 feet per second (ft/s)
Calibrate model using field data from many
points within the distribution system AWWA. (2011c). Partnership for Safe
Include minimum water quality parameters of
chlorine residual and water age
AWWA. (2012b). M32 Computer
Establish a model maintenance plan for
Modeling of Water Distribution Systems,
updating and calibrating the model
Third Edition
Update the model at least every 6 months
Model must include:
 Extended-period simulations
Hydraulic Modeling
 A minimum of 75 percent of pipe
volume and 50 percent of pipe length
 All pressure zones
 All 12-inch diameter and larger pipes USEPA. (2006). Stage 2 DBPR
 All diameter pipes which are
hydraulically significant
 All storage facilities with typical
operations represented
 All active pumping stations and all
active controls
90th percentile lead ≤ 0.015 mg/L
90th percentile copper ≤ 1.3 mg/L
Establish a routine monitoring program for Friedman et al. (2010b). Criteria for
corrosion-related water quality parameters Optimized Distribution Systems
Regularly inspect internal condition of piping AWWA. (2011c). Partnership for Safe
for tuberculation, pitting, holes or scaling Water Self-Assessment Guide for
Internal Corrosion Develop written goals and SOPs for control of Distribution System Optimization
Control corrosion-related byproducts; water should
meet secondary standards for corrosion-
related parameters (i.e., 0.3 mg/L for iron)
100% samples for lead and copper below their
respective action levels (AL) (all points in the Kirmeyer et al. (2002). Guidance Manual
distribution system) for Monitoring Distribution System Water
Iron < 0.1 mg/L (all points in the distribution Quality
system)
arcadis.com 5

Develop and maintain an asset inventory AWWA. (2006a). M17 Installation, Field
Develop a process and system to assess Testing and Maintenance of Fire
condition Hydrants, Fourth Edition
Develop a work order management process to AWWA. (2006c). M44 Distribution
track and prioritize maintenance Valves: Selection, Installation, Field
Testing and Maintenance, Second
Installation, Establish a planning process that includes
identified metrics for operability, a Edition
Operation and
Maintenance of replacement program to meet those metrics, Friedman et al. (2010b). Criteria for
Valves and Hydrants budgets, and coordination with other Optimized Distribution Systems
programs AWWA. (2011c). Partnership for Safe
Develop a communication plan to notify Water Self-Assessment Guide for
stakeholders of maintenance activities Distribution System Optimization
Includes design specifications for locations,

spacing and sizing of hydrants and valves
Implement a continuous monitoring system to
monitor flows and pressures at the major point
of entry, tanks, pump stations and other key
points in the distribution system Friedman et al. (2010b). Criteria for
Install online water quality monitoring at least Optimized Distribution Systems
Online Monitoring at the largest point of entry and at 3 AWWA. (2011c). Partnership for Safe
distribution locations for chlorine residual, pH, Water Self-Assessment Guide for
turbidity temperature and conductivity; monitor Distribution System Optimization
total organic carbon at a minimum of one
location (such as largest point of entry)
Install at least 5 pressure monitoring sites
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
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
Turbidity 0.5 – 1.0 NTU (2.0 NTU at flushed
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

Develop and maintain an asset inventory
Establish a defined operation and
maintenance program
Pump Station Track and report the Planned Maintenance AWWA. (2011c). Partnership for Safe
Design, Operation Ratio (%) = 100 (hours planned maintenance) Water Self-Assessment Guide for
and Maintenance / (hours of planned + corrective maintenance); Distribution System Optimization
well-maintained systems exceed 80% for this
metric
Develop a motor and pump rehabilitation and
replacement program
AWWA. (2001). M19 Emergency
Planning for Water Utilities, Fourth
Develop an Emergency Response Plan (ERP) Edition
that meets Federal Emergency Management
Agency (FEMA) standards and conduct AWWA. (2010). J100-10 Risk and
regular exercises Resilience Management of Water and
Wastewater Systems
Conduct a J100 Vulnerability Assessment and
implement risk mitigation measures FEMA (2010). Developing and
Maintaining Emergency Operations
Security and Cultivate a security culture Plans: Comprehensive Preparedness
Emergency Implementation of cybersecurity measures Guide 101, Version 2.0
Management
Hardening of physical security AWWA (2011b). G440-11 Emergency
Develop robust internal and external Preparedness Practices
communication plans AWWA. (2011c). Partnership for Safe
Develop strategic partnerships Water Self-Assessment Guide for
Distribution System Optimization Ten
Provide standby power required to meet States Standards (2012).
average day demand
AWWA. (2014a). G430-14 Security
Practices for Operation and Management
Develop standard operating procedures
(SOPs):
 That define optimum water level
changes to ensure acceptable pressures
and flows and targeted turnover rates to
minimize water age
 To address water quality
 For routine inspections of storage tanks Friedman et al. (2010b). Criteria for
Storage Facility Assess condition of tank and coatings, Optimized Distribution Systems
Design, Operation corrosion control and any safety, sanitary or AWWA. (2011c). Partnership for Safe
and Maintenance structural deficiencies annually Water Self-Assessment Guide for
Provide critical security and Distribution System Optimization
telecommunications equipment
Monitor weekly at a minimum for disinfectant
residual
Conduct periodic sampling surveys to assess
residual uniformity
Track and control water age in the storage
facility
arcadis.com 7

Minimum chlorine residual of 0.5 mg/L and full Kirmeyer et al. (2002). Guidance Manual
turnover every 5 days for Maintaining Distribution System Water
Daily turnover of at least 20% Quality
AWWA) & Economic and Engineering

Storage tank inspection every 3-5 years Services, Inc. (2002). Finished Water
Storage Facilities
AWWA. (2011a). C652-11 Disinfection of
General guidelines for tank disinfection
Water-Storage Facilities
Detailed procedures for inspecting and AWWA. (2013b). M42 Steel Water
repairing steel tanks Storage Tanks, Revised Edition
Routine inspections daily/weekly
Periodic inspections monthly/quarterly
AWWA. (2015j). G200-15 Distribution
Comprehensive inspections every 3-5 years
Systems Operation and Management
Cleaning at least every 5 years
Disinfect after draining
Commit qualified staff and resources to review
distribution system and storage configurations
Establish surveillance testing schedules at key
locations in the system for water age
Implement proper maintenance practices (i.e.,
unidirectional flushing)
Water Age Distribution System Optimization
Management Train staff to recognize and response to
customer complaints resulting from water age
Water age < 5-7 days at all points in the Friedman et al. (2005). Development of
distribution system Distribution System Water Quality Plans
Reduce and establish target turnover rates
arcadis.com 8

AWWA. (2016b). M36 Water Audits and
Loss Control Programs, Fourth Edition
AWWA. (2006). M33 Flowmeters in
Establish internal water loss management Water Supply, Second Edition
team Friedman et al. (2010b). Criteria for
Perform annual water audits Optimized Distribution Systems
Improve audit data validity score AWWA. (2011c). Partnership for Safe
Determine present expenditures on water loss Distribution System Optimization
control
AWWA. (2012). M6 Water Meters –
Water Loss Control Determine and achieve economic levels of Selection, Installation, Testing and
water loss (apparent and real) Maintenance, Fifth Edition
Make business case for activities to achieve AWWA. (2014). M22 Sizing Water
economic levels of water loss Service Lines and Meters, Third Edition
Develop cost-effective water loss control plan AWWA. (2014). Policy Statement on
Select performance targets and conduct water Metering & Accountability
loss performance monitoring AWWA. (2015). G200-15 Distribution
AWWA. (2015 – 2016). Standards C700
– C750 for Water Meters
Develop a sampling plan and SOPs for
training, tap flushing, field measurements,
sample collection, sample storage and
transport and documentation
Perform monitoring at:
 Known problem locations and establish
trigger levels with responses (routine) Friedman et al. (2010b). Criteria for
Water Quality Optimized Distribution Systems
 Storage facilities (at least weekly) for
Sampling and AWWA. (2011c). Partnership for Safe
bacteria, disinfectant residual and other
Response Water Self-Assessment Guide for
appropriate parameters
 Chemical treatment points in the
distribution system (for process control)
 Upstream and downstream of main
breaks and repairs (non-routine)
 Customer complaint locations (non-
routine)
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
Table ES-3: Areas that Meet or Exceed Industry Standards
Areas that Meet or Exceed Industry Standards

Category Description
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. As of October 2016, the City is also meeting a more rigorous goal
of ≥ 0.5 mg/L in 95% of samples collected on a monthly basis.
Disinfectant Residual
DBPs are below 50% of their respective 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.
Disinfection of repaired mains and placement into service appears to be conducted in
accordance with standard practice.
Main Breaks
There is sufficient (possibly excessive) redundancy in the system, so main breaks can
be isolated despite valve problems, while still providing service to customers.
Pressure Management Not possible to determine at this time.
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
Internal Corrosion
systems in the U.S.
Control
Water quality parameter (WQP) monitoring for an extended list of parameters is
currently conducted at least once per week at 10 locations.
Cross-Connection The existing cross-connection control program appears to align with AWWA Manual of
Control Practice M14.
Customer Complaint Detailed analyses are conducted on water quality samples collected from complaint
Tracking and Response locations. The analyses provide insight into metals, aesthetic, and microbial conditions
The City’s conventional and dead-end flushing practices currently used to manage
water age meet and exceed industry standards.
Specific flow rates have been targeted that are known to avoid sediment and biofilm
disruption and spreading of contaminants.
Flushing
Numerous programmable auto-flushers are used and can be rotated around the City
as needed to meet water quality objectives.
A hydrant sampling device has been created and is used to control flow rates and
collect sanitary water quality samples
The model has been setup to perform real-time modeling, although it is not directly
connected to the City’s supervisory control and data acquisition (SCADA) system at
Hydraulic Modelling
this time. The real-time modeling method to achieve calibration for a specific period of
time is above expectations.
The City has made significant investment progress on the valve assessment and
exercising program. This progress demonstrates the commitment of City resources to
Installation, Operation, ensure the information flows from the office to the field and that current business
and Maintenance of processes do allow the completed work to be reported up to departmental
Valves and Hydrants management.
The fire hydrants within the City of Flint have been able to meet the needs of
emergency services.
Chlorine residual analyzers have been installed upstream and downstream of
Online Monitoring
chemical addition points to provide dosing feedback.
arcadis.com 10

Pipeline Installation,
Disinfection of new mains and placement into service appears to be conducted in
Rehabilitation, and
Replacement
Pump Station Design,
Operation, and None identified.
Maintenance
The City meets industry standards according to the Ten States Standards requiring all
Security and Emergency exterior entrances to lock, fencing around facilities (with the notable exception of the
Management Cedar Street Reservoir and Pump Station, which is not entirely enclosed in fencing),
intrusion alarms, and some security camera monitoring.
Storage Facility Design, Recent modifications to improve turnover, conduct monitoring with depth to verify
Operation, and conditions, and to improve controls of boosting operations are positive operational
Maintenance steps that meet industry standards.
Use of a hydraulic model to understand areas of highest water age.
Water Age Management
Use of automated flushing devices (seasonal).
Water Loss Control None identified.
RTCR Plan and chlorine surveillance monitoring plans meet industry standards.
Water Quality Sampling
and Response Extended water quality parameter (EWQP) monitoring exceeds industry standards in
terms of monitoring frequency and parameters included.
Table ES- 4: Areas that Need Minor Strengthening
Areas that Need Minor Strengthening

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)
Use hydraulic model to identify potential areas of system with low chlorine residuals
Disinfectant Residual and conduct surveillance monitoring and site-specific improvements at these locations.
Identify optimal chlorine dosing to target 0.2- 0.5 mg/L in 95% of sample locations
throughout distribution system.
Identify seasonal optimal boosting strategies at plant and reservoirs.
Continue flushing in areas where maintaining a detectable chlorine residual has been
challenging and identify opportunities to improve residual levels without further
increasing dosages.
Update the procedure for documenting and responding to main breaks.
Additional information regarding the approximate time of the break, Type
categorization, visual description (shear break, longitudinal, hole) should be captured.
Main Breaks 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 evaluation 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.
arcadis.com 11

Newly added pressure monitoring stations and site selection for portable data loggers
need to be confirmed by the City’s hydraulic model and incorporated into the City’s
Pressure Management
SCADA. Additional recommendations for improved pressure control may be
developed once results from distribution system data loggers are available for review.
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. High orthophosphate doses
could contribute to formation of white precipitate (calcium phosphate) in the
Internal Corrosion distribution system.
Control
The orthophosphate level at Site D9/D13 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.
Analyze recorded data 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.
Customer Complaint Complaints should be entered into a geographic information system (GIS) or a
Tracking and Response computerized maintenance management system (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.
Cross-Connection
None identified.
Control
Flowrates during conventional and auto-flushing can likely be increased after field
Flushing verification. 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 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
Hydraulic Modelling
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.
Demand allocation and representation within the model is unintuitive and does not
align with industry standards.
Continue efforts related to obtaining global positioning system (GPS) coordinates for
all appurtenances.
Enhance the GIS database schema to include other valve and hydrant attributes,
types, and functions.
Installation, Operation, Establish a comprehensive data model review to consolidate information into a single
and Maintenance of source.
Valves and Hydrants Consider modifying existing asset IDs for consistency with industry standards.
Strengthening may be needed to ensure asset information is accessible to the specific
person who is going to locate, access, and operate the asset in a reasonable amount
of time (including historical maintenance records and reliability information, currently in
paper and excel).
Online Monitoring Timely export of SCADA data into Grafana or similar visualization software.
arcadis.com 12

Review existing technical specifications and construction details and update as
needed to current standards based on the Ten States Standards and AWWA.
Develop an established schedule for updating and maintaining technical specifications
and standard construction details for new water main installations and replacements in
Rehabilitation, and
accordance with AWWA Standard G200.
Replacement
Develop an approval procedure for verifying the installation was completed according
to these newly developed City standards, and that pressure testing and disinfection
was in compliance with the appropriate regulatory agencies.
Install globe body style control valves at West Side Reservoir to reduce the potential
for pressure surges during startup and shutdown of the pumps.
Develop written SOPs 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.
Pump Station Design, Develop detailed maintenance procedures for each pump station defining the
Operation, and frequency, procedures, and maintenance of records for each piece of mechanical and
Maintenance electrical equipment. Expand any existing SOPs with additional items for
consideration listed below:
Monthly inspections including equipment lubrication, vibration checks, bearing
temperature readings, packing adjustments, etc.
Annual electrical system maintenance including variable speed drives, transformers,
and motor starters.
Establish of formal long-term maintenance and rehabilitation program.
Increase control and intrusion detection per guidance in the AWWA Standard G430-
14.
Security and Emergency Perform annual inspections of identified critical assets.
Management Completely enclose the Cedar Street Reservoir site in fencing, per guidance in the
Ten States Standard and AWWA Standard G430-14. Additionally, the fence should be
maintained clear of plant growth.
Storage Facility Design, Replace inlet valves to West Side and Cedar Street Reservoirs with globe body style
Operation, and altitude valves having back pressure sustaining pilots. The opening and closing rate
Maintenance of these valves can be adjusted to minimize the potential for pressure surges.
Use hydraulic model to investigate optimal storage reservoir operations to reduce
detention time and avoid sequential draining/filling.
Manual flushing practices must be strategically located to turnover water in the highest
age areas.
Streamline monitoring locations across monitoring plans.
Ensure that the current quality assurance program (QAP) is up to date and captures
most recent and approved monitoring methods and sampling procedures.
and Response
Conduct side-by-side monitoring of turbidity, iron, and apparent color for further
streamlining of surveillance monitoring program.
arcadis.com 13
Table ES- 5: Areas that Need Major Strengthening
Areas that Need Major Strengthening

Continue efforts and approach outlined in the Chlorine Residual Management Program to
optimize residual levels while reducing dosages and minimizing variability.
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 immediately.
Standard operating procedures for pump start/stops and valve operations are required
Main Breaks 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. Initiate a regular leak detection program.
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 in SCADA and operate the
Pressure system to achieve Partnership pressure goals.
Management Conduct surge modelling to optimize operations of the reservoir fill valves, pump
sequencing, filter backwash cycles, and elevated tank operations. Determine the need for
variable frequency drives (VFDs), soft starts, and other surge control devices.
Create standard procedures for valve open and close operations.
Lead levels at the higher end of the percentile curve are still excessive and do not
Internal Corrosion
currently meet the performance target recommended in Kirmeyer et al. (2002) of 100
Control
percent of the samples for lead below the action level of 15 µg/L.
Staff levels need to be increased to improve the management and supervision of the
Cross-Connection
cross-connection control program and maintain re-certification testing of the 1,000
Control
backflow prevention devices.
Customer Complaint
Tracking and Develop a Customer Complaint Response SOP.
Response
The City should assess the effectiveness of unidirectional flushing (UDF) for reducing
pipe wall chlorine demands and removing accumulated sediment from the distribution
Flushing system by conducting a UDF pilot study.
Estimated level of effort for developing and implementing a UDF program will be prepared
as part of the Resource Analysis and Needs Assessment.
Facilities such as pumps should be confirmed and better represented within the
distribution system model using available pump curves, modified as needed to reflect
current operation.
Hydraulic Modelling
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.
arcadis.com 14

The Valve Exercising and the Hydrant Replacement Program is currently prioritized in the
capital improvement plan (CIP) as level 2 (Important) in the 2016 Capital Improvement
Plan and should be re-evaluated to consider the feasibility to moving to a level 1 (Urgent).
Installation, It is recommended that the City consider departmental organization and other options to
Operation, and optimize resources.
Maintenance of Establish a documented plan for valve assessment and exercising and hydrant
Valves and Hydrants assessment and replacement in accordance with AWWA Standard G200.
Re-evaluate risk criteria used in previous studies to identify critical pipeline segments and
apply to appurtenances (current CIP identifies critical valves based solely on size).
Incorporate additional scoring criteria based on a triple-bottom line approach.
Online Monitoring None identified.
Develop a formal risk-based approach as the foundation of their watermain
replacement/rehabilitation program in accordance with AWWA Standard G200.
Replace a minimum of 1% of watermains per year with a goal of maintaining the age of all
piping within the system to less than 100 years.
Develop a program for assessing the physical condition of the pipes including the
following:
 A GIS for the distribution system that includes pipe attribute information such as
Pipeline Installation, age, diameter, material, length, and location
Rehabilitation, and  Maintenance records for all leaks and breaks. Data to be recorded as part of a
Replacement repair include diameter, material, pipe class, interior/exterior coatings or linings,
repair type, joint type, soil conditions, and GPS coordinates. This data should
then be incorporated into the City’s GIS.
 Surrounding environmental information at the leak site such as soil type,
corrosion potential, and potentially hazardous material identified during the
excavation.
 Distribution system data such as fire-flow tests, c-factor tests, and pressure data
that can be incorporated into the City’s GIS.
Maintenance
Update the existing Emergency Response Plan (ERP) to meet Federal Emergency
Management Agency (FEMA) standards.
Regular exercise of the ERP requires major strengthening and is recommended in the
Partnership for Safe Water Self-Assessment Guide, AWWA Standard G430-14, and the
AWWA Standard G440-11
Security and
Emergency Development and promote a security culture. This includes commitment from senior
Management leadership and promotion of security awareness, as recommended in AWWA Standards
G430-14 and G440-11.
Development and strengthen communications and partnerships to enable Flint to best
respond in emergency situations as recommended in AWWA Standards G430-14 and
G440-11. It is also recommended that the City develop mutual aid and assistance
partnerships.
Per AWWA Standard G200 and Manual of Practice M42, establish formal procedures for
Storage Facility
the inspections of each storage facility. These inspections should focus on structural,
Design, Operation,
safety, security, coatings system, and sanitary conditions. It is recommended that the
and Maintenance
following timeframes be established for the inspections:
arcadis.com 15

 Monthly Inspection – To be completed by city personnel
 Annual Inspection – To be performed by city personnel and/or contractor
 Comprehensive Inspection – To be performed by a professional services
contractor with support from city personnel
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.
Water Age Review the overall system water demands, storage needs and capacity of their existing
Management storage facilities.
Perform annual water audits in accordance with the AWWA Manual of Practice M36
methodology.
Develop a Water Loss Control Plan that aligns with the Water Department’s
organizational goals and objectives, and adjust, as appropriate, going forward. Key
activities needed for the plan development include:
 Perform an annual water audit, using AWWA Free Water Audit Software
(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 level needed to establish performance targets for activities comprising Flint’s
Water Loss Control Plan.
 Determine present expenditures for water loss management-related activities
 Perform an annual meter accuracy determination of all distribution system entry
point flow meters, including volumetric test and calibration of associated
instrumentation and apply errors in the annual water audit.
 Verify the presence of all customer connections, meter readings, and billed
Water Loss Control metered consumption in the Read-to-Bill System.
 Quantify all types of authorized consumption (billed metered, billed unmetered,
unbilled metered, unbilled unmetered) and correlate with SCADA-reported flow.
 Manage apparent loss control through determining the accuracy of residential
and commercial meters and maintaining/replacing as necessary. Evaluate
sources of unauthorized consumption and systematic data handling errors.
 Control real losses by accounting for storage leakage, main, service line, and
appurtenance leakage, and coordinating findings with an ongoing asset
management program. Evaluate pressure management and pipe/service line
repair activities to minimize leakage.
 Perform cost:benefit analysis of candidate activities and select activities for
implementation
 Establish performance targets, emphasizing recommended financial and
operational performance indicators (while minimizing reliance on percentage-
based operational indicators) and monitor on an ongoing basis
 Benchmark activities against those of comparable water systems and update as
appropriate.
Electronic capture of monitoring results – lack of a data management system.
Water Quality Routine tracking and trending of key parameters and relationships (monthly, quarterly).
Sampling and Convert heterotrophic plate count (HPC) analyses from Standard Plate Count (SPC) agar
Response to R2A agar (7-day hold time at 20oC). A new incubator may be required. Conduct side-
by-side monitoring for a few months to assist with interpretation of R2A results.
arcadis.com 16

Purchase new deionized water system to ensure quality of standards and analytical
results.
Table ES-6: Areas Where No Program Exists
Areas Where No Program Exists

Verify and track chlorine dosing/chemical usage at storage facilities.
On-going tracking and trending of residual levels throughout distribution system.
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.
Main Breaks
There is a lack of city vehicles to efficiently respond to and/or complete maintenance.
There are no GPS coordinates of pipes and valves, and no properly updated water
system map.
As far as is understood, no regular pressure monitoring and assessment is currently
conducted within the distribution system.
Data from pressure data loggers is needed to provide insight to distribution system
Pressure response to current operations and to compare to Partnership pressure management
Management goals.
Surge modelling to verify and optimize WTP, storage, and pump station operations.
Evaluation of use of one pressure zone for entire distribution system.
Internal Corrosion
None identified.
Control
Implement a cross-connection control plan on the residential level in accordance with
Cross-Connection AWWA Manual of Practice M14.
Control Establish a formal backflow incident response plan including tracking and investigation of
backflow incidents.
Customer Complaint
Develop a SOP for customer complaint response and tracking. Include a Communications
Tracking and
Plan between the WQL and WSC.
Response
Flushing Establish a UDF Program.
The City has not developed a model maintenance plan, or regular schedule for updating
the model components and performing periodic calibration.
Training for City staff should include basic model usage, model capabilities, modeling
goals, location of input/output data, instructions on how to export model results for
Hydraulic Modelling
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.
arcadis.com 17

Need for documented current or future levels 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:
 System renewal/replacement rate
 Water distribution system integrity
 Operations & maintenance cost ratios
 Planned maintenance ratios (%)
 Average time to address planned and unplanned service disruption (hours)
 Customer service complaints per 1,000 accounts
Installation,  Leaks/breaks per 100 miles of pipe
Operation, and  System Inspection (%)
Maintenance of
Develop SOPs including valve/hydrant installation and procedures, valve/hydrant
Valves and Hydrants
preventative maintenance program and procedures, valve emergency and corrective
maintenance procedures, work order documentation protocols, and community fire risk
assessment.
Confirm manufacturer manuals are available or used to establish/verify preventive
maintenance requirements.
Need for hydrant/valve coordination/training with other departments/agencies (fire
department, traffic accident reporting, street cleaning, sewer cleaning, building
construction, recreation, contractors, etc.).
Public notification of a potential impact to service during planned or unplanned
maintenance.
Incorporation of hydrant testing results into the hydraulic model calibration.
For systems with populations greater than 50,000, the Partnership recommends, at a
minimum, the following security-related online monitoring:
 Point of entry (POE) monitoring chlorine residual, pH, turbidity temperature,
pressure and conductivity
 Three distribution sites monitoring chlorine residual, pH, turbidity temperature
and conductivity,
Online Monitoring  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.
Apart from TOC monitoring, the HACH Distribution System Monitoring (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.
Rehabilitation, and None identified.
Replacement
Pump Station Design, Conduct a needs assessment at each pump station to identify short and long-term
Operation, and capacity requirements as well as other capital needs. With the downward trend in water
Maintenance consumption within the City, smaller capacity pumps may be necessary.
arcadis.com 18

No existing cyber security or firewall for SCADA or instruments outside of the plant on a
wireless network. Use of firewalls is recommended in the Ten States Standard and
AWWA Standard G430-14.
No confirmed level of monitoring and security of most distribution systems assets,
particularly pertinent distribution system assets (e.g., critical valves or mains), as
recommended in the Ten States Standard.
Security and
Emergency No confirmed existing all-hazards vulnerability assessment. An up-to-date all-hazards
Management vulnerability assessment is recommended by the Partnership, AWWA Standard G430-14,
AWWA Standard G440-11, and the AWWA Standard J100-10. It is also recommended
that a vulnerability assessment be regularly updated. Additionally, a funding analysis is
recommended to determine if federal or state funds are available to the City for
implementation of risk mitigation measures.
It is recommended that the City participate in the Genesee County Hazard Mitigation Plan
revision process to better understand regional risks that may affect City operations.
Detailed hydraulic analysis of storage facility capacity and optimal operations, considering
Storage Facility
decreased system demand, need for adequate capacity during main breaks, and planned
Design, Operation,
operational changes associated with switch to Karegnondi Water Authority (KWA) and
and Maintenance
operation of Flint WTP.
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,
Water Age commercial, industrial), and develop a plan for system operation as demand decline
Management which maintains adequate water age.
Need for a UDF program.
Tracer evaluations to confirm hydraulic model predictions of water age.
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.
Water Loss Control
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.
Water Quality No data management system.
Sampling and
Response No trending of water quality results.
Immediate Recommendations and Next Steps

Some gaps were identified for immediate action because they directly impact the City’s ability to achieve
compliance with current regulations, minimize public health risks, or inform later tasks. Based on the
findings of this Gap Analysis, immediate recommendations include:
 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
 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
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.
1.1 Project Background

For several decades, the City of Flint purchased finished water from GLWA’s Lake Huron Water
Treatment Plant, which treats water from Lake Huron for compliance with all Safe Drinking Water Act
(SDWA) standards. The water contains a free chlorine residual to prevent biological regrowth and an
orthophosphate residual to control corrosion in the water distribution system. Purchased water from
GLWA was the sole source of supply and no additional treatment was performed prior to distribution to
the City of Flint.
The City maintains the Flint WTP, which can treat water from the Flint River, as a backup emergency
supply. On April 25, 2014, the City stopped purchasing water from GLWA and began treating the Flint
River at the Flint WTP. This change occurred as part of a plan to join the Karegnondi Water Authority
(KWA) upon the original pipeline anticipated completion date of late 2016. From April 25, 2014 through
October 16, 2015, the City continued to treat and distribute water from the Flint River. During this time,
orthophosphate addition was not in place at the plant.
Several water quality concerns arose while on the Flint River supply including the following:
 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
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
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.
1.2 Report Objectives

The purpose of this report is to document the approach and results of the assessment and gap analysis.
This report serves as a summary of data gathering efforts, comparison to industry standards generally
following the American Water Works Association (AWWA) Partnership for Safe Water (Partnership)
approach, and analysis of system gaps.
1.3 Report Organization

The remaining sections of the report are organized as follows:
 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
2.0 WATER SYSTEM OVERVIEW

The following section describes the existing system, including the supply, treatment, distribution system
and demands.
2.1 Water Treatment Plant

The Flint WTP was constructed in 1952, and has primarily served as an emergency water source, except
for the period from April 25, 2014 through October 16, 2015, when it was operated full time. The plant
was maintained and upgraded to treat and distribute water from the Flint River as needed for up to six
weeks (Rowe, 2016a).
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
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.
2.1.3 Finished Water Storage and Pumping

There are three finished water storage facilities on the site of the WTP. The elevated tank has a capacity
of 2 MG and is used as one measure to control pressure throughout the single pressure zone of the
distribution system. There is a 4 MG clearwell/suction well at Pumping Station 4. The Dort Reservoir is a
20 MG partially underground storage facility constructed in 1966 (Rowe, 2016a). At the time of this
memo, Dort reservoir currently houses highly chlorinated water that cannot be released to the distribution
system, thereby rendering Dort Reservoir unusable for the time being. Plans to drain the reservoir to
waste are underway.
Pumping Station 4 is located on the WTP property connected to Dort Reservoir and contains a total of 5
pumps – three 20 MGD, one 25 MGD, and one 6 MGD. However, the 25 MGD pump and one of the 20
MGD pumps were installed in 1949 and are inoperable. The remaining three pumps were installed in
1994 and yield total and firm capacities of 46 MGD and 26 MGD, respectively (Rowe, 2016a).
These facilities are also in the process of being updated as part of the Flint WTP improvements to be
completed by 2019.
2.2 Water Distribution System

The Flint distribution system is operated and maintained by the Water Service Center (WSC), apart from
the storage facilities and pump stations, which are maintained by WTP personnel.
2.2.1 Distribution System Storage

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. The facility fills and empties through a common 36-
inch header. Based on personnel interviews, the oversized exit valve on the header is widely assumed to
cause water hammers and main breaks throughout the distribution system due to difficulty controlling and
lack of SOPs for operators at the plant. The pumping station houses three pumps rated at 12, 9, and 9
MGD yielding a total pumping capacity of 30 MGD and a firm pumping capacity of 18 MGD (Rowe,
2016a). The pumps are not equipped with soft starts or variable frequency drives (VFDs) which leads to
large pressure variations when they are turned on or off. The pumps can be controlled locally or remotely
from the WTP. Chlorine is currently boosted on the fill line into the Cedar Street Reservoir using liquid
sodium hypochlorite. Chlorine samples are measured continuously on-site, as well as collected by lab
personnel at the same time as RTCR sampling.
West Side Reservoir, which houses Pumping Station 5, was constructed in 1970 and has a capacity of 12
MG. The pumping station consists of four turbine pumps, two 8 MGD and two 4 MGD for a total capacity
of 24 MGD and a firm capacity of 16 MGD (Rowe, 2016a). Pumps are controlled remotely from the WTP
or locally. Chlorine is boosted on the fill line to the reservoir using liquid sodium hypochlorite. In-line
analyzers are currently in place to monitor chlorine concentrations. Additionally, chlorine samples are
arcadis.com 25
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.
2.2.2 Pumping Stations

The Flint distribution system also maintains one booster station known as the Torrey Road Pumping
Station. Two pumps, 2.8 MGD and 4 MGD, were installed in 1954 for total and firm capacities of 6.8 and
4 MGD, respectively. Upgrades to the pumps have been purchased and are awaiting installation.
2.2.3 Piping Network

The Flint distribution system has conflicting reports on the length of operating pipelines. A 2016 leak
detection report (Echologics, 2016) asserts the system includes approximately 795 miles of pipelines. The
City’s hydraulic model indicates that there are 579 miles of pipeline. A recent asset management report
(Rowe, 2016b) describes the system as having 583 miles of pipe (nearly identical to what is listed in the
hydraulic model) with the following characteristics:
Pipe Diameter
 52% as 6-in pipe (~300 miles),

 24% as 8-in pipe (~138 miles),
 8% as 12-in pipe (~47 miles), and
 16% of remaining pipe ranges from 2-in to 72-in (~98 miles).
Pipe Material
 96.6% is unlined cast iron pipe (564 miles),

 2.6% is ductile iron (31 miles) and,
 Remaining material includes steel, concrete, galvanized, and unidentified.
Pipe Age
 Everything installed after 1900,

 63% installed between 1900 and 1930,
 33% installed between 1930 and 1960, and
 4% installed after 1960.
Based upon the above, it was assumed in this report that there is approximately 580 miles of drinking
water distribution system pipelines operated by the City.
2.2.4 System Demands

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
arcadis.com 26
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.
Figure 2-1: Purchased/Treated Water Flow (2010 – 2016)
arcadis.com 27
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.
3.1 Industry Best Practices

This Gap Analysis provides an assessment of the City’s 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. The Partnership 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.
Although the Partnership recognizes that it is difficult for any utility to achieve fully optimized status as
defined by its performance goals, it offers a process in which all utilities can advance their system
operations to attain performance nearer the Partnership goals.
Enhancement of system performance is based upon improving water quality, physical, and hydraulic
characteristics of the system. Specifically, system enhancement is based upon three performance
indicators, which include:
Water Quality Integrity. The water quality indicator is based upon the capability of maintaining an
adequate distribution system residual disinfectant level. Based upon Partnership criteria, an optimized
distribution system using free chlorine as a secondary disinfectant will maintain the following:
 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:
 Distribution system pressure should be maintained at or above 20 psi in 99.5% of measurements

(based on daily minimums from hourly readings),
arcadis.com 28
 Maximum pressure does not exceed utility-specified maximum in 95% of measurements

(maximum values vary based on system piping and recommended operating pressures)
 Pressure should be met under normal conditions including maximum day demand and fire flow
conditions (excluding emergencies),
 Utilities have pre-approved procedures to protect public health during emergency conditions
(main breaks, power outages), and
 Utilities have pre-determined goals for maximum pressure ranges (differences between minimum
and maximum pressure) within each pressure zone.
Many other factors (administrative, design, maintenance, and operations) influence, or can be influenced
by, these integrity performance indicators. Optimization through self-assessment of some or all of these
factors (known as improvement variables) is another critical step in the program. In addition to meeting
the goals for the performance indicators above, there are key minimum data requirements related to
these variables that have been established by the Partnership for utilities to be able to successfully
complete the self-assessment of their distribution system. Some of these requirements include the
following
 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
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

Performance Indicators
≥0.2 mg/L and ≤4.0 mg/L in 95% of samples Optimized Distribution Systems
Chlorine Residual
collected on a monthly basis from locations AWWA. (2011c). Partnership for Safe
above and beyond TCR sites Water Self-Assessment Guide for
arcadis.com 30

Alexander & Dugan. (n. d.) Area-Wide
≥0.2 mg/L at all locations in the distribution Optimization Program’s Approach to
system at all times Maintaining Distribution System Water
Quality
≤15 reported breaks and leaks per 100 miles
of utility controlled distribution system piping
per year1 AWWA. (2011c). Partnership for Safe
Main Breaks Water Self-Assessment Guide for
Rolling 5-year reduction trend
Establish an annual goal and system for AWWA (2015j). G200-15 Distribution
recording and reporting breaks Systems Operation and Management
20 psi (minimum including peak day and fire
flow) from 99.5% of readings taken from AWWA. (2011c). Partnership for Safe
lowest pressure locations Water Self-Assessment Guide for
Maximum pressures and fluctuations meet Distribution System Optimization
Pressure
utility-specified goals in 95% of readings
35 psi (minimum normal working)
60-80 psi (normal working) GLUMRB. (2012). Ten States Standards
100 psi (maximum)
Improvement Variables
Establish and enforce cross connection and AWWA. (2011c). Partnership for Safe
backflow prevention program Water Self-Assessment Guide for
Install backflow prevention, where required, Distribution System Optimization
and test at least annually AWWA (2015j). G200-15 Distribution
Cross-Connection Systems Operation and Management
Develop and maintain a backflow incident
Control
response plan AWWA. (2015k). M14 Recommended
Practice for Backflow Prevention and
Cross Connection Control
USEPA. (2001). Potential Contamination
Maintain pressure >20 psi (all points in
Due to Cross-Connections and Backflow
and the Associated Health Risks
Establish a program to track and respond to
customer complaints Friedman et al. (2010b). Criteria for
Customer Complaint Maintain up-to-date contact information for Optimized Distribution Systems
Tracking and critical or special needs customers (e.g., AWWA. (2011c). Partnership for Safe
Response health care providers, schools, and regulators) Water Self-Assessment Guide for
<2.5 Technical Quality Complaints per 1,000 Distribution System Optimization
customer accounts annually
Each TTHM sample ≤ 80 µg/L AWWA. (2011c). Partnership for Safe
Each HAA5 sample ≤ 60 µg/L Water Self-Assessment Guide for
Employ methods to maintain or reduce DBPs Distribution System Optimization
Disinfection
Byproduct Control Each TTHM LRAA ≤ 80 µg/L
USEPA. (2006). Stage 2 Disinfectants
Each HAA5 LRAA ≤ 60 µg/L
and Disinfection Byproducts Rule (DBPR)
Bromate RAA ≤ 10 µg/L (at POE only)
arcadis.com 31

Individual sites: Quarterly maximum TTHM
LRAA ≤ 70 µg/L and HAA5 LRAA ≤ 50 µg/L Alexander & Dugan. (n. d.) Area-Wide
Optimization Program’s Approach to
Long-term: Maximum TTHM LRAA ≤ 60 µg/L Maintaining Distribution System Water
and HAA5 LRAA ≤ 40 µg/L (based on 11 Quality
quarters)
Develop a systematic and localized flushing
plan and metrics to monitor flushing
Develop a communication plan to notify
stakeholders of planned flushing
Flush to 2.0 Nephelometric Turbidity Units
(NTU) turbidity at hydrants
Flushing Optimization Plans.
HDR Engineering, Inc. et al. (2006).
Flush velocity >3.5 ft/s to avoid particle
Inorganic Contaminant Accumulation in
deposition
Distribution Systems
Avoid damaging or dislodging tubercles from
AWWA. (In Publication). M68 Water
unlined cast iron pipes
Quality in Distribution Systems
Equip dead ends with flushing devices
targeting a velocity > 2.5 feet per second (ft/s)
Calibrate model using field data from many
points within the distribution system AWWA. (2011c). Partnership for Safe
Include minimum water quality parameters of
chlorine residual and water age
AWWA. (2012b). M32 Computer
Establish a model maintenance plan for
Modeling of Water Distribution Systems,
updating and calibrating the model
Third Edition
Update the model at least every 6 months
Model must include:
 Extended-period simulations
Hydraulic Modeling
 A minimum of 75 percent of pipe
volume and 50 percent of pipe length
 All pressure zones
 All 12-inch diameter and larger pipes USEPA. (2006). Stage 2 DBPR
 All diameter pipes which are
hydraulically significant
 All storage facilities with typical
operations represented
 All active pumping stations and all
active controls
arcadis.com 32

90th percentile lead ≤ 0.015 mg/L
90th percentile copper ≤ 1.3 mg/L
Establish a routine monitoring program for Friedman et al. (2010b). Criteria for
corrosion-related water quality parameters Optimized Distribution Systems
Regularly inspect internal condition of piping AWWA. (2011c). Partnership for Safe
for tuberculation, pitting, holes or scaling Water Self-Assessment Guide for
Internal Corrosion Develop written goals and SOPs for control of Distribution System Optimization
Control corrosion-related byproducts; water should
meet secondary standards for corrosion-
related parameters (i.e., 0.3 mg/L for iron)
100% samples for lead and copper below their
respective action levels (AL) (all points in the Kirmeyer et al. (2002). Guidance Manual
distribution system) for Monitoring Distribution System Water
Iron < 0.1 mg/L (all points in the distribution Quality
system)
Develop and maintain an asset inventory AWWA. (2006a). M17 Installation, Field
Develop a process and system to assess Testing and Maintenance of Fire
condition Hydrants, Fourth Edition
Develop a work order management process to AWWA. (2006c). M44 Distribution
track and prioritize maintenance Valves: Selection, Installation, Field
Testing and Maintenance, Second
Installation, Establish a planning process that includes
identified metrics for operability, a Edition
Operation and
Maintenance of replacement program to meet those metrics, Friedman et al. (2010b). Criteria for
Valves and Hydrants budgets, and coordination with other Optimized Distribution Systems
programs AWWA. (2011c). Partnership for Safe
Develop a communication plan to notify Water Self-Assessment Guide for
stakeholders of maintenance activities Distribution System Optimization
Includes design specifications for locations,

spacing and sizing of hydrants and valves
Implement a continuous monitoring system to
monitor flows and pressures at the major point
of entry, tanks, pump stations and other key
points in the distribution system Friedman et al. (2010b). Criteria for
Install online water quality monitoring at least Optimized Distribution Systems
Online Monitoring at the largest point of entry and at 3 AWWA. (2011c). Partnership for Safe
distribution locations for chlorine residual, pH, Water Self-Assessment Guide for
turbidity temperature and conductivity; monitor Distribution System Optimization
total organic carbon at a minimum of one
location (such as largest point of entry)
Install at least 5 pressure monitoring sites
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

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
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
Turbidity 0.5 – 1.0 NTU (2.0 NTU at flushed
hydrants)
Optimization Plans.
pH > 7.0 (unless State approved)
Smith. (2005). Water Distribution System
Calcium carbonate precipitation potential Assessment Workbook
(CCPP) within established guidelines
Develop and maintain an asset inventory
Establish a defined operation and
maintenance program
Pump Station Track and report the Planned Maintenance AWWA. (2011c). Partnership for Safe
Design, Operation Ratio (%) = 100 (hours planned maintenance) Water Self-Assessment Guide for
and Maintenance / (hours of planned + corrective maintenance); Distribution System Optimization
well-maintained systems exceed 80% for this
metric
Develop a motor and pump rehabilitation and
replacement program
AWWA. (2001). M19 Emergency
Planning for Water Utilities, Fourth
Develop an Emergency Response Plan (ERP) Edition
that meets Federal Emergency Management
Agency (FEMA) standards and conduct AWWA. (2010). J100-10 Risk and
regular exercises Resilience Management of Water and
Wastewater Systems
Conduct a J100 Vulnerability Assessment and
implement risk mitigation measures FEMA (2010). Developing and
Maintaining Emergency Operations
Security and Cultivate a security culture Plans: Comprehensive Preparedness
Emergency Implementation of cybersecurity measures Guide 101, Version 2.0
Management
Hardening of physical security AWWA (2011b). G440-11 Emergency
Develop robust internal and external Preparedness Practices
communication plans AWWA. (2011c). Partnership for Safe
Develop strategic partnerships Water Self-Assessment Guide for
Distribution System Optimization Ten
Provide standby power required to meet States Standards (2012).
average day demand
AWWA. (2014a). G430-14 Security
Practices for Operation and Management
arcadis.com 34

Develop SOPs:
 That define optimum water level
changes to ensure acceptable pressures
and flows and targeted turnover rates to
minimize water age
 To address water quality
 For routine inspections of storage tanks
Assess condition of tank and coatings, Optimized Distribution Systems
corrosion control and any safety, sanitary or
structural deficiencies annually
Provide critical security and Distribution System Optimization
telecommunications equipment
Monitor weekly at a minimum for disinfectant
residual
Conduct periodic sampling surveys to assess
residual uniformity
Storage Facility Track and control water age in the storage
Design, Operation facility
and Maintenance
Minimum chlorine residual of 0.5 mg/L and full Kirmeyer et al. (2002). Guidance Manual
turnover every 5 days for Maintaining Distribution System Water
Daily turnover of at least 20% Quality
AWWA) & Economic and Engineering

Storage tank inspection every 3-5 years Services, Inc. (2002). Finished Water
Storage Facilities
AWWA. (2011a). C652-11 Disinfection of
General guidelines for tank disinfection
Water-Storage Facilities
Detailed procedures for inspecting and AWWA. (2013b). M42 Steel Water
repairing steel tanks Storage Tanks, Revised Edition
Routine inspections daily/weekly
Periodic inspections monthly/quarterly
Comprehensive inspections every 3-5 years
Cleaning at least every 5 years
Disinfect after draining
Commit qualified staff and resources to review
distribution system and storage configurations
Establish surveillance testing schedules at key
locations in the system for water age
Implement proper maintenance practices (i.e.,
unidirectional flushing)
Water Age Distribution System Optimization
Management Train staff to recognize and response to
customer complaints resulting from water age
Water age < 5-7 days at all points in the Friedman et al. (2005). Development of
distribution system Distribution System Water Quality Plans
Reduce and establish target turnover rates
arcadis.com 35

AWWA. (2016b). M36 Water Audits and
Loss Control Programs, Fourth Edition
AWWA. (2006). M33 Flowmeters in
Establish internal water loss management Water Supply, Second Edition
team Friedman et al. (2010b). Criteria for
Perform annual water audits Optimized Distribution Systems
Improve audit data validity score AWWA. (2011c). Partnership for Safe
Determine present expenditures on water loss Distribution System Optimization
control
AWWA. (2012). M6 Water Meters –
Water Loss Control Determine and achieve economic levels of Selection, Installation, Testing and
water loss (apparent and real) Maintenance, Fifth Edition
Make business case for activities to achieve AWWA. (2014). M22 Sizing Water
economic levels of water loss Service Lines and Meters, Third Edition
Develop cost-effective water loss control plan AWWA. (2014). Policy Statement on
Select performance targets and conduct water Metering & Accountability
loss performance monitoring AWWA. (2015). G200-15 Distribution
AWWA. (2015 – 2016). Standards C700
– C750 for Water Meters
Develop a sampling plan and SOPs for
training, tap flushing, field measurements,
sample collection, sample storage and
transport and documentation
Perform monitoring at:
 Known problem locations and establish
trigger levels with responses (routine) Friedman et al. (2010b). Criteria for
Water Quality Optimized Distribution Systems
 Storage facilities (at least weekly) for
Sampling and AWWA. (2011c). Partnership for Safe
bacteria, disinfectant residual and other
Response Water Self-Assessment Guide for
appropriate parameters
 Chemical treatment points in the
distribution system (for process control)
 Upstream and downstream of main
breaks and repairs (non-routine)
 Customer complaint locations (non-
routine)
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.
3.2 Gap Analysis

The specific approach used to identify the gaps and opportunities for improvement for the City of Flint
included four components as illustrated in Figure 3-1. Each step is discussed in detail in the following
sections.
arcadis.com 36
Coordinate with
regulatory
agencies
Review avaialble
information
Conduct staff interviews

and site visits
Identify gaps and

opportunities for
improvement
Figure 3-1: Assessment and Gap Analysis Approach
3.2.1 Coordinate with Regulatory Agencies

The project kickoff workshop consisted of several presentations by the USEPA and MDEQ staff to update
the Arcadis project team on the status of the various ongoing efforts being performed by each
organization. A summary of the efforts is provided in Table 3-3. Follow-up conversations were conducted
with various USEPA and MDEQ staff as necessary to obtain additional information or clarification.
Table 3-3: Description of Key Distribution System Activities Performed by Regulatory Agencies
Topic Agency Description

Distribution system optimization activities focused on
maintaining chlorine residual (i.e., hydrant flushing,
Area-Wide Optimization Program USEPA
chlorine decay studies) based on USEPA AWOP
guidelines
Investigative chlorine sampling at businesses and
homes to assess distribution system chlorine residual
Chlorine Residual and DBP Monitoring USEPA
Recurring chlorine and DBP monitoring at 24
distribution locations
Home investigation of all lead sources offered if blood
Elevated Blood Lead Level Investigations MDEQ
lead level (BLL) > 5 µg/deciliter (dL)
Local field office consisting of MDEQ staff and trained
team of local residents providing filter training to
Filter Education MDEQ
customers and responding to inquiries on filters for lead
removal
Program designed to educate residents on fixtures (not
Fixtures MDEQ
yet launched as of September 2016)
Food Establishments MDEQ Implemented protocols for point-of-use devices
arcadis.com 37
Topic Agency Description

Two-week flushing campaign conducted in May 2016 to
increase usage, flush pipes and aid in water system re-
stabilization efforts
Residents received a credit on their water bill to
Flush for Flint MDEQ
encourage participation
Flushing instructions included a high flow 5-minute
bathtub flush followed by a high flow kitchen faucet
flush daily for 14 days
Installed 13 automatic flushers and 13 manual flushers
in distribution system to reduce water age and improve
water quality
Hydrant Flushing USEPA
Modified flow to flush at low, controlled rate
Routinely monitored chlorine residual and adjusted
settings as needed
Performed update and calibration of hydraulic model
Hydraulic Modeling USEPA
Ongoing efforts include water quality calibration
Constructed 4 pipe loop rigs, with lead pipe from 4
different homes – 2 from homes near “high” chlorine
residual areas and 2 from homes near “low” chlorine
Pipe Loop Testing USEPA residual areas.
Loops operated under on/off flow cycles with purchased
water from GLWA
Offers a free lead test kit and sample analysis to Flint
Residential Lead Sampling MDEQ residents
Collected over 23,000 samples to date
Performed complete evaluation of schools and day care
Schools and Day Care Facilities MDEQ
facilities, including sampling and fixture replacement
Lead testing performed at select sites around the City
Sentinel Testing MDEQ Includes 5 rounds of sentinel monitoring from March –
May 2016 and six rounds of extended sentinel
monitoring from June – December 2016
Performed sequential sampling (i.e., collection of 15-20
consecutive samples) every 2 months to assess
Sequential Lead Sampling USEPA corrosion control effectiveness
Includes 5 rounds of monitoring from January –
November 2016
Collected water quality samples throughout both
Storage Reservoir Sampling USEPA distribution system reservoirs to assess water quality
concerns (e.g., stagnation, dead zones)
Tested whole house filters at 5 residential homes
Whole House Filter Pilot MDEQ Limited success achieved as devices are not point of
use (POUs)
arcadis.com 38
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.
3.2.2 Review Available Information

An initial request for information was provided to the City to obtain available data, tools and
documentation necessary to assess distribution system performance, practices, procedures and studies
conducted to-date. Available information that has been review as part of this project is summarized in
Table 3-4.
Table 3-4: Summary of Information Reviewed
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
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
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
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
3.2.3 Conduct Staff Interviews and Site Visits

Members of the Arcadis project team spent three days onsite to interview select staff from various
divisions including the WTP, WSC, and/or other pertinent utility departments. Where possible,
management was interviewed separately from field staff as we have found that this approach allows for
the most candid discussions regarding practices that are actually implemented versus those that might
exist on paper. Staff interviewed included:
 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:
 Cedar Street Reservoir and Pump Station

 West Side Reservoir and Pump Station
 Dort Reservoir (located at the WTP)
 Control Station No. 2 (located at the WTP)
 Hydrant Flusher (located at the intersection of South Ballenger Highway & Miller Road)
 Water Treatment Plant Laboratory (located at the WTP)
 Chemical booster feed facilities (located at the WTP)
3.2.4 Identify Gaps

Information gathered from the steps listed above was used to assess the City’s existing distribution
system data, practices and procedures as compared to industry standards and optimization metrics (refer
to Section 3.1) to identify strengths, weaknesses and gaps in current practices as well as opportunities for
improvement. With the exception of the administrative topics and asset management (which will be
assessed in greater detail as part of future work on this project), results for each optimization topic were
categorized into one of the four qualitative designations shown in Figure 3-2. This categorization will help
to identify potential early action items and to prioritize elements for implementation, which will occur as
part of the development of the overall Optimization Plan.
arcadis.com 42
Areas that Meet

Areas that Need Areas that Need Areas where No
or Exceed
Minor Major Program or
Industry
Strengthening Strengthening Practice Exists
Standards
Figure 3-2: Gap Analysis Categories
arcadis.com 43
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 Disinfectant Residual and Disinfection Byproducts
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
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:
 Reducing water age

 Reducing free chlorine application at entry points and adding booster chlorination in the system to
target specific areas of low residual
 Flushing
More aggressive DBP control strategies may be required when DBPs exceed MCLs. These strategies
could include treatment improvements for NOM reduction, switching to chloramines, tank aeration, or
switching or blending source waters.
The AWWA Partnership for Safe Water and Criteria for Optimized Distribution Systems (AWWA, 2010)
provide the following criteria for distribution system chlorine residual:
 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
4.1.2 Data Sources
4.1.2.1 Chlorine Residual

Through 2016, the City monitored chlorine residual at the entry point (daily) and at ten sample collection
sites in the distribution system, including at the two main reservoirs. Historically, the City has monitored
as many as thirteen sites in the distribution system, including the two reservoirs. A minimum of 100
samples were collected and analyzed each month as part of routine regulatory coliform monitoring.
Beginning in 2017, this number of routine sites increased to 20 and each will be monitored a minimum of
five times each month. The City is targeting a goal of 25 routine coliform samples, so this program is
anticipated to expand in the near future. In the meantime, chlorine residual readings are being collected
from five additional surveillance monitoring locations each week to provide a combined total of 35 chlorine
readings from 25 unique locations each week.
While drinking water regulations require compliance with detectable chlorine residual concentrations
measured in routine coliform samples, other regulatory and non-regulatory sampling provides additional
distribution system chlorine residual measurements. Samples collected for DBP analysis related to the
Stage 2 DBPR are tested for chlorine residual to ensure concentrations do not exceed the MRDL for
chlorine of 4 mg/L. In March 2016, the USEPA initiated weekly monitoring of 24 system sample sites to
evaluate chlorine residual levels outside of the sites monitored for regulatory compliance. The 24 sites
monitored by the USEPA are also tested monthly for DBPs for investigative purposes.
arcadis.com 45
4.1.2.2 Disinfection Byproducts

As a Schedule 1 system (i.e., community water system serving 100,000 or more at the time the Stage 2
DBPR was established in 2006), the City was required to begin monitoring for DBPs under the Stage 2
DBPR in April 2012, when the compliance basis changed from a running annual average (RAA) of all
samples collected to a LRAA at sampling locations. Due to the City’s historically low levels of
trihalomethanes (THMs) and HAAs, it was eligible for reduced monitoring under the Stage 2 DBPR. For
the City’s population, this meant continued monitoring of its four sampling locations identified under the
Stage 1 DBPR. When the City switched sources to the Flint River, it was required to increase monitoring
to eight sample locations.
4.1.3 Data Analysis
4.1.3.1 Chlorine Residual

The following key objectives were developed in collaboration with the City to improve chlorine residual
management in the distribution system and to lay the foundation for an overall Chlorine Residual
Management Program (outlined in Appendix A):
 Describe process to manage chlorine residual

 Use a data-driven process to identify practical targets or set-points
 Develop seasonal recommendations and operational guidance for maintaining goals
The City’s current approach to chlorine residual management is to monitor free chlorine residual in
wholesale water purchased from GLWA and boost chlorine at the plant (system point of entry) and at the
two main storage reservoirs, West Side Reservoir and Cedar Street Reservoir. In late summer 2016, the
City began targeting a free chlorine residual concentration of 1.5 mg/L leaving the plant.
To evaluate the success of the City’s current chlorine residual management practices in achieving the
optimization criteria and identify gaps in the strategy, chlorine data collected at the plant entry point, in
routine coliform samples, and in sites monitored by the USEPA were reviewed.
The following figures show the 5th, 10th, 50th (median), 90th, and 95th percentiles of monthly system-wide
chlorine residual measurements for coliform (TCR) sites (Figure 4-2) and USEPA sites (Figure 4-3).
Monthly percentiles were calculated to compare with optimization goals. Using data from TCR sites alone,
the City has largely met the optimization criteria of 0.2 mg/L in 95% of samples collected while using
Detroit/GLWA water (Periods 1 and 3). During the Flint River operational period (April 25, 2014 to October
16, 2015), the City was unable to maintain 95% of residual measurements above 0.2 mg/L in most
months of the period, despite adding chlorine to 2 mg/L or greater entering the distribution system.
Review of USEPA sites, which include more challenging locations compared to TCR sites, indicates that
the optimization goal of 0.2 mg/L in 95% of monthly samples has been met since September of 2016.
Chlorine residual boosting began in January 2016 at the Cedar Street and Westside reservoirs and at the
plant in May 2016. Since September 2016, the City has maintained chlorine residual concentrations
above 0.5 mg/L in 95% of samples collected throughout the distribution system. The 0.5 mg/L level, while
greater than the Partnership target of 0.2 mg/L in 95% of samples, is viewed as appropriate for Flint
arcadis.com 46
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
Figure 4-2: Monthly Free Chlorine Residual at TCR Sites and System Entry Point (2013 – 2016)
arcadis.com 48
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
3.0
Flint River Operational Period
2.5
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)
Figure 4-4: Free Chlorine Concentrations at TCR Sites
3.0
2.5
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
3.0
2.5
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
3.0
2.5
2.0
1.5
1.0
0.5
0.0
Figure 4-7: Free Chlorine Residual at USEPA Sites (July – Nov. 2016)
4.1.3.2 Disinfection Byproducts

Figure 4-8 and Figure 4-9 present Stage 1 RAAs and Stage 2 LRAAs for TTHM and HAA5 for the
previous five years. Figure 4-10 and Figure 4-11 present TTHM and HAA5 concentrations for individual
samples (not running annual averages) collected from 2013 to 2016.
arcadis.com 52
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
Figure 4-8: TTHM LRAAs at Historical Compliance Sites (2010 - 2016)
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
MCL Stage 1 RAA
Figure 4-9: HAA5 LRAAs at Historical Compliance Sites (2010 - 2016)
arcadis.com 53
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
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
MCL
Figure 4-11: HAA5 Concentrations at Historical Compliance Sites (2013 - 2016)
arcadis.com 54
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
Figure 4-12: THM Concentrations at USEPA Sites (Mar. – Oct. 2016)
arcadis.com 55
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
Figure 4-13: HAA Concentrations at USEPA Sites (Mar. – Oct. 2016)
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
 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.
Areas where No Program or Practice Exists
 Verify and track chlorine dosing at storage facilities

 On-going tracking and trending of residual levels throughout distribution system
4.2 Main Breaks and Impacts of Pressure
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
Figure 4-14: Issues That Influence Frequency of Main Breaks
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.
4.2.2 Data Sources

City data on main breaks were reviewed to determine performance compared with recommendations
outlined by Partnership guidelines as described above.
A summary of data made available by the City for review for this assessment is presented in Figure 4-15
below. Information summarized includes a reference to Flint water system operations events and
distribution system operations data received. Water system operations of note include the period of
operation on Flint River Water supply (April 2014 through October 2015) and a period where water
system pump and valve operations were adjusted to minimize pressure spikes and depressions
(anecdotal information provided by City staff indicating operations adjustments made in spring/summer
2016).
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
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:
 2016 main break data could not be assessed, and

 Trends in main break occurrence due to system pressure management practices could not be
assessed (i.e., breaks occurring due to pressure variations could not be assessed independent of
Flint River Water operations).
4.2.3 Data Analysis
4.2.1.1 Spatial Analysis

Water main breaks were evaluated according to geographical location as shown in and Figure 4-16 and
Figure 4-17 below. This was completed to attempt to determine if breaks occurred more frequently in
various locations compared to others. Figure 4-16 illustrates all breaks from 2008 through 2015 plotted on
one system map, while Figure 4-17 provides location plots for each individual year.
These figures show the water system defined by five separate geographical areas based upon an attempt
to identify water system locations as follows:
 Near the Flint Water Treatment Plant (Northeast, green)

 After Flint Water Treatment Plant and north, not influenced by reservoirs (Northwest, blue)
 After Flint Water Treatment Plant and south, not influenced by reservoirs (Southeast, grey)
 Downstream of both Cedar Street and West Side Reservoirs (Center-west, light red), and
 Downstream of the Cedar Street Reservoir (Southwest, purple).
These areas have not been confirmed by hydraulic modeling of the system and should be refined. They
are discussed here as a first attempt in evaluating breaks per geography/potential operational influences.
Once the areas are properly defined, main breaks can be normalized to pipe miles for further analysis.
arcadis.com 59
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
Figure 4-17: Flint Water System Main Breaks Plotted Annually
arcadis.com 61
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.
Table 4-1: Summary of Main Breaks per Geographic Location
Note: St. Dev = Standard Deviation
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.
4.2.1.2 Annual Trends

The total breaks per year (listed in Table 4-1), are plotted below in Figure 4-18 to illustrate the trend from
2008 through 2015. The data is plotted with an estimated number of breaks for 2016 (based on
discussions with City staff). Without including 2014 and 2015 breaks, the actual trend indicates a
decrease in number over time. Inclusion of the 2014 and 2015 breaks change the trend to be increasing
arcadis.com 62
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.)
Figure 4-18: Main Breaks Per Year
4.2.1.3 Monthly Trends

A monthly trend of main breaks is provided below in Figure 4-19. This trend shows both the breaks and
the range between the monthly median of high and low daily temperatures (recorded at Bishop
International Airport) for each month. Plotted on the figure is the grey highlighted period of Flint River
operations from April 2014 through October 2015, and the monthly Partnership main break goals for both
an 800 and 580 mile water system (15 breaks / 100 miles / year / 12 months per year).
arcadis.com 63
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
Figure 4-19: Summary of Main Breaks Per Month
arcadis.com 64
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.
4.2.1.4 Impacts of System Pressure and Valve Operations

As referenced above in Figure 4-15, there is very little overlapping SCADA data available to attempt
identifying water system pressure, pump, or valve operations impacts on the occurrence of main breaks.
Furthermore, the reporting of main breaks is on a per-day without a specific time-stamp, making it difficult
to associate breaks with time-stamped pressure or operational events in SCADA.
Regardless, an attempt was made to determine if daily-recorded main breaks could be correlated with
events such as operation of the West Side Reservoir Fill Valve, Cedar Street Reservoir Fill Valve, and
Distribution System Entry Point Pressure (leaving the WTP). This correlation was investigated during
warm weather months of June through August 2015 when data was available, and air temperature was
not likely a controlling variable for main break occurrences.
arcadis.com 65
Breaks (No.) and Percent of Breaks (%)

arcadis.com 66

Figure 4-20: Breaks Per Month Characterized according to their Percentage of the Total Breaks Per Year
arcadis.com 67
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
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):
 57 main breaks occurred,

 West Side Reservoir fill valve operated (open and close) 81 different times,
 Cedar Street Reservoir fill valve operated 68 different times,
 There were 4,144 distribution system pressure readings near the WTP
o Maximum value = 91 psi (unclear to be true or a data collection error)
o 90th Percentile = 75 psi
o Average = 73 psi
o Minimum = 10 psi (unclear to be true or a data collection error)
 There were 117,064 Cedar Street Reservoir Header Pressure readings
o Maximum value = 81 psi
o Average = 64 psi
o Minimum = 24 psi
arcadis.com 69
Figure 4-22: Distribution System Pressure Plotted with Main Breaks (June – Aug. 2015; Flint WTP in operation over entire period)
arcadis.com 70
Figure 4-23: Cedar Street Reservoir Operations Plotted with Main Breaks (June – Aug. 2015; Flint WTP in operation over entire period)
arcadis.com 71
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:
 rapid open and rapid close,

 rapid open and slow close,
 slow open and rapid close, and
 slow open and slow close.
There were also fill events (not shown) that had rapid open and close and very short fill durations. These
types of operation were noticed at both West Side Reservoir and Cedar Street Reservoir and should be
corrected such that they (1) allow valves to always be opened/closed slowly, and (2) are documented into
a SOP such that valve operation practice is consistent across all staff.
arcadis.com 72
25
20 July 27, 2015

507 Minutes
2 Min ramp open
43 Min ramp close
15
10
June 26-27, 2015

544 Minutes June 18-19, 2015
30 Min ramp open June 12-13, 2015 188 Minutes
5 241 Minutes 17 Min ramp open
51 Min ramp close
1 Min ramp open 1 Min ramp close
1 Min ramp close
0
0 100 200 300 400 500 600
Time of Operation (min)
Figure 4-24: Selected Operation Summaries for West Side Reservoir Fill Valve
4.2.1.5 Main Repair and Replacement Practices

Interviews with staff indicated that responding to main breaks is exceedingly draining on resources. The
City budgets for approximately 250 breaks per year. City staff can make timely repairs to mains and
service lines, disinfect them in accordance with best practices, and restore them to service. A written SOP
was not available at the time of this writing, but during staff interviews the following procedures were
described:
 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
Table 4-2: Main Break Classifications (Source: Kirmeyer et al., 2014)
Main replacement activities tend to focused on:

• Rebuilding areas with high lead levels/lead service lines
• Areas with large number of breaks in viable neighborhoods
• Focus on transmission mains
• Prioritize based on historical break data and Echologics pipe wall assessment
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
 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,
 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
4.3 Pressure Management
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.
Partnership for Safe Water Numeric Goals:

 Minimum Pressure
o ≥ 20 psi in 99.5% of measurements (daily minimum from hourly readings)
arcadis.com 76
 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
4.3.2 Data Sources

Pressure readings are not available from the distribution system locations needed for comparison to Ten
States Standards or Partnership goals. Rather, some SCADA pressure data is available from the entry
point to the distribution system and from the West Side and Cedar Street Reservoirs as summarized in
the Main Breaks section. A brief summary of findings from analysis of available data and staff interviews
is provided below.
4.3.3 Data Analysis

Some findings that have resulted from interviews and information assessment include the following:
 Pressure control using elevated tank
o Evaluate need for modifications to tank altitude valve to minimize occurrence of pressure
transients.
o When system demands go up, tank levels drop, causing pressure problems
 Pressure transients are believed to be a root cause of many main breaks. Transients are
believed to be associated with pump capacity (relative to demand), pump operating
characteristics (constant or variable speed) and pump control valve operations (closing/opening
speeds), and valve operations (closing/opening speeds) at storage tank and possibly at ground
storage reservoirs. An assessment of potential impacts of system operations on main breaks was
provided in the Main Break section of the report.
 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.
 Use of one pressure zone for entire distribution system

o Portions of the single pressure zone require booster pumping to maintain adequate
pressure. Recommendations have been made to replace 4-inch mains where fire flows
are inadequate, or to parallel them with larger mains. Service lines would need to be
relocated if the 4-inch mains are replaced or abandoned. Evaluate a purely single zone,
if feasible.
arcadis.com 77
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.
 Not possible to determine at this time.
 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
 As far as is understood, no regular pressure monitoring and assessment is currently conducted

within the distribution system.
 It is possible that data from pressure data loggers could provide insight to distribution system
response to current operations.
 Evaluation of use of one pressure zone for entire distribution system
4.4 Internal Corrosion Control
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 Data Sources

Data from both MDEQ and USEPA were analyzed for this technical memorandum. Both agencies
maintained multiple data collection programs, each with their own sampling procedures, locations and
time periods. Descriptions of each of the datasets from these sampling are provided below.
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
4.4.2.1.1 LCR Compliance Dataset

The LCR Compliance dataset comprised of monthly/daily compliance data collected by the City of Flint as
part of their compliance program with the LCR. A total of 627 samples were collected from January 15 to
July 21, 2016 at 154 locations. Samples for this dataset were collected based on LCR sampling protocols.
4.4.2.1.2 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. Sample bottles and
instructions were distributed to residents with instructions to allow a 6-hour stagnation period before
sampling, sample from a faucet without a filter that is commonly used for drinking (not a hose bib or any
other), sample with cold water from the first draw, and to answer questions on the provided forms. There
are a total of 24,251 samples across 14,193 locations collected and analyzed between September 3,
2015 and September 15, 2016. Information that was collected includes a unique sample number; date
sample submitted; lead and copper concentrations (µg/L); as well as the street number, street name, city
and zip code where the sample was taken.
4.4.2.1.3 Sentinel Dataset

As part of their response to the high lead levels observed in Flint, MDEQ began a dedicated lead
sampling program known as the “Sentinel” program on February 16, 2016. The sampling sites were
established through coordination between MDEQ and Flint residents, and were reviewed by USEPA prior
to sampling. The sites include Tier One sites as defined in the LCR, meaning they are single family
residences containing either copper pipes with lead solder built after 1982 or contain lead pipes and/or
lead service lines (LSLs). The Sentinel sampling program encompassed 10 rounds total (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. Data were collected at
a total of 840 sites in the Flint system, although not all sites were included in every sampling round. The
original Sentinel sampling rounds averaged 630 sites, while on average the Extended Sentinel sampling
rounds included around 168 sites. In total, 3,988 lead data were collected through the Sentinel sampling
program.
During the sampling a variety of information was collected. The data set includes the following information
for each of the samples submitted:
 Sample site address

 Date the sample was submitted to MDEQ
 Lead and copper concentrations (µg/L)
 Service line material
 Location where sample was collected (e.g. kitchen, bathroom, etc.)
 Sentinel group number
MDEQ also assigned unique sample identification (ID) numbers and site codes to each sample and
sample collection location, respectively.
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
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
Sampling Round Sample Dates No. of Profiles No. of Sample Sites

Sequential 1 February 2016 94 94
Sequential 2 May 2015 42 41
Sequential 3 July 2016 47 47
Sequential 4 September 2016 51 51
Sequential 5 November 2016 45 45
Total 279 115
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
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.
4.4.2.3 Common Database

For this study, a comprehensive analysis has been 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.
It was also necessary to review the data for applicability. The focus of this investigation was on high-risk
sites known to have or thought to have lead service lines or lead-containing premise plumbing. MDEQ
and USEPA structured the Sentinel and Sequential sampling programs, respectively, to focus on high-risk
sites. Therefore, all of the locations from the USEPA Sequential and MDEQ Sentinel datasets were
arcadis.com 82
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.
4.4.2.4 Water Quality Parameters

In addition to lead sampling programs that were conducted by MDEQ and USEPA, the City of Flint
continued to collect operational data both at the plant and at 10 WQP monitoring sites in the distribution
system on a regular basis. Per the City’s MDEQ permit, it is required to monitor WQPs from 10 locations
weekly, and from 25 locations on a quarterly basis. Prior to January 1, 2017 data were collected from
each of the 10 WQP monitoring sites on a twice-weekly basis. After January 1, 2017, in conjunction with
the expansion of the RTCR program and sample scheduling changes, data have been collected at least
once each week from the 10 WQP locations. Quarterly monitoring at an additional 15 sites has been on-
going, resulting in a total of 25 quarterly WQP samples.
During each sampling event, Flint staff collected the following data:
 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.
4.4.3 Data Analysis

The analysis of the common database evaluated spatial and temporal distributions of maximum lead
values for the locations included in the dataset. However, this analysis includes some nuances that
should be explained prior to presenting data. As described above, the Residential and Sentinel datasets
collected by MDEQ include only one sample per sampling event, which consists of the first liter of water
collected after a stagnation period. The USEPA Sequential sampling data, on the other hand, collects
many different samples from each location for each sampling event. Figure 4-25 illustrates the lead data
collected during a typical sampling event included in the USEPA Sequential database. In this particular
case, the first sample collected, which was 250 mL in volume, had a lead concentration of 2.04 µg/L. This
first sample collected would be similar to the first liter collected in the MDEQ Residential and Sentinel
databases. However, in this case, the first sample does not represent the peak lead concentration
arcadis.com 83
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
Sample Volume (mL)

1,000
750
500
250
120 0
100
80
Pb Conc. (µg/L)
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
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.
4.4.3.1 Seasonal Lead Distribution

The distribution of maximum lead values for all of the 904 locations in the common database was
evaluated across four seasons: winter (Jan-Feb), spring (Mar- May), summer (Jun- Aug), and fall (Sept-
Oct). Although the distribution of maximum (peak) lead values is fairly similar for all four seasons, it
should be noted that the number of locations sampled during winter and spring 2016 exceeded
significantly than those sampled during summer and fall of 2016. Figures B-3 and B-4 in Appendix B
show the max lead distribution for each location by season for unpaired data. At first glance, it appears
that the distribution of maximum lead concentrations was similar between the first three seasons,
although there is a marked improvement in lead levels in Fall 2016.
However, the above analysis does not account for the change in sample population over the course of
2016. During Winter and Spring 2016, many more locations were sampled that during Summer and Fall.
To better understand the changes in lead levels at specific locations over time, a subset was created from
only those sites that were sampled during all four seasons. Table 4-4 summarizes the difference between
the sample populations of each dataset.
Table 4-4: Summary of Paired and Unpaired Locations by Season
Season Unpaired locations Paired locations

Winter 750 126
Spring 820 126
Summer 241 126
Fall 196 126
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
max lead distribution for each location by season for paired data. As these figures show, it appears that
lead levels are improving over time.
4.4.3.2 Flushing Event

On May 1, the City of Flint in association with USEPA and MDEQ initiated a 14-day flushing program
across the City. Residents and businesses were asked to flush the bathroom tub and kitchen faucet full-
on for five minutes a day. The State announced that the residents would not be charged for their water
usage in May to encourage maximum participation from people to move water through the system as part
of the City-wide flushing effort. It was also announced that the residents would be compensated for the
cost associated with flushing in addition to receiving water credits already awarded by the state. The main
objective behind this flushing program was to better distribute the treated water throughout the distribution
system and into premise plumbing. The City had increased their orthophosphate dose in December 2015,
but due to low water usage the phosphate-boosted water had not been distributed to many locations.
The impact of this flushing event was evaluated by grouping data from the common dataset into Pre- and
Post-Flushing datasets, using May 15, 2016 as the delineator. As with the seasonal data, there were
significantly more data available prior to the flushing event comparted to after the flushing event. Figure
B-7 shows that the maximum lead levels at each of the location before and after the flushing event are
significantly different. 90th percentile before and after the flushing event were 55.1 and 38.0 µg/L.
As described in the seasonal lead distribution analysis, to better understand the change in lead values at
specific locations over time, a subset was created from only those sites with lead samples collected both
prior to and after the May flushing event. Out of the 876 locations sampled prior to the flushing event, and
the 259 location sampled after the flushing event, there were 232 locations that were sampled both prior
to and after the flushing event.
As with the seasonal data analysis, excluding the unpaired data significantly impacted the distribution of
the data collected prior to the flushing event. Figure B-8 in Appendix B shows the max lead distribution for
each location by flushing event for paired data
Figure B-8 does not fully reflect what occurred prior to and after the May flush event at specific locations,
because it only represents the data as a percentile distribution. Although Figure B-8 does show that lead
levels across the system tended down after flushing, lead levels at specific locations may have increased,
decreased, or stayed the same. The impact at specific locations was evaluated by comparing a
percentile distribution of the percent change in lead concentrations at that location, which is shown in
Figure B-9 in Appendix B.
Figure B-9 shows the impact of the May flushing event on lead levels for the paired locations. Before the
May flushing event, lead levels for 46 percent of the paired locations exceeded the action level of 15 µg/L
whereas after the flushing event, 22 percent of the paired locations showed lead values higher than 15
µg/L. Around 62 percent (144 /232) of the paired locations that had lead values greater than 15 µg/L
showed a reduction/drop (< 15 µg/L) in the lead values after the flushing event. 9 percent (21/232) of the
locations that had lead values lower than 15 µg/L showed an increase (> 15 µg/L) in the lead results after
the flushing event. A total of 11 percent (26/232) locations showed no change in the lead levels after the
flushing event.
arcadis.com 86
4.4.3.3 Spatial Analysis

The lead data in the common database was classified spatially based on the zip-codes for each of the
sample locations. It was observed that majority of the sample locations were situated in the City center
(48503) and the area northwest of the City center (48504).
Figure B-10 shows the lead distribution before and after the flushing event for all of the zip-codes where
samples were collected. Figure B-11 illustrates the 90th percentile of the maximum lead values at each
location sampled in the zip code. The 90th percentile of samples collected in the City center, zip code
48502, did not exceed the action level with a 90th percentile value of 13.8 µg/L. Other locations saw
higher values. The 90th percentile of zip code 48503, which encircles the City center, was 53.7 µg/L.
Even higher 90th percentile values were observed to the northeast (zip code 48506, 90th percentile = 78.6
µg/L) and the northwest (zip code 48504, 90th percentile = 115 µg/L). The zip codes directly north, west,
and south of zip code 48503 had lower 90th percentile values, although all exceeded the action level of 15
µg/L.
4.4.3.4 Service Line Material

Service line materials significantly impact lead levels in drinking water, as the largest source of lead in
most drinking water systems are lead service lines. Service line material information was extracted from
the Sentinel and USEPA Sequential Plumbing datasets to be included in the common database. The
distributions of maximum lead levels from each location based on the service line material type was
evaluated seasonally. Both unpaired and paired data were evaluated. Paired dataset included locations
that had data for all four seasons. Figures B-12 through B-17 illustrate the results from this analysis. For
this specific analysis, lead levels from the USEPA dataset were based on the peak lead value for each
profile, not the first sample lead value.
As expected, it can be seen from figures B-12 through B-16 that lead concentrations for locations with
lead service lines tend to be higher than locations with copper or galvanized pipes. However, by Fall
2016, the difference in lead levels between locations with all three service line types is minimal. Figure B-
17 shows maximum lead levels at paired locations with either lead or galvanized pipes are decreasing
over time. Although the data set was relatively sparse, with only 34 locations served by lead or
galvanized service lines with paired data in all four seasons, amongst those 34 locations the 90th
percentile lead values fell from a high of 114.2 µg/L in Spring 2016 to 12.4 µg/L in Fall 2016.
4.4.3.5 USEPA Sequential Data Analysis

As discussed in section 4.4.2.2.1, the USEPA Sequential dataset included lead profile samples for all of
the 115 locations. While the analysis of the common database focused only on the lead results from the
first draw, this section discusses about the observations/trends in the lead results through the entire
profile at a particular location. The analyses discussed in this section/ report focus on evaluating the
profile volumes where the peak lead values occur, distribution of the occurrence of peak lead values,
occurrence of any high lead values (>15 or 50 µg/L) and location of the profile peaks relative to the
premise plumbing
Figure B-18 shows the profile volume associated with the peak lead values for all of the 234 profiles
across 115 locations. Since sample volumes were not consistent among the samples collected, sample
data were condensed based on the cumulative sample volume collected. For example, for Rounds 1, 3,
4, and 5, the samples collected within the first liter include samples 1 and 2, each of which were 125 mL.
In Figure B-18, the highest lead value between samples 1 and 2 for those rounds is included in the 1st L
arcadis.com 87
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
As discussed in section 1.2.2.2, USEPA provided detailed plumbing breakdown of 77 residences,

indicating all the plumbing pipes, fixtures, service line material, etc. from tap to the service line. With this
information and the peak lead analysis, the location of the profile peaks was evaluated with respect to the
premise plumbing. It involved comparison of volume of the premise plumbing to the sample volume where
the peak lead occurred. This helped in understanding the contribution of the premise plumbing in peak
lead values.
Premise plumbing was considered from the tap to the meter. All locations in this database had meters
located inside of house. Volume of plumbing inside the house was calculated using the pipe size and
length information from the Plumbing dataset. Figure B-24 shows the results of this analysis. It can be
seen that most profile peaks occurred inside of premise plumbing.
4.4.3.6 Water Quality Parameter Analysis

Preliminary analysis of the weekly WQP data was conducted. Figure B-25 shows the spatial distribution of
the WQP monitoring sites. Data review for this analysis was limited to only 2016 data. Particularly, results/
data for pH, total alkalinity and orthophosphate were evaluated using percentile distribution plots and
control charts. Control charting is a very useful tool to demonstrate performance of the distribution system
with respect to water quality control. It indicates if any existing system variance is due to some inherent
disturbance in the system or if a controllable external factor is responsible. Control charts are not used for
compliance purposes but are intended for process optimization and control.
Figure B-26 shows an example control chart developed for this analysis. This particular control chart was
generated from the available total alkalinity data collected at WQP monitoring site D6. The total alkalinity
data, which was collected weekly prior to November 2016 and bi-weekly afterwards, was grouped into 30-
day periods. The median of each group is plotted as a single point on the control chart. The median of
medians is then plotted as a green line on the chart, with the upper control limit (UCL) and the lower
control limit (LCL) as the red and purple lines, respectively. The red and purple lines are also referred to
as the upper and lower control limits, respectively. The differential between the upper and lower control
limits is a measure of how well-controlled the system is; if the differential is small, the system is well-
controlled and variance is limited, if the differential is large, the system is not well-controlled.
Control charts were generated for each of the three WQPs most relevant to internal corrosion control
(total alkalinity, pH, and orthophosphate) were generated for each WQP monitoring site. For presentation
purposes, the control charts were compressed to display the group medians form each location on a
single chart. Figures B-27 through B-29 show the compressed control charts for total alkalinity, pH, and
orthophosphate, respectively.
Reviewing Figures B-27 through B-29 provides some insight into the water quality throughout 2016.
Figure B-27 shows that total alkalinity was relatively stable within a ±4 mg/L as CaCO3 band, centered
around 71.5 mg/L as CaCO3. The results for pH, on the other hand, vary considerably but stay within the
generally acceptable (7.2 < pH < 7.8) range for orthophosphate corrosion control. Although the trend
varies over time, the pH values between the eight distribution system monitoring locations vary around
±0.1 s.u.
Orthophosphate data for all of the WQP locations varies between 3.0 – 3.8 mg/L as PO4 except for one
location, site D9/D13, which trended consistently lower than other stations (2.7 – 3.3 mg/L as PO4). Aside
from site D9/D13, the orthophosphate dose generally varied by less than 0.3 mg/L as PO4 between WQP
monitoring sites for any given period after February 2016.
arcadis.com 89
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.
All of 2016 Post-July 1, 2016
1.4
Differential between UCL and LCL
1.2
1
(mg/L as PO4)
0.8
0.6
0.4
0.2
0
WQP Monitoring Site
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
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
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
Figure 4-27: Current and potential WQP monitoring sites
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
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
 None identified.
4.5 Distribution System Practices

The following subsections provide an assessment of the City’s distribution system practices as compared
to industry best practices for optimization for each of the relevant improvement variables identified under
the Partnership program.
4.5.1 Cross-Connection Control
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.
4.5.1.2 Summary of Current Practices and Gaps

The City of Flint has a Cross-Connection Control Program (CCCP) currently being managed by the City’s
Plumbing and Mechanical Inspector. The January 2013 Water System Sanitary Survey prepared by
MDEQ indicated that the City was maintaining an adequate implementation of the Cross-Connection
Control Program. This report utilized data collected in 2012 to indicate that the City was meeting the
appropriate number of inspections and tests of high hazard and low hazard accounts as listed below:
• Annual cross-connection inspections of all 300 high hazard accounts
• Approximately 600 out of 2,000 low hazard accounts were inspected which falls within the
acceptable range of inspecting these accounts on a 3 to 5-year cycle.
• Approximately 75% of the 1,200 testable devices were tested for compliance with backflow
protection. AWWA M14 recommends that 100% of devices be tested for compliance on an
annual basis. However, MDEQ indicated that “based upon the survey, the City is meeting the 1-
year and 3-year objective for device testing”.
MDEQ also indicated that recordkeeping was adequate and a database of all backflow protection devices
was maintained based on the degree of hazard.
Since 2014, the staff dedicated to managing the CCCP has been reassigned to plumbing and mechanical
inspections which has reduced the ability to properly oversee the program. Clerical support has also
been reduced. In 2016, less than 50% of the testable devices were tested for compliance with backflow
protection. Also, the number of cross-connection inspections has been reduced such that they are only
performed during a scheduled plumbing/mechanical inspection.
In an effort to support the CCCP with limited staff, the Utilities Accountant is assisting the
plumbing/mechanical inspector with record keeping and forwarding reminder letters to customers that
have not yet completed the annual re-certification testing for their device(s).
arcadis.com 95
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
• 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 Customer Complaint Tracking and Response
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.

Anecdotal conversations with City staff indicate that there were many discolored water complaints when
treating the Flint River, but there are very few complaints now. Water Quality Laboratory (WQL) personnel
typically respond to customer complaints, and WSC personnel may be called to conduct flushing in
response to a water quality complaint. A site visit is made and a sample is collected from the tap at which
the complaint originated. Field analyses include pH, chlorine, conductivity, and temperature. Samples are
also brought back to the laboratory for coliform bacteria, iron, lead, copper, phosphate, turbidity, alkalinity,
hardness, and chloride analyses. According to Laboratory personnel, results are recorded in the system
on a complaint tracking form. However, these results were not available for review by the project team.
Thus, an assessment of the adequacy of the City’s complaint response and tracking program is not
possible at this time.
Based on the information provided by the City, the analyses conducted on each water quality complaint
sample are quite comprehensive. Additional data review is needed to assess the City’s customer
complaint tracking and response program, and to better understand the nature of complaints and follow-
up activities. It is unclear at this time if data from the past several years are available in a reviewable
format.
arcadis.com 97
 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.
 Development of Customer Complaint Response SOP.
 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
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.
Figure 4-28: Comparison of Conventional vs. Unidirectional Flushing

The City does not have an established UDF program. Historically, flushing consisted of running hydrants
fully open in the fall for winterization. In response to a recommendation from the Flint Drinking Water Task
Force, the City avoided its routine flushing during the fall of 2016 to allow the distribution system to
continue to stabilize and recover. Instead, the City is planning for a unidirectional flushing pilot program
arcadis.com 99
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:
 13 portable automated flushers

 13 manual flushers
 4 permanent automated flushers (purchased but not yet installed)
The auto flushers have been set to operate at a low flow rate of 20 gallons per minute (gpm) to meet the
objectives summarized above. Hydrant sampling devices have been constructed and are used to control
the flow rate while allowing for clean sample collection. The manual flushers have been used on
approximately 130 hydrants since the event. Chlorine residual grab samples are collected during flushing,
and the duration of the flush and/or location of flushing devices is modified accordingly. Photos of the
auto-flushers and manual flushing device are shown in Figure 4-29. Given water discharge and freezing
issues, flushing cannot be conducted during the winter months.
Figure 4-29: Automated flushing and manual flushing devices

(photos courtesy of Matt Alexander, USEPA Region 5)
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
o Specific flow rates have been targeted that are known to avoid sediment and biofilm
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.
 UDF Program. See recommendations above.
4.5.4 Hydraulic Modeling
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
 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).

In 2009, the City of Flint contracted with various professional services firms to develop and calibrate a
hydraulic model representing the City’s water distribution system. Since that time, the model was
periodically updated, but additional calibration was not performed. Recently the USEPA collected select
field data and the latest information available within the City’s geographical information system (GIS),
SCADA and customer billing systems to update and calibrate the model to represent current system
conditions.
The current model is setup as an EPS and calibrated to reflect a specific set of conditions experienced by
the system during October 2016. The model consists of a single scenario with one set of controls,
demands, infrastructure and settings. The model well-represents system hydraulics during October 2016
and is capable of simulating water age throughout the system. No verification of water age was
performed nor was there any comparison to actual water quality data to determine if water age
reasonably reflects actual system conditions. However, based on recent discussions with the USEPA,
additional updates to the model are currently in progress, including a water quality calibration.
It should also be noted that while the model does include physical attribute data for Dort Reservoir and
Pump Station No. 4, it does not include the correct operational parameters for these facilities. Additional
information would be needed to understand how these facilities would operate (either once the Flint WTP
comes back online or if improvements were made to allow use of this reservoir while receiving water from
GLWA), and modifications would need to be completed prior to performing any model simulations.
While the model does simulate conditions reasonably well, the model setup does not align with industry
practices. This is a result of the model being exported from a proprietary real-time modeling software into
a traditional software package, which is EPANET from the USEPA. This results in a cumbersome
process to utilize the model for evaluating various operating schemes and future infrastructure
modifications and improvements. A summary of the water distribution system model is provided in
Appendix C.
Areas That Meet or Exceed Industry Standards
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
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.
Areas That Require Minor Strengthening
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.
Areas That Require Major Strengthening
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
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
4.5.5 Installation, Operation and Maintenance of Valves and Hydrants
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.

Like many water utilities across the country, the City of Flint continues to struggle in finding a balance
between reactive and preventive maintenance – specifically related to appurtenances. The City has
made multiple efforts in the past to establish goals for annual valve exercising, hydrant assessments, and
replacement programs for appurtenances approaching 50 years of age (see Section 4.5.12 for discussion
of meter replacement efforts and needs). However, City resources often are shifted away from this
preventive work to respond to corrective issues, such as emergency breaks or construction activities.
Most recently, the services of Wachs Water Services were retained in 2015 to “catch-up” on valve
assessment and exercising. Information on each valve was collected and delivered in a GIS database,
as well as a list of inoperable valves (and other issues which require City follow-up). This information was
used to establish a budget within the 2015 Capital Improvement Plan (CIP) related to valve/hydrant
preventive maintenance and replacement. Some of these work orders have been completed, however
the City continues to struggle with adequate resources for timely follow-up and continues to get pulled
away on emergency issues. In summary, operational changes are needed to continue momentum
towards maintaining appurtenances consistent with industry standards.
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
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
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
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
Appurtenance Proposed Preventive

Size Count Age Proposed R&R Goal
Type Maintenance Goal
72in 20
Exercised/Assessed
24-72in 661 once per year (not 2-yr Identified
cycle)1 Repairs/Replacements to
Valves 16-20in 737
be completed during the
12in 1,398 Exercised/Assessed course of that year
every 3-5 years (not
< 12in 4,462 10-yr cycle)
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
Appurtenance Proposed Preventive

Size Count Age Proposed R&R Goal
Type Maintenance Goal
Identified
Repairs/Replacements to
Assessed 1-2 times per be completed during the
year (per manufacturer course of that year
Hydrants 3,605 requirements)2 Replace hydrants 50-years
Flow tests every 10 or older, then continue
years replacement hydrants to
achieve/maintain 25-year
life span
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:

 Leaks/breaks per 100 miles of pipe
 System Inspection (%)
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
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.
Table 4-6:Recommended Valve/Hydrant SOPs to be Developed
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
Other lower priority gaps were identified, such include:
 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
4.5.6 Online Monitoring
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.

The City’s existing OWQM instrumentation includes six free chlorine analyzers installed at the following
locations:
 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
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.
Areas that Meet or Exceed Current Standards
Chlorine residual analyzers have been installed upstream and downstream of chemical addition points to
provide dosing feedback.
Timely export of SCADA data into Grafana or similar visualization software.
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.
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 Pipeline Installation, Rehabilitation and Replacement
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
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.

Watermain installation and replacement projects within the City are designed by professional engineering
firms and installed by local contractors selected through a competitive public bidding process. The City
provides inspection of the Contractor during the installation and pressure testing of the new watermains.
They are also involved with the disinfection process and bacteriological testing. The City follows AWWA
C651 for disinfecting new watermains.
The City has developed technical specifications and standard construction details to provide consistency
of the new water system components and the installation practices from project to project.
The City has replaced an average of one mile of watermain per year from 1999 to 2011. However, no
watermain replacement projects have been completed since 2011. Industry best practices generally
recommend that utilities replace approximately 1% of their water system per year and maintain the age of
watermains at under 100 years. Table 3 in the 2016 MDEQ Asset Management Report indicates that
approximately 362.5 miles of watermain exist in the system with installation dates between 1900 and
1930. To comply with this practice by 2030 (maintaining an age of less than 100 years) would require
approximately 28 miles of water main to be replaced each year from 2017 to 2030.
AWWA Standard G200 recommends that utilities develop a prioritization strategy for the replacement
and/or rehabilitation of watermains with the distribution system. The strategy should incorporate concepts
of asset management to prioritize capital improvements based on risk by developing probability of failure
and consequence of failure estimates for individual pipe segments. The 2016 MDEQ Asset Management
Report developed a prioritized plan for replacing the aging watermains using watermain breaks and
neighborhood vacancy rates to establish risk but it is unclear if this plan was developed with input from
the City.
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
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.
Areas Where No Existing or Practice Exists
None identified.
4.5.8 Pump Station Design, Operation and Maintenance
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
 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

The City has indicated that the majority of the maintenance activities at the pump stations are reactionary.
Some preventative maintenance is performed approximately once per month including lubrication and
visual inspection of each pump. Other maintenance activities that do take place, but not on a regular,
recurring basis, include pump and motor bearing oil changes, checks and adjustment of packing seals,
and pump/motor vibration testing.
A preventative maintenance program is in place for the electrical switch gear and transformers.
A description of each pump station and current operating procedures are listed below.
 Cedar Street Pump Station (PS No. 3)

o This pumping station is equipped with three single-stage horizontal centrifugal pumps
which draw water from the adjacent 20 MG Cedar Street Reservoir during peak demand
periods or emergency conditions. The pumps were installed in 1948. The control valves
were replaced for pumps 2 and 3 while the original cone valve for Pump 1 remains in
service. These pumps draw water from the adjacent Cedar Street Reservoir.
o Pump operation and filling of the reservoir is controlled from the WTP or manually at the
pumping station. Typically, the adjacent reservoir is refilled during the night. Based on
direction from MDEQ, the City has revised their operating procedures for the West Side
and Cedar Street reservoirs to provide for a larger fluctuation in water levels. This is
described in detail in Section 4.5.10.2. This new operating procedure results in Cedar
Street Pump Station and West Side Pump Station alternating their operations every one
to two days. Typically only one station will operate while the reservoir at the other pump
station is filling. Once a reservoir is full, the corresponding pump station commences
operation while the other station is shutdown to refill its reservoir.
o A rechlorination facility is located at this site to boost the chlorine residual as it enters the
reservoir.
 West Side Pump Station (PS No. 5)
o This pump station consists of two 3-stage turbine pumps and two 2-stage turbine pumps
that are supplied by the West Side Reservoir. This station is operated similarly to the
Cedar Street Pump Station. Pumps 1 and 2 are rated at 4 MGD while pumps 3 and 4 are
rated at 8 MGD.
o Two of the pumps at this station have swing check valves. The City believes that these
swing check valves create a water hammer during startup and shut down of the pumps
which may be causing some of the water main breaks in the system.
o A rechlorination facility is located at this site to boost the chlorine residual as it enters the
reservoir.
 Torrey Road Pump Station
o The City has purchased new pumps, motors, and variable speed drives for this pump
station. Records indicate that the electrical equipment was installed but the pumps are
not yet in service. The station is equipped with two centrifugal pumps rated at 2.8 MGD
and 4 MGD. The large pump is used infrequently because it is oversized and over-
arcadis.com 114
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 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.
Areas that Meet or Exceed Industry Practices
None identified.
Areas that Require Minor Strengthening
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:
 Monthly inspections including equipment lubrication, vibration checks, bearing temperature

readings, packing adjustments, etc. A detailed list of all monthly inspection items should be
developed with input from the City maintenance personnel.
 Annual electrical system maintenance including variable speed drives, transformers, and motor
starters. Vibration testing should be performed on each pump and motor. Values should be
compared against previous years for developing trends and predicting future maintenance and/or
rehabilitation needs. Many utilities choose to hire qualified electricians or the original equipment
manufacturers for the electric equipment maintenance. If so, detailed specifications and
contractual requirements should be developed to provide consistency of the work from year to
year.
 Establish of formal long-term maintenance and rehabilitation program. This work typically occurs
every three to five years depending on past equipment performance and any previous equipment
arcadis.com 115
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.
Areas that Require Major Strengthening
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 Security and Emergency Management
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.

The City has several established security practices, including:
 All doors are locked and alarmed;

 Staff visit pump stations approximately twice per day; and,
 A portion of the distribution system currently has security camera coverage.
Additionally, the West Side Reservoir and Pump Station is entirely fenced in, although the Cedar Street
Reservoir and Pump Station is not because it backs up to a densely-wooded area. The Cedar Street
facility has experienced some issues with graffiti as a result.
While there are some existing physical security practices, several identified gaps in the City’s existing
physical security include the following:
 Operators are not always out in teams when visiting dangerous locations at night;
arcadis.com 116
 Additional lighting are desired; and,

 Additional security cameras are desired.
Some identified gaps in the City’s existing cyber security practices include that there is no existing firewall
for SCADA, although one has been requested, and that instruments outside the water plant are on a
wireless network with an unknown level of security. Finally, it is assumed that a vulnerability assessment
was conducted in 2003 as required by the Bioterrorism Act. It is also assumed that the vulnerability has
not been updated to meet the current industry standards.
For emergency management, the City has an existing ERP for the water supply system that was last
updated in January 2013 and is next slated for update in 2018. However, the existing ERP does not meet
current FEMA standards. Additionally, regular training and exercising of the ERP is not conducted as
recommended in current FEMA and industry standards.
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
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.
For the following areas, no program exists:

No existing cyber security or firewall for SCADA or instruments outside of the plant on a wireless network.
Use of firewalls is recommended in the Ten States Standards and AWWA Standard G430-14. In addition
to these standards, extensive technical resources addressing design and maintaining secure SCADA
networks. Two of the most prominent include the National Institute of Standards and Technology (NIST)
Special Publication 800-82 Guide to Industrial Control Systems Security (NIST, 2015) and the
International Society of Automation (ISA) ISA99 Security Guidelines and User Resources for Industrial
Automation and Control Systems, Fourth Edition (ISA, 2015).
No confirmed level of monitoring and security of most distribution systems assets, particularly pertinent
distribution system assets (e.g., critical valves or mains), as recommended in the Ten States Standard.
No confirmed existing all-hazards vulnerability assessment. An up-to-date all-hazards vulnerability
assessment is recommended by the Partnership, AWWA Standard G430-14, AWWA Standard G440-11,
and the AWWA Standard J100-10 All-Hazards Vulnerability Assessment (AWWA, 2010b). It is also
recommended that a vulnerability assessment be regularly updated. An all-hazards vulnerability
assessment entails the following steps: identification and prioritization of assets that are critical to the
utility achieving its mission, identification and characterization of threats and hazards relevant to the City's
assets and operations, determination of the risk presented to the City's assets by the relevant threats and
hazards, and development of risk mitigation measures to reduce risk to the City's assets and operations.
Additionally, a funding analysis is recommended to determine if federal or state funds are available to the
City for implementation of risk mitigation measures.
Finally, it is recommended that the City participate in the Genesee County Hazard Mitigation Plan revision
process to better understand regional risks that may affect City operations.
arcadis.com 118
4.5.10 Storage Facility Design, Operation and Maintenance
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

The City has established operating procedures with minimum operating levels in the storage facilities
based on pressure in the distribution system, fire flow requirements, emergency storage requirements,
and chlorine residuals. In previous years, the City had difficulty maintaining chlorine residuals within the
Cedar Street and West Side storage reservoirs. Based on direction from MDEQ, the operating
procedures were revised to provide more fluctuation of the reservoir levels. Current procedures dictate
that one reservoir fills while water within the other reservoir is pumped into the system for 1-2 days. Once
a reservoir is full it rests for 24 hours before the City reverses the procedure and begins pumping out of
the full reservoir for 1-2 days while the other reservoir begins filling. This procedure has increased the
chlorine residuals in Cedar Street and West Side storage reservoirs. It also appears to mitigate some
thermal stratification as evidenced by the reservoir sampling study conducted at Cedar Street reservoir on
October 27, 2016. However, this practice has the potential for negatively influencing distribution system
residual levels if water pumped out of one reservoir is pumped into the other. This will result in significant
water age increases and should be evaluated with the hydraulic model.
All treated water facilities are covered and protected from contamination. In addition, West Side and
Cedar Street reservoirs have rechlorination facilities to boost the residual as the water enters the storage
tank.
The City does not have a formal storage tank inspection program that includes procedures for routine
inspections (weekly or monthly), annual, and a comprehensive inspection every 3-5 years.
The following dates are the last known inspections for each storage tank:
o West Side: 2012
o Elevated Tank: 2011
o Dort: 2008
arcadis.com 119
o Cedar Street: 1997

o WTP Ground Storage: 1996
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
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:

o Visual inspection at ground level checking the foundation, signs of leakage, exterior
corrosion, vandalism, lighting, and ladders, platforms, tank overflow, manholes, and vents
if they are accessible for inspection without climbing safety equipment.
o Other utilities perform this type of routine inspection on a daily or weekly basis. With the
limited city staff currently available it is recommended to perform this inspection on a
monthly basis and increase the frequency of the inspections if staffing levels allows.
o This inspection is similar to the monthly inspections but includes areas not visible from
the ground. This investigation should also include a visual inspection of the storage tank
interior if there are accessible roof hatches.
o This work can be performed by city personnel if they have appropriate climbing and fall
protection training. Otherwise, this work should be contracted to a properly certified
contractor or professional services firm.
 Comprehensive Inspection – To be performed by a professional services contractor with support
from city personnel
o Every three to five years, storage tanks should be removed from service, drained,
cleaned, and inspected by a professional services contractor experienced in this type of
work.
arcadis.com 120
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 Water Age Management
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)
Problem Parameter/measure Potential causes

Reduction in disinfectant residual due to long
retention time
Total Coliforms
Poor quality source water treatment (high
Microbial water quality Fecal Coliforms AOC*/BDOC**)
HPCs Intrusion
Unlined cast-iron pipe
Long retention time
Chloramines
High levels of TOC in finished water
Loss of disinfectant residual Free chlorine
Excessive amounts of unlined cast-iron pipe
Total chlorine
Sediments
Long retention time
THMs
Disinfection byproducts Inadequate pH control
HAAs
Poor quality source water treatment
arcadis.com 121
Problem Parameter/measure Potential causes

Excessive retention time causing release of
Nitrate
ammonia
Ammonia
Nitrification Microbiological activity
Alkalinity
Poor quality source water
pH
High organic content in source water
Red water
Unlined cast-iron
Long retention time in unlined cast-iron pipe
Iron
Water chemistry characteristics (e.g. low
Manganese alkalinity)
Discolored water Turbidity Change of flow regime
Color Sediment
Consumer complaints Can be related to slow velocities (retention time)
that result in deposition of particulates.
Accumulated deposits can be swept up when
velocity increases
Inadequate chemical treatment (e.g., pH,
Lead carbonate, and orthophosphate)
Corrosion of metals Copper Long standing times in pipe
Iron Long retention time may reduce pH which could
increase corrosion
Chlorine residual
Phenolic compounds
Taste and odor Consumer complaints
Microbiological activity
Leaching from material
*AOC Assimilable organic carbon
**BDOC Biodegradable organic carbon
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:
 Changing the flow path of water by altering valves

 Installing time varying valves
 Manual flushing
 Automated flushing
 Abandoning mains
 Downsizing mains
 Increasing turnover of storage facilities
 Adjusting pump schedules
 Reducing the fill level within storage facilities or removing storage facilities, and
 Altering reservoir configurations to avoid dead zones
arcadis.com 122

Water age is a major factor in water quality deterioration within the distribution system. The City’s water
infrastructure was designed and built for a population nearly three times its current level. As a result,
maintenance of distributed water quality can be a challenge in areas with high water age. The recent
water crisis may have also exacerbated these problems in some areas, particularly in premise plumbing
of homes with little to no water use.
Water age cannot be easily monitored, and there is no “ideal” water age for all systems. Water age can
be estimated using hydraulic models, tracer studies, computational fluid dynamic modeling, and spatial
sampling (Walski 2003, Brandt 2004, Mahmood 2009). Water quality models can be used with hydraulic
models to predict concentrations of chlorine, DBPs, and other constituents in a distribution system
(Vasconcelos 1996, by Rossman 1994).
Effective water age management includes the following strategies:
1) Evaluation of system configuration and operations to improve system and storage turnover.
Calibrated hydraulic models are an effective tool for evaluation of system operation and
identification of strategies to improve water age. USEPA has improved the nearly all-pipes,
calibrated hydraulic model for the City. Future phases of this project include a storage analysis
and evaluation of operational strategies to reduce water age. More on hydraulic modeling and
storage facility design and operation is provided in sections 4.5.4 and 4.5.10, respectively.
2) Effective water quality monitoring. The City regularly conducts compliance monitoring and
expanded water quality parameter (EWQP) monitoring in the distribution system. Modifications to
those monitoring plans to comply with the RTCR and provide better surveillance monitoring of the
distribution system have been made as a part of this project. In addition, on-line water quality
monitoring is recommended to help the City respond proactively to potential water quality upsets
before they occur. More discussion on on-line monitoring (section 4.5.6) and water quality
sampling (section 0) is provided elsewhere in this document.
3) Water quality maintenance practices, such as flushing. The City utilizes 13 autoflushers and 13
manual flushers around the city to move water and reduce water age during times of year when
freezing is not a concern. The hydraulic model can be used to verify optimal locations for
installation of the autoflushers. The City has a reactive-maintenance flushing program and
conducts routine flushing during winterization of its hydrants. A more rigorous flushing program
(section 0) is being developed as a part of this project and will be provided in future phases.
Removal of biofilm and sediments that exert a chlorine demand can help reduce negative impacts
of increased water age.
4) Effective training program. The City currently lacks a distribution system operator training
program. Staff interviews will be conducted during the next phase of this project to determine
training needs, and a training program will be developed and implemented prior to completion of
the Optimization Plan.
 Use of a hydraulic model to understand areas of highest water age

 Use of automated flushing devices (seasonal)
arcadis.com 123
 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 Water Loss Control
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
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.

Prior to the collection of information requested by Echologics, Inc. relevant to a water loss audit
conducted after their condition assessment and leak detection investigations performed in 2015, Flint did
not perform a water loss audit in accordance with the AWWA M36 Methodology. Recommendations from
the Echologics investigation are included in the following discussion as being consistent with a
programmatic approach for achieving cost-effective water loss control described in Figure 4-30.
arcadis.com 125
Figure 4-30: Programmatic Approach for Cost-Effective Water Loss Control
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
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
 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
 Coordinate findings with the ongoing asset management program to

establish priorities for asset management and water loss control program
elements
o Evaluate pressure management
 Address root causes of pressure transients and determine the impact of
corrective measures, such as pump control valve modification, storage tank
altitude valve modification, modifying pump drives and operating curves to
better match system demand
 Evaluate the benefit of employing proactive pressure management in
portions of system on a diurnal basis
o Evaluate pipe/service line repair activities to optimize response and minimize runtime
for leakage occurrence
 Perform benefit/cost analysis of candidate activities and select activities for implementation
 Establish performance targets, emphasizing recommended financial and operational
performance indicators (while minimizing reliance on percentage-based operational
indicators) and monitor on an ongoing basis
 Benchmark activities against those of comparable water systems.
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 Water Quality Sampling and Response
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:
 Revised Total Coliform Rule

o Coliform bacteria
 Stage 2 Disinfection Byproduct Rule
o TTHM
o HAA5
o Chlorine residual
 Lead and Copper Rule
o Lead
o Copper
o WQPs
 Surface Water Treatment Rule
o Chlorine residual
arcadis.com 128
o Heterotrophic bacteria

The City’s monitoring and compliance status for chlorine residual, HPCs, and DBPs along with
comparison to optimization goals is described in Section 4.1. Recent modifications to the City’s RTCR
Plan are summarized below.
RTCR Plan Modifications:
Flint’s existing coliform monitoring plan was reviewed and updated to meet the following objectives:
 Update the City’s coliform monitoring plan to comply with and describe requirements of the
RTCR, which replaced the TCR in April 2016.
 Ensure representative routine coliform sampling in the City’s system by expanding the number of
routine samples collected. A sample site selection table was developed to summarize
characteristics of existing and proposed sites.
o Historically, the City monitored 10 routine sites three times per week.
o Representatives from the City, MDEQ, USEPA Region 5, and the Project Team visited
existing and candidate sites to aid in selection of the most representative and suitable
sites for on-going compliance monitoring.
o A total of 20 sites were selected that represent various characteristics of the distribution
system, including geographic location, water age, distribution main size, vulnerable
conditions, storage reservoir source, etc. Appropriate repeat sampling sites were also
identified.
o Six locations are being sampled five days per week, resulting in 30 coliform samples per
week from 20 unique locations.
o Five additional chlorine surveillance monitoring sites were added to the Plan, resulting in
collection of 35 chlorine residual measurements from 25 unique locations each week.
o The goal is to identify five additional RTCR sites, for a total of 25 RTCR/chlorine
monitoring locations. This work will continue once the expanded RTCR plan has been
successfully implemented. Ideally, these 25 sites can also serve as the City’s quarterly
WQP monitoring locations. Participants, variability and access challenge the city’s ability
to rapidly expand the number of monitoring sites.
 The City’s SOPs for coliform sampling and regulatory reporting were reviewed.
 Provide emergency response and public notification procedures in the event of a total coliform or
E. coli positive sample. Key contact information at the State and the City were updated.
 Changes to violations that may be issued due to MCL exceedance and failure to comply with
treatment technique, monitoring, or reporting requirements were summarized.
 Information for the completion of Level 1 and Level 2 Assessments as required by the RTCR was
provided.
 Enhanced water quality surveillance was implemented.
An updated RTCR Plan was submitted to the State in December 2016. Comments were received from
the State and incorporated into the Interim Final RTCR Monitoring Plan, February 2017, which is provided
as Appendix D. Additionally, a communications briefing was prepared at the City’s request for delivery to
sample location facility owners to describe the purpose of monitoring and how they can assist with
sampling efforts.
arcadis.com 129
Other Water Quality Monitoring Programs

An optimized water system conducts distribution system water quality monitoring above and beyond
regulatory compliance requirements to better understand water quality conditions, trends, causes of
customer complaints, need for maintenance and/or water age management activities, and to focus
resources where needed most.
Optimized systems will have Surveillance Water Quality Monitoring programs in place to meet the
following objectives:
 Develop baseline data to help City staff better understand prevailing conditions within the
 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
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.
 Streamline monitoring locations across monitoring plans.

 Ensure quality assurance plan (QAP) is up to date and captures most recent and approved
monitoring methods and sampling procedures.
 Conduct side-by-side monitoring of turbidity, iron, and apparent color for further streamlining of
surveillance monitoring program
 Electronic capture of monitoring results – lack of a Data Management System.

 Routine tracking and trending of key parameters and relationships (monthly, quarterly).
 Convert HPC analyses from SPC agar to R2A agar (7-day hold time at 20oC). A new incubator
may be required.
 Purchase new deionized (DI) water system to ensure quality of standards and analytical results.
 Improvements to the laboratory heating and cooling system are needed to better control humidity.
 No Data Management systems

 No trending of water quality results
arcadis.com 131
4.6 Supporting Utility Programs

The following subsections provide a high-level overview of the City’s existing administrative policies,
budget and financial planning, staffing levels and asset management program as outlined in the
Partnership program. While some gaps or areas for improvement have been identified, a more complete
assessment of these areas will be performed and included in the final Distribution System Optimization
Plan.
4.6.1 Administration
4.6.1.1 Administrative Policies

The City of Flint Water Department is faced with multiple challenges that compete for limited resources.
For a utility to strive for optimized performance, there needs to be a commitment to excellence at all levels
of the organization. This commitment must be based on an understanding of the importance of
maintaining distribution system integrity to protect public health and enhance service continuity. For the
City of Flint, administrators should first support policies and provide the resources needed to improve
system performance to ensure both regulatory compliance and customer acceptance of the delivered
water, and, ultimately, pursue optimized system operation.
Water Department administrators and supervisors must also understand that to reduce the potential for
exposure of consumers to contaminants in their drinking water, all system components must be working
together on a continuous basis. Accordingly, the Water Department administration should develop goals
for providing high quality water through proper balance of treatment and O&M of the distribution system.
The importance of achieving these goals needs to be emphasized to the operating staff.
Policy should support goals of low cost, safe working conditions, good system performance, and high
employee morale. However, administrative policies should also focus both on water quality and system
reliability. Utility administrators also need to be aware of the impact that their policies have on system
performance. The WSC distribution operations staff may need to educate administrators, who may not be
technically trained, on the importance of both maintaining water quality in the distribution system and
system reliability through timely replacement of failing water mains, investment of pressure control
strategies, implementing proactive flushing programs, and other measures. The WSC currently does not
have access to the SCADA system, so it is difficult to react to changes in demand, reservoir levels, pump
station operation, etc.
Another area in which administrators and policy can significantly affect performance is through personnel
motivation. Performance can be enhanced if administrators encourage personal and professional growth
through support of training; involvement in professional organizations; and providing tangible awards for
higher certification levels. If, however, administrators limit operator training, offer non-competitive salaries,
or otherwise undermine staff morale, this can result in sub-optimum performance. A comprehensive
training program will be developed as a part of the final Distribution System Optimization Plan.
Policy and utility management needs to discourage complacency. A strong commitment to excellence can
be the best defense against complacency. This requires the involvement of all personnel to create an
empowered, knowledgeable staff that can effectively respond to changing conditions.
arcadis.com 132
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
Table 4-8: Water Department Staffing Summary
Position Description Budgeted Filled Vacant
Water Service Center Distribution
Account Clerk 1 1 0
Building Maintainer 1 0 1
Deputy Supervisor 1 1 0
Senior Water Distribution Operator 13 13 0
Water Distribution Foreman 5 4 1
Water Distribution Operator 11 6 5
Water Distribution Operator Trainee 6 1 5
Water Service Center Supervisor 1 1 0
TOTAL 39 27 12
Water Treatment Plant
Principal Clerk Typist Utilities 1 1 0
Utilities Instrumentation Tech - Water Plant 1 1 0
Water Plant Laboratory Technician 2 2 0
Water Plant Maintenance Supervisor 2 2 0
Water Plant Operations Foreman 2 2 0
Water Plant Operations Foreman (Provisional) 2 2 0
Water Plant Operations Supervisor Provisional 1 0 1
Water Plant Operator Maintainer 0 0 0
Water Plant Operator Trainee 6 3 3
Water Plant Senior Operator/Maintainer 5 5 0
Water Plant Supervisor 1 1 0
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
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.
Table 4-9:Comparison of Survey Results to Current Flint Staffing and Training
Customer Accounts per MGD Delivered per Training Hours per

Employee Employee Employee
25th Percentile 400 0.17 12
Median 500 0.25 20
75th Percentile 700 0.36 32
Flint (Budgeted/Actual) 808/1167 0.33/0.48 0/0
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
Figure 4-31: Utilities Department Organization Chart (May 2016)
arcadis.com 136
4.6.2 Asset Management

Asset management is the basis by which operations and maintenance decisions are made, rehabilitation
and replacement efforts are prioritized, and capital and operations budgets are developed and managed.
The Partnership for Safe Water outlines an approach to asset management in line with the USEPA “Asset
Management: A Best Practices Guide” published in 2008. By this analysis method, utilities track assets
on an up to date distribution system mapping software, maintain inventory of distribution system assets,
maintain maintenance records of procedures and observations, and develop a comprehensive asset
management program.
In 2016, the City took its first steps towards implementation of an asset management program with the
development of an asset management plan, which aimed to develop an inventory and identify gaps for
the City’s roadways and water and wastewater systems. The report, however, was reviewed by MDEQ
and found to be lacking several key items of an “approvable” asset management plan, including, but not
limited to:
 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
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
5.0 SUMMARY AND IMMEDIATE RECOMMENDATIONS

This section presents a summary of the approach and results of the gap analysis and recommends early
action items for immediate implementation.
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
Table 5-1: Areas that Meet or Exceed Industry Standards

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. As of October 2016, the City is also meeting a more rigorous goal
of ≥ 0.5 mg/L in 95% of samples collected on a monthly basis.
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.
Disinfection of repaired mains and placement into service appears to be conducted in
Main Breaks
There is sufficient (possibly excessive) redundancy in the system, so main breaks can
be isolated despite valve problems, while still providing service to customers.
Pressure Management Not possible to determine at this time.
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
Internal Corrosion
systems in the U.S.
Control
WQP monitoring for an extended list of parameters is currently conducted at least
once per week at 10 locations.
Cross-Connection The existing cross-connection control program appears to align with AWWA Manual of
Control Practice M14.
Detailed analyses are conducted on water quality samples collected from complaint
Customer Complaint
locations. The analyses provide insight into metals, aesthetic, and microbial
Tracking and Response
conditions.
The City’s conventional and dead-end flushing practices currently used to manage
water age meet and exceed industry standards.
Specific flow rates have been targeted that are known to avoid sediment and biofilm
Flushing
Numerous programmable auto-flushers are used and can be rotated around the City
as needed to meet water quality objectives.
A hydrant sampling device has been created and is used to control flow rates and
collect sanitary water quality samples
The model has been setup to perform real-time modeling, although it is not directly
Hydraulic Modelling connected to the City’s SCADA system at this time. The real-time modeling method to
achieve calibration for a specific period of time is above expectations.
The City has made significant investment progress on the valve assessment and
exercising program. This progress demonstrates the commitment of City resources to
Installation, Operation, ensure the information flows from the office to the field and that current business
and Maintenance of processes do allow the completed work to be reported up to departmental
Valves and Hydrants management.
The fire hydrants within the City of Flint have been able to meet the needs of
emergency services.
Chlorine residual analyzers have been installed upstream and downstream of
Online Monitoring
chemical addition points to provide dosing feedback.
arcadis.com 140

Disinfection of new mains and placement into service appears to be conducted in
Rehabilitation, and
Replacement
Maintenance
The City meets industry standards according to the Ten States Standards requiring all
Security and Emergency exterior entrances to lock, fencing around facilities (with the notable exception of the
Management Cedar Street Pump Station and Reservoir, which is not entirely enclosed in fencing),
intrusion alarms, and some security camera monitoring.
Storage Facility Design, Recent modifications to improve turnover, conduct monitoring with depth to verify
Operation, and conditions, and to improve controls of boosting operations are positive operational
Maintenance steps that meet industry standards.
Use of a hydraulic model to understand areas of highest water age.
Use of automated flushing devices (seasonal).
RTCR Plan and chlorine surveillance monitoring plans meet industry standards
and Response EWQP monitoring exceeds industry standards in terms of monitoring frequency and
parameters included.
Table 5-2: Areas that Need Minor Strengthening

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).
Use hydraulic model to identify potential areas of system with low chlorine residuals
Disinfectant Residual and conduct surveillance monitoring and site-specific improvements at these locations.
Identify optimal chlorine dosing to target 0.2- 0.5 mg/L in 95% of sample locations
throughout distribution system.
Identify seasonal optimal boosting strategies at plant and reservoirs.
Continue flushing in areas where maintaining a detectable chlorine residual has been
challenging and identify opportunities to improve residual levels without further
increasing dosages.
Update the procedure for documenting and responding to main breaks.
Additional information regarding the approximate time of the break, Type
categorization, visual description (shear break, longitudinal, hole) should be captured.
Main Breaks 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 evaluation 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.
arcadis.com 141

Newly added pressure monitoring stations and site selection for portable data loggers
need to be confirmed by the City’s hydraulic model and incorporated into the City’s
Pressure Management
SCADA. Additional recommendations for improved pressure control may be
developed once results from distribution system data loggers are available for review.
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. High orthophosphate doses
could contribute to formation of white precipitate (calcium phosphate) in the
Internal Corrosion distribution system.
Control
The orthophosphate level at Site D9/D13 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.
Analyze recorded data 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.
Customer Complaint
Complaints should be entered into GIS or a CMMS (once implemented) for better
Tracking and Response
tracking/analysis
Compare current complaint rates to industry benchmark of no more than 2.5 technical
water quality complaints per 1,000 customers.
Cross-Connection
None identified.
Control
Flowrates during conventional and auto-flushing can likely be increased after field
Flushing verification. 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 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
Hydraulic Modelling
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.
Demand allocation and representation within the model is unintuitive and does not
align with industry standards.
Continue efforts related to obtaining GPS coordinates for all appurtenances.
Enhance the GIS database schema to include other valve and hydrant attributes,
types, and functions.
Establish a comprehensive data model review to consolidate information into a single
Installation, Operation,
source.
and Maintenance of
Valves and Hydrants Consider modifying existing asset IDs for consistency with industry standards.
Strengthening may be needed to ensure asset information is accessible to the specific
person who is going to locate, access, and operate the asset in a reasonable amount
of time (including historical maintenance records and reliability information, currently in
paper and excel).
Online Monitoring Timely export of SCADA data into Grafana or similar visualization software.
arcadis.com 142

Review existing technical specifications and construction details and update as
needed to current standards based on the Ten States Standards and AWWA.
Develop an established schedule for updating and maintaining technical specifications
and standard construction details for new water main installations and replacements in
Rehabilitation, and
accordance with AWWA Standard G200.
Replacement
Develop an approval procedure for verifying the installation was completed according
to these newly developed City standards, and that pressure testing and disinfection
was in compliance with the appropriate regulatory agencies.
Install globe body style control valves at West Side Reservoir to reduce the potential
for pressure surges during startup and shutdown of the pumps.
Develop written SOPs 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.
Pump Station Design, Develop detailed maintenance procedures for each pump station defining the
Operation, and frequency, procedures, and maintenance of records for each piece of mechanical and
Maintenance electrical equipment. Expand any existing SOPs with additional items for
consideration listed below:
Monthly inspections including equipment lubrication, vibration checks, bearing
temperature readings, packing adjustments, etc.
Annual electrical system maintenance including variable speed drives, transformers,
and motor starters.
Establish of formal long-term maintenance and rehabilitation program.
Increase control and intrusion detection per guidance in the AWWA Standard G430-
14.
Security and Emergency Perform annual inspections of identified critical assets.
Management Completely enclose the Cedar Street Reservoir site in fencing, per guidance in the
Ten States Standard and AWWA Standard G430-14. Additionally, the fence should be
maintained clear of plant growth.
Storage Facility Design, Replace inlet valves to West Side and Cedar Street Reservoirs with globe body style
Operation, and altitude valves having back pressure sustaining pilots. The opening and closing rate
Maintenance of these valves can be adjusted to minimize the potential for pressure surges.
Use hydraulic model to investigate optimal storage reservoir operations to reduce
detention time and avoid sequential draining/filling.
Manual flushing practices must be strategically located to turnover water in the highest
age areas.
Streamline monitoring locations across monitoring plans.
Ensure that the current QAP is up to date and captures most recent and approved
monitoring methods and sampling procedures.
and Response
Conduct side-by-side monitoring of turbidity, iron, and apparent color for further
streamlining of surveillance monitoring program.
arcadis.com 143
Table 5-3: Areas that Need Major Strengthening

Continue efforts and approach outlined in Chlorine Residual Management Program to
optimize residual levels while reducing dosages and minimizing variability.
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 immediately.
Standard operating procedures for pump start/stops and valve operations are required
Main Breaks 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,
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. Initiate a regular leak detection program.
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 in SCADA and operate the
Pressure system to achieve Partnership pressure goals.
Management Conduct surge modelling to optimize 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.
Create standard procedures for valve open and close operations.
Lead levels at the higher end of the percentile curve are still excessive and do not
Internal Corrosion
currently meet the performance target recommended in Kirmeyer et al. (2002) of 100
Control
percent of the samples for lead below the action level of 15 µg/L.
Staff levels need to be increased to improve the management and supervision of the
Cross-Connection
cross-connection control program and maintain re-certification testing of the 1,000
Control
backflow prevention devices.
Customer Complaint
Tracking and Develop a Customer Complaint Response SOP.
Response
The City should assess the effectiveness of UDF for reducing pipe wall chlorine demands
and removing accumulated sediment from the distribution system by conducting a UDF
Flushing pilot study.
Estimated level of effort for developing and implementing a UDF program will be prepared
as part of the Resource Analysis and Needs Assessment.
Facilities such as pumps should be confirmed and better represented within the
distribution system model using available pump curves, modified as needed to reflect
current operation.
Hydraulic Modelling
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.
arcadis.com 144

The Valve Exercising and the Hydrant Replacement Program is currently prioritized in the
CIP as level 2 (Important) in the 2016 Capital Improvement Plan and should be re-
evaluated to consider the feasibility to moving to a level 1 (Urgent).
Installation, It is recommended that the City consider departmental organization and other options to
Operation, and optimize resources.
Maintenance of Establish a documented plan for valve assessment and exercising and hydrant
Valves and Hydrants assessment and replacement in accordance with AWWA Standard G200.
Re-evaluate risk criteria used in previous studies to identify critical pipeline segments and
apply to appurtenances (current CIP identifies critical valves based solely on size).
Incorporate additional scoring criteria based on a triple-bottom line approach.
Online Monitoring None identified.
Develop a formal risk-based approach as the foundation of their watermain
replacement/rehabilitation program in accordance with AWWA Standard G200.
Replace a minimum of 1% of watermains per year with a goal of maintaining the age of all
piping within the system to less than 100 years.
Develop a program for assessing the physical condition of the pipes including the
following:
 A GIS for the distribution system that includes pipe attribute information such as
Pipeline Installation, age, diameter, material, length, and location
Rehabilitation, and  Maintenance records for all leaks and breaks. Data to be recorded as part of a
Replacement repair include diameter, material, pipe class, interior/exterior coatings or linings,
repair type, joint type, soil conditions, and GPS coordinates. This data should
then be incorporated into the City’s GIS.
 Surrounding environmental information at the leak site such as soil type,
corrosion potential, and potentially hazardous material identified during the
excavation.
 Distribution system data such as fire-flow tests, c-factor tests, and pressure data
that can be incorporated into the City’s GIS.
Maintenance
Update the existing ERP to meet FEMA standards.
Regular exercise of the ERP requires major strengthening and is recommended in the
Partnership for Safe Water Self-Assessment Guide, AWWA Standard G430-14, and the
Security and Development and promote a security culture. This includes commitment from senior
Emergency leadership and promotion of security awareness, as recommended in AWWA Standards
Management G430-14 and G440-11.
Development and strengthen communications and partnerships to enable Flint to best
respond in emergency situations as recommended in AWWA Standards G430-14 and
G440-11. It is also recommended that the City develop mutual aid and assistance
partnerships.
Per AWWA Standard G200 and Manual of Practice M42, establish formal procedures for
Storage Facility the inspections of each storage facility. These inspections should focus on structural,
Design, Operation, safety, security, coatings system, and sanitary conditions. It is recommended that the
and Maintenance following timeframes be established for the inspections:
arcadis.com 145

 Comprehensive Inspection – To be performed by a professional services
contractor with support from city personnel
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.
Water Age Review the overall system water demands, storage needs and capacity of their existing
Management storage facilities.
Perform annual water audits in accordance with the AWWA Manual of Practice M36
methodology.
Develop a Water Loss Control Plan that aligns with the Water Department’s
organizational goals and objectives, and adjust, as appropriate, going forward. Key
activities needed for the plan development include:
 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 level needed to establish performance targets for activities comprising Flint’s
Water Loss Control Plan.
 Perform an annual meter accuracy determination of all distribution system entry
point flow meters, including volumetric test and calibration of associated
instrumentation and apply errors in the annual water audit.
 Verify the presence of all customer connections, meter readings, and billed
Water Loss Control metered consumption in the Read-to-Bill System.
 Quantify all types of authorized consumption (billed metered, billed unmetered,
unbilled metered, unbilled unmetered) and correlate with SCADA-reported flow.
 Manage apparent loss control through determining the accuracy of residential
and commercial meters and maintaining/replacing as necessary. Evaluate
sources of unauthorized consumption and systematic data handling errors.
 Control real losses by accounting for storage leakage, main, service line, and
appurtenance leakage, and coordinating findings with an ongoing asset
management program. Evaluate pressure management and pipe/service line
repair activities to minimize leakage.
 Perform cost:benefit analysis of candidate activities and select activities for
implementation
 Establish performance targets, emphasizing recommended financial and
operational performance indicators (while minimizing reliance on percentage-
based operational indicators) and monitor on an ongoing basis
 Benchmark activities against those of comparable water systems and update as
appropriate.
Electronic capture of monitoring results – lack of a data management system.
Routine tracking and trending of key parameters and relationships (monthly, quarterly).
Water Quality
Sampling and Convert HPC analyses from SPC agar to R2A agar (7-day hold time at 20oC). A new
Response incubator may be required. Conduct side-by-side monitoring for a few months to assist
with interpretation of R2A results.
Purchase new DI water system to ensure quality of standards and analytical results.
arcadis.com 146
Table 5-4: Areas Where No Program Exists

Verify and track chlorine dosing/chemical usage at storage facilities.
On-going tracking and trending of residual levels throughout distribution system.
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.
Main Breaks
There is a lack of city vehicles to efficiently respond to and/or complete maintenance.
There are no GPS coordinates of pipes and valves, and no properly updated water
system map.
As far as is understood, no regular pressure monitoring and assessment is currently
conducted within the distribution system.
Data from pressure data loggers is needed to provide insight to distribution system
Pressure response to current operations and to compare to Partnership pressure management
Management goals.
Surge modelling to verify and optimize WTP, storage, and pump station operations.
Evaluation of use of one pressure zone for entire distribution system
Internal Corrosion
None identified.
Control
Implement a cross-connection control plan on the residential level in accordance with
Cross-Connection AWWA Manual of Practice M14.
Control Establish a formal backflow incident response plan including tracking and investigation of
backflow incidents.
Customer Complaint
Develop a SOP for customer complaint response and tracking. Include a Communications
Tracking and
Plan between the WQL and WSC.
Response
Flushing Establish a UDF Program.
The City has not developed a model maintenance plan, or regular schedule for updating
the model components and performing periodic calibration.
Training for City staff should include basic model usage, model capabilities, modeling
goals, location of input/output data, instructions on how to export model results for
Hydraulic Modelling
viewing, analysis of various fictitious standard model simulations, and demonstration of
A surge model is needed to assess and prevent pressure transients caused by
operational activities.
arcadis.com 147

Need for documented current or future 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:
Installation,  Leaks/breaks per 100 miles of pipe
Operation, and  System Inspection (%)
Maintenance of
Develop SOPs including valve/hydrant installation and procedures, valve/hydrant
Valves and Hydrants
preventative maintenance program and procedures, valve emergency and corrective
maintenance procedures, work order documentation protocols, and community fire risk
assessment.
Confirm manufacturer manuals are available or used to establish/verify preventive
maintenance requirements.
Need for hydrant/valve coordination/training with other departments/agencies (fire
department, traffic accident reporting, street cleaning, sewer cleaning, building
construction, recreation, contractors, etc.)
Public notification of a potential impact to service during planned or unplanned
maintenance
Incorporation of hydrant testing results into the hydraulic model calibration
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,
Online Monitoring
 Five distribution system pressure monitoring sites
 TOC monitoring at the largest POE and at one point within the system.
Apart from 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.
Rehabilitation, and None identified.
Replacement
Pump Station Design, Conduct a needs assessment at each pump station to identify short and long-term
Operation, and capacity requirements as well as other capital needs. With the downward trend in water
Maintenance consumption within the City, smaller capacity pumps may be necessary.
arcadis.com 148

No existing cyber security or firewall for SCADA or instruments outside of the plant on a
wireless network. Use of firewalls is recommended in the Ten States Standard and
No confirmed level of monitoring and security of most distribution systems assets,
particularly pertinent distribution system assets (e.g., critical valves or mains), as
recommended in the Ten States Standard.
Security and
Emergency No confirmed existing all-hazards vulnerability assessment. An up-to-date all-hazards
Management vulnerability assessment is recommended by the Partnership, AWWA Standard G430-14,
AWWA Standard G440-11, and the AWWA Standard J100-10. It is also recommended
that a vulnerability assessment be regularly updated. Additionally, a funding analysis is
recommended to determine if federal or state funds are available to the City for
implementation of risk mitigation measures.
It is recommended that the City participate in the Genesee County Hazard Mitigation Plan
revision process to better understand regional risks that may affect City operations.
Storage Facility Detailed hydraulic analysis of storage facility capacity and optimal operations, considering
Design, Operation, decreased system demand, need for adequate capacity during main breaks, and planned
and Maintenance operational changes associated with switch to KWA and operation of Flint WTP.
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,
Water Age commercial, industrial), and develop a plan for system operation as demand decline
Management which maintains adequate water age.
Need for a UDF program.
Tracer evaluations to confirm hydraulic model predictions of water age.
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.
Water Loss Control
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.
Water Quality No data management system.
Sampling and
Response No trending of water quality results.
5.2 Immediate Recommendations

Immediate recommendations focus on those that directly impact the City’s ability to achieve compliance
with current regulations or to minimize risks to public health over those that will achieve optimization as
outlined in the Partnership program. They may also include action items that are needed to inform later
tasks.
Based on the findings of this Gap Analysis, immediate recommendations include:
 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
 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
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
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). (2015h). C713-15 Cold-Water Meters—Fluidic-Oscillator

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). (2016a). C707-10(R16) Encoder-Type Remote-Registration

Systems for Cold-Water Meters. AWWA.
American Water Works Association (AWWA). (2016b). M36 Water Audits and Loss Control Programs (4th
ed.). Denver, CO: AWWA.
arcadis.com 152
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.
Echologics. (2016, June 1). EchoShore-M Leak Detection Report.
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
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.
Rowe Professional Services Company (Rowe). (2016b). Asset Management Report.
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. Review and Analyze Existing Data
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) Chlorine dose and residual

a. Applied at the WTP – data available in monthly operating reports
b. Applied at Reservoir during Fill – need data to complete analysis
2) Distribution system practices
a. Water age impacts
i. Auto-flushers
ii. Reservoir operations
iii. Seasonal demands
iv. Valve positioning
b. Pipe wall demands
i. Flushing in response to main repairs
ii. Unidirectional flushing (UDF)
iii. Spot flushing in response to customer complaints
3) Fill data gaps through additional surveillance monitoring – locations, frequency, and parameters
to be determined- likely to include heterotrophic bacteria, DBPs, other.
3. Recommend Residual Goals

Note: disinfectant residual goals developed for Flint through this process are not meant to be interpreted
or misconstrued as regulatory requirements. The recommended goals are more in line with optimization
principles, which go beyond regulatory compliance requirements. Goals may be adjusted seasonally, by
location, or as appropriate as system conditions warrant.
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
Draft 11-8-16
4. Develop Monitoring Plan to Demonstrate On-Going Status
1) Monitoring locations, parameters, and frequency; sites may include:

a. Final RTCR sites
b. Additional surveillance sites and parameters as needed (to be determined)
c. DBP sites
d. LCR and WQP monitoring
e. Available SCADA and/or data logger data regarding pressures and flows
f. Locations of future HACH water quality monitoring panels
g. Customer complaints, supplemental household data
2) Data tracking, analysis, feedback process
5. Develop Tools for Maintaining/Improving Residuals
1) Chorine dose modifications at WTP

2) Chlorine dose modifications at reservoirs
3) Water age reduction – supported by hydraulic model when possible
a. Within storage facilities
i. Increased turnover
ii. Elimination of excess storage
b. Within distribution system
i. Valve management
ii. Looping dead ends
iii. Auto-flushing
iv. Other
c. At customer taps through flushing initiatives, customer outreach
4) Pipe wall demand reduction
a. Prioritized UDF
b. Rehab/replacement schedule
5) Other consideraitons
3
Draft 11-8-16
APPENDIX B
Corrosion Control Figures
APPENDIX B: ASSESSMENT OF CURRENT PRACTICES
CORROSION CONTROL FIGURES AND GAP ANALYSIS TECHNICAL MEMRANDUM
Maximum Pb Distribution by Database

(# of Locations)
EPA-1S (115) Residential (801)
Sentinel (840) Action Level
100%
Percent of Observations less than Value
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

Maximum Pb Distribution by Database

(# of Location)
EPA-Peak (115) Residential (801)
Sentinel (840) Action Level
100%
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

Max Pb Level for each Location, by Season

(# of Locations)
Winter (750) Spring (820) Summer (241)
Fall (199) Action Level
100%
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

Binned Distribution of Seasonal Max Pb Data

Winter Spring Summer Fall
100%
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

Max Pb Level for Locations with Paired Data, by Season

(# of Locations)
Winter (128) Spring (128) Summer (128)
Fall (128) Action Level
100%
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

Binned Distribution of Seasonal Max Pb - Paired Data

Winter-Paired Spring-Paired Summer-Paired Fall-Paired
100%
90%
80%
70%
% of Locations
60%
50%
40%
30%
20%
10%
0%

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

Locational Pb Levels Before/After May Flush

(# of Locations)
Before (876) After (259) Action Level
100%
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

Locational Pb Levels Before/After May Flush at Paired

Locations
(# of Locations)
Before (232) After (232) Action Level
100%
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

Impact of May Flushing on Pb Levels by Paired

Location
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

Locational Pb Levels Before/After May Flush

(# of Locations)
48502 (3) 48503 (328) 48504 (221)
48505 (134) 48506 (65) 48507 (120)
48517 (1) 48532 (32) Action Level
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


Figure B-11: 90th percentile of maximum lead values at locations sampled in each zip code in
2016, using EPA 1st sample values

B11

Maximum Pb at Location by Service Line Type, All 2016

(# of Locations)
Copper (592) Galvanized (167) Pb (134)
100%
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

Maximum Pb at Location by Service Line Type,

Winter 2016
(# of Locations)
100%
90%
80%
70%
60%
50%
40%
30%
20%
10%
0%
0 1 10 100 1,000 10,000
Pb (µg/L)

or galvanized service lines sampled in Winter 2016, using EPA peak lead values

B13


Spring 2016
(# of Locations)
100%
90%
80%
70%
60%
50%
40%
30%
20%
10%
0%
0 1 10 100 1,000 10,000
Pb (µg/L)

or galvanized service lines sampled in Spring 2016, using EPA peak lead values
B14


Summer 2016
(# of Locations)
100%
90%
80%
70%
60%
50%
40%
30%
20%
10%
0%
0 1 10 100 1,000 10,000
Pb (µg/L)

or galvanized service lines sampled in Summer 2016, using EPA peak lead values
B15

Maximum Pb at Location by Service Line Type, Fall

2016
(# of Locations)
100%
90%
80%
70%
60%
50%
40%
30%
20%
10%
0%
0 1 10 100 1,000
Pb (µg/L)
or galvanized service lines sampled in Fall 2016, using EPA peak lead values
B16

Maximum Pb Levels at Locations with Pb or

Galvanized Service Lines, Paired by Season
(# of Locations)
Winter (34) Spring (34) Summer (34) Fall (34)
100%
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

EPA Sequential Data

Profile Volume where peak occurred
"2nd L" means peak occurred >1,000 but ≤2,000 mL
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

EPA Sequential - >15 vs >50 µg/L - Peak

Peak >15 µg/L 95 samples Peak >50 µg/L 42 samples
45%
40%
35%
Percent of Values
30%
25%
20%
15%
10%
5%
0%
Cumulative Profile Volume (L)

Figure B-19: Distribution of sample volumes were “high lead level” peaks were observed during
EPA Sequential sampling
B19

EPA Sequential - >15 vs >50 µg/L - Any or Peak

Peak >15 µg/L 95 samples Peak >50 µg/L 42 samples
Any >15 µg/L 495 samples Any >50 µg/L 114 samples
45%
40%
35%
Percent of Values
30%
25%
20%
15%
10%
5%
0%
Cumulative Profile Volume (L)

Figure B-20: Distribution of sample volumes where “high lead levels” were observed during
EPA Sequential sampling

B20

Comparison of Peak Lead Values Versus 1st Sample

Values
(typically 125 mL) - EPA Sequential
1st sample each profile (234 profiles)
max of each profile (234 profiles)
peak at all profiles at each house (115 household locations)
Observations Less Than Value (%)
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

Comparison of Peak Lead in a House

EPA Sequential - By Sample Period
Sequential Sequential_2 Sequential_3
Sequential_4 Sequential_5
100%
Observations Less Than Value (%)
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

Figure B-23: Percentage of locations in the EPA Sequential dataset with “high lead level” peaks,
compared to datasets from other locations
B23


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

B-25: Locations of water quality parameter monitoring sites within Flint
B25

Median Total Alkalinity - D6 - 30 day grouping

group median median + (A3 * group range)
median of group medians median - (A3 * group range)
78
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

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

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

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.
CLIENT BY THE NUMBERS

Flint, Michigan
99,000 Population Pipes: 17,935 Reservoirs: 4
579 Miles of Pipe Nodes: 15,938 Pump Stations: 4
Average/Max Day Demands: Tanks: 5 Demand Patterns: 12 (1 in use)
12.1 MGD / Unknown Pumps: 14 Flow Units: gpm
Valves: 10 Roughness Coeff. Range: 90 to 150
AUDIT DATA
‘Flint_RTX_realtime_20160930-
20161028.inp’
EPANET SCENARIOS
Audit Date: 12/9/2016
☒ Average Day Demands ☐ Peak Hour Demands
CALIBRATION ☐ Maximum Day Demands ☐ Fire Flow
Last calibration by: Regan Murray ☐ Minimum Day Demands ☒ Existing System Conditions
(USEPA) and Jim Uber (CitiLogics)
Last calibration date: Oct. 2016
☐ Peak Hour Demands ☐ Future / Planning Conditions
Last edited by: Jim Uber ☐ Other
Last edited date: Dec. 2016
Scenario Comments
Model originally developed in 2009 Single scenario representing actual demands and operations
in InfoWater software. Previous during the period of September 30, 2016 to October 28, 2016.
model development, editing and
No data available to compare date range conditions to annual
calibration by: Rowe Professional
Services Company, Potter average, maximum or minimum day conditions.
Consulting
SIMULATION SETTINGS
☒ Hydraulics ☒ Extended Period ☐ Steady State
☐ Water Quality ☒ Time-Based Controls
☐ Energy ☒ Rule-Based Controls
Duration of EPS Simulation: 672 hours
Hydraulic Simulation Time Step: 0.25 hours

Water Quality Simulation Time Step: 0.1 hours
Page:
1/2
WATER DISTRIBUTION MODEL AUDIT
City of Flint, Michigan
CONDITION ASSESSMENT
☐ Does not complete simulation
Comments
☐ Runs with Warning Messages
Comments
☒ Nodes with zero elevation
☒ Pumps without curves
☐ Pumps cannot meet demand or head conditions
☐ Pressures less than 20 psi
☒ Nodes with negative demand/pressure
☒ 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
RTCR and Chlorine Sample Siting Map

Appendix A – MDEQ Level 1 and 2 Assessment Forms
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.
This Plan has been developed to meet the following objectives:
● 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.
B. Coliform Monitoring Requirements

The following sub-sections outline coliform monitoring requirements which apply to the City.
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
February 2017 2
at the following locations:
● The same tap as the original unsatisfactory Routine sample.

● An active service within five active connections upstream from where the original unsatisfactory
sample was taken*.
● An active service within five active connections downstream from where the original
unsatisfactory sample was taken*.
*If sample sites are not available within five connections up or downstream of the routine sample
location, alternative sites may be selected that represent the up or downstream water quality.
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.
Table 1. System Information
Water System Name: City of Flint

Address: 4500 North Dort Highway
Flint, MI 48505
County: Genesee
WSSN: 2310
Population Served: 98,310 (2015 census)
City of Flint
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.
Table 2. Key Contact Information
Name Title E-mail Phone number(s)

City of Flint – Water System
JoLisa McDay Water Plant Supervisor jmcday@cityofflint.com (810) 787-6537
Vacant Water Plant Ops. (810) 787-6537
Supervisor
William Bradley Lab Supervisor wbradley@cityofflint.com (810) 787-6537
Michigan DEQ
Bob London (primary Corrosion Control londonr@michigan.gov (989) 450-7834
contact for RTCR) Specialist
Jon Bloemker KWA Team Leader bloemkerj@michigan.gov (989) 460-7254
Brian Thurston Distribution Specialist thurstonb@michigan.gov (231) 590-3430
Pollution Emergency 1-800-292-4706
Alerting System
City of Flint
Dr. Karen Weaver Mayor kweaver@cityofflint.com (810) 766-7346
Sylvester Jones City Administrator sjones@cityofflint.com (810) 237-2057
Kristin Moore Public Relations kmoore@cityofflint.com (810) 875-2576
Director
Dr. Pamela Pugh Chief Public Health ppugh@cityofflint.com (810) 237-2041
Advisor
Genesee County Health Department
Jim Henry Environmental Health jhenry@gchd.us (810) 257-3612
Director (517) 404-8401 (cell)
Mark Valacak Health Officer mvalacak@gchd.us (810) 257-3588
Public Notification - Media
mlive – Flint Journal flnews@mlive.com (810) 766-6100
Channel 25 - WEYI (810) 687-9612
E. Sample Locations and Schedule

Under this revised plan, a total of 30 coliform samples will be collected each week from 20 unique
locations.* Previously, the City collected 30 coliform samples each week from 10 unique locations. The
RTCR requires systems to identify sample sites that are representative of the system. Features that
should be represented through coliform monitoring include pressure zones, unique sources, areas
served by reservoirs, and vulnerable conditions such as dead ends, transient usage, high water age, high
hazard cross connections, and sensitive populations. While the ten historical coliform sample sites are
City of Flint
February 2017 4
geographically spread throughout the system, recent water quality concerns and identification of new
areas of low chlorine residual have prompted a re-evaluation of the sample sites under the RTCR. 1The
City’s 10 existing TCR sites have been retained, and 10 new sites have been selected to capture a variety
of conditions to ensure that microbiological water quality is maintained throughout the system. In
addition to coliform, the sites are sampled for chlorine residual. These sites were selected with input
from Flint staff, EPA, MDEQ and Arcadis staff after consideration for site accessibility,
upstream/downstream access, and safety.
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.
Table 3. Monthly Total Coliform Monitoring Locations
Dist. Sampled Historical

Planned Upstream Planned Downstream
Site Routine Site Address on Which Sampling
Site Address1 Site Address1
# Days Site
2501 Flushing Rd M
1 TCR3 2702 Flushing 1117 N. Chevrolet Ave
University Market F
3 1621 Saginaw Street M
2 New 1638 Saginaw St 1709 Saginaw St
Ten Fu Chinese Gourmet F
3609 Beecher M EPA Chlorine
3 3617 Beecher Rd 3505 Beecher Rd
Palace Liquor Store F Monitoring4
Quarterly
(DBP)5;
3521 Corunna M
4 former Taco 3606 Corunna 3409 Corunna
Rite-Aid F
Bell site
upstream3,
1100 Cedar M
5 TCR3 702 12th St 1035 Ann Arbor
Cedar St Reservoir Th(e/o)2
611 W. Court Street, Ste. 200 M
6 New 620 W. Court 521 W. Court
Dean T. Yeotis Law Offices Th(e/o)2
1159 Foss Ave. Tu EPA Chlorine
7 1185 Foss Ave 1164 E. Holbrook Ave.
Foss Avenue Church Th(e/o)2 Monitoring4
8 6204 N. Saginaw St Tu TCR3 6101 N. Saginaw St 6509 N. Saginaw St
1
*Instances, including but not limited to holidays, declining participation, and site closings, may reduce the number
of weekly samples collected.
City of Flint
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
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).
Repeat Sample Sites

Most of the repeat sample sites shown in Table 3 are within five active connections up/downstream of
the routine sample site. The City will attempt to use the repeat sample locations shown in Table 3 for
repeat coliform sample collection; however, because they are not routine sample locations, there are
times when they may be unavailable. In that case, the City will use an alternate site that is
representative of the water quality for the routine sample site. If an alternative site is needed, the City
will send a written justification to DEQ for use of the alternative site.
F. Sampling and Reporting SOP

Monthly Monitoring
1. Collect samples at regular time intervals throughout the month.
2. Failure to collect all required routine samples in a monitoring period is a violation. Another total
coliform monitoring violation in the following 12 months will result in a fine. Notify the DEQ if
monitoring was not performed as required.
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:
1) Remove any tap attachments such as aerators or filters.

2) If tap cleanliness is questionable, apply disinfectant (100 mg/L solution of sodium hypochlorite
or alcohol wipes) to sample tap before flushing service line.
3) Open tap fully and let water run until the water is cold, to permit turnover of the water in the
service line and premise plumbing prior to the sample point.
City of Flint
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.
Procedure When Sample Result Is POSITIVE

1) If a distribution system sample result is positive for total coliform bacteria or E.coli, then collect
repeat samples within 24 hours of learning of the positive result from all of the following sites:
a. The site of the positive sample result; and
b. An upstream site within five service connections of the original positive site; and
c. A downstream site within five service connections of the original positive site; and
Note: If approved by the DEQ, when an upstream repeat site is temporarily not available, sample the
closest available location to the repeat site that is also upstream of the routine site and similarly for the
downstream side.
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
February 2017 8
City of Flint E. coli Response Plan
If we have E. coli in our distribution system we will immediately:

1. Call Lab Supervisor and MDEQ by end of the day that the City is notified of positive
sample.
2. Contact GLWA
● Review source water data
● Identify operational changes
3. Contact Genesee County Public Health
4. Collect repeat samples outlined in section E or alternate locations that are expected to
represent pathways of contamination into the distribution system.
● Collect investigative samples as needed.
5. Inspection of potential pathways and correct as needed:
● Review SCADA records for pressure irregularities.
● Reservoir inspection for potential pathways; screens, hatches, locks, etc.
● Interview staff for operational changes, water main breaks, pressure outages,
etc.
● Review cross connection program and status.
● Review construction activities.
See Appendix A for MDEQ Level 1 and 2 Assessment Forms
6. Implement health advisory as directed by MDEQ if deemed necessary.
7. Review repeat sample results.
● Any repeat samples unsatisfactory: Issue health advisory. Work with MDEQ to
Conduct Level 2 Assessment
● Establish criteria for lifting health advisory.
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.
E. coli MCL Violations

An E. coli MCL violation occurs if a routine sample and at least one related repeat sample both have
coliform bacteria present and one of the samples is positive for the presence of E. coli. If this occurs,
contamination is confirmed in the water supply.
If an E. coli MCL violation occurs, follow the steps outlined in Section F.
Other Types of Violations

Other types of violations related to RTCR compliance include treatment technique violations, monitoring
violations, and reporting violations:
City of Flint
February 2017 9
● Treatment Technique (TT) violations result from failure to perform appropriate assessments or
correction actions. TT violations require Tier 2 public notification (public notice as soon as
practicable, but not later than 30 days following the violation).
● Monitoring violations result from a failure to collect any coliform sample (routine or repeat) or
a failure to analyze an unsatisfactory total coliform positive sample for E. coli bacteria. Tier 3
public notification is required for monitoring violations. For this type of notification, the City
must communicate the violation to customers within 12 months of the violation, which could be
included in the Consumer Confidence Report (if it is issued within 12 months).
● Reporting violations result from failure to submit a monitoring report or completed assessment
form in a timely manner or failure to notify MDEQ following an E. coli positive sample in a timely
manner. Reporting violations require Tier 3 public notification.
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.
Table 4 provides a basic comparison of Level 1 and Level 2 Assessments.
City of Flint
February 2017 10
Table 4. Comparison of Level 1 and Level 2 Assessment Requirements
Level 1 Assessment Level 2 Assessment

Trigger for assessment ● TC+ in >5% of monthly ● E. Coli MCL violation, or
samples, or ● Two Level 1 Assessments
● Failure to collect any repeat required within 12-month
samples following TC+ sample period
Party responsible for Supply owner or operator MDEQ
completing assessment
Assessment Deadline ● City to initiate assessment as ● MDEQ to initiate assessment as
soon as practicable after soon as practicable after
learning of trigger learning of trigger
● Submit assessment form to ● Correct sanitary defects as soon
DEQ within 30 days of as practicable or on State-
learning that a trigger was approved timeline identified in
exceeded form
Assessment Form Level 1 assessment form for Level 2 assessment form for
community water supplies community water supplies
(Appendix A) (Appendix A)
Minimum assessment ● Review and identification of atypical events that could have
elements affected or impaired distributed water quality
● Changes in DS O&M that could affect DS water quality (inc.
storage)
● Source and treatment considerations affecting DS water quality
● Review of water quality data
● Inadequacies in sample sites, sampling protocols, and sample
processing
● Form describing sanitary defects detected, corrective actions
completed, and a proposed timetable for each corrective action
not already completed
City of Flint
February 2017 11
City of Flint
RTCR Sampling Sites
7 Foss Avenue Baptist Church

8 North Flint Automotive
´
22 Hasselbring Senior Center
9 Rite-Aid Pharmacy 23 Sam's Rollingwood

"WaterMarket
Treatment Plant
13 Grandma's Recipes
10 Auto Zone
19 Salem Housing 25 Fire House #5
15 Luigi's Restaurant 14 Admiral Station

20 Fire House #3
11 West Side Reservoir
1 University Market
12 Hoffman's Deco Deli & Café
16 Rite-Aid Pharmacy
6 Dean T. Yeotis Law Offices

3 Palace Liquor Store
5 Cedar St. Reservoir
2 Ten Fu Chinese Gourmet
4 Rite-Aid Pharmacy
24 Genesee Community Health Center

17 Liquor Palace
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
Reservoir and RTCR Sites

Chlorine Only Sites 0 0.4 0.8 1.6 2.4 3.2
Miles
Appendix A - Level 1 and 2 Assessment Forms
MICHIGAN DEPARTMENT OF ENVIRONMENTAL QUALITY
OFFICE OF DRINKING WATER AND MUNICIPAL ASSISTANCE
LEVEL 1 ASSESSMENT FORM FOR COMMUNITY WATER SUPPLIES

Issued under authority of the Safe Drinking Water Act, 1976 PA 399, as amended,
MCL 325.1001 et seq., and its Administrative Rules (Act 399).
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:
Assessor Name: Assessor Title:
Phone Number: E-mail:
Trigger Event: Greater Than 5% Total Coliform Positives or Failure to Collect All Repeat Samples
Date Assessment Triggered: Date Assessment Completed:
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.
A. Sample Site Selection and Sample Collection

Answer
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?
DEQ Environmental Assistance Center www.michigan.gov/deq

Telephone: 1-800-662-9278 Page 1 of 3 EQP 5826 (Rev. 3/2016)
C. Source – Surface Water (if surface water is not used check here and go to subsection D) Answer
Yes No NA
Are there any new potential contamination sources, or visible signs of unsanitary conditions near the raw
water intake?
Any signs of vandalism or unauthorized access to source facilities?
Was there any heavy precipitation, rapid snowmelt or flooding recently?
Any unusual changes to quality of the raw water like a spike in turbidity, sudden change in pH or very high
heterotrophic plate counts?
D. Treatment (if no treatment check here and go to subsection E) Answer

Yes No NA
Have there been additions or modifications to any treatment process?
Have there been interruptions in any treatment process?
Any signs of vandalism or unauthorized access to treatment equipment or facilities?
Are there any issues with operation or maintenance of treatment equipment, units or processes?
Is there any water quality data that indicates treatment is ineffective?
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?
F. Distribution System Answer

Yes No NA
Have there been any low pressure events (< 20 psi)?
Have there been any water main breaks, repairs, or new main installations?
Have there been any recent fires or hydrant flushing?
Have there been any booster pump issues, repairs or new installations?
Is the supply actively performing cross connection control inspections, including frequent testing of all
testable backflow preventers?
Have there been other construction activities like hydrant or valve replacement that could have introduced
contamination into the system?
If samples were collected from inside a building, has there been any recent plumbing work performed
within the building?
G. Operation and Maintenance (O & M) Answer

Yes No NA
Any changes in procedures or staff effecting O & M activities?
Any water quality data collected from the treated water tap or distribution system show results are
indicative of an issue?
Any complaints from customers related to water quality or low pressure?
Any other issues or items that may have caused bacteriological contamination?

4. Issue Description: For any answer in Part 2, Subsections A - G that are 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 page(s) if needed. Include date(s) of low pressure events, 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(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.
Assessor’s Name (printed):
Assessor’s Signature: Date:
-----------------------------------------------------------------------------------------------------------------------------------------
DEQ USE ONLY: This section is to be completed by DEQ.
Reviewer 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:

MICHIGAN DEPARTMENT OF ENVIRONMENTAL QUALITY
OFFICE OF DRINKING WATER AND MUNICIPAL ASSISTANCE
LEVEL 2 ASSESSMENT FORM FOR COMMUNITY WATER SUPPLIES

Issued under authority of the Safe Drinking Water Act, 1976 PA 399, as amended,
MCL 325.1001 et seq., and its Administrative Rules (Act 399).
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
CWS Name: WSSN:
DEQ Staff Completing Assessment:
Name & Title of Person Representing the CWS During Assessment:
Level 2 Trigger: E. coli MCL or 2nd Level 1 Assessment in 12 months

Date Assessment Triggered: Date Assessment Completed:
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?

B. Source – Wells (if wells are not used check here and go to subsection C) Answer
Is the electrical conduit damaged or not sealed to the well cap?
Is the ground graded to prevent water flow towards the wells?
Is there standing water or other unsanitary conditions near the wells?
Have any wells/pumps undergone any recent repairs or maintenance activities?
Do the wells have adequate isolation distances from sources of contamination?
Does the raw water quality data indicate changes to the source water quality?
Has the pumping capacity of the well(s) changed recently?
Have there been any sewer or chemicals spills, or other disturbances near the wells?
Have any backup or emergency wells been placed into service?
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
E. Treatment (if no treatment check here and go to subsection F)
Yes No NA
If sand/gravel or other mixed media filtration is used, are the media depths near the original design depths
and are the underdrains in good condition?
Did the plant flow exceed the state rated treatment capacity?
For surface water plants, did a review of the turbidity data reveal any anomalies?
For membrane plants, is daily integrity testing being performed every 24 hours of operation and do the
results indicate that the membranes are in good condition?
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?
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.
Assessor’s Name (printed):
Assessor’s Signature: Date:
-----------------------------------------------------------------------------------------------------------------------------------------
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:

28550 Cabot Drive

Suite 500
Novi, Michigan 48377
Tel 248 994 2240
Fax 248 994 2241
www.arcadis.com
ADDENDUM 1
Update to Disinfectant Residual and Disinfection Byproducts
MEMO

To: Copies:
28550 Cabot Drive
City of Flint, MI Chris Hill and Rebecca Slabaugh, Suite 500
Arcadis Novi
Michigan 48377
Tel 248 994 2240
From: Fax 248 994 2241
Melinda Friedman, Confluence Engineering Group, LLC
Date:
April 6, 2018
Subject:
Addendum No. 1 to the Assessment of Current Practices and Gap Analysis

Technical Memorandum, March 2017
Update to Disinfectant Residual and Disinfection Byproducts
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.
UPDATE TO ORIGINAL ANALYSIS
Updated Data Sources

The analysis of chlorine levels in the distribution system that was presented in the original memo was
updated to include the following data:
 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.
CHLORINE RESIDUAL MANAGEMENT TOOL
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.
Input Data and Sources

The CRM Tool can be used to develop statistical and graphical analysis of:
 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.
Data Tracking and Analyses

The CRM Tool provides the following analyses for free chlorine residuals across all 26 locations (25
distribution sites plus the POE):
 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

To: Copies:
28550 Cabot Drive
City of Flint, MI Chris Hill and Rebecca Slabaugh, Suite 500
Arcadis Novi
Michigan 48377
Tel 248 994 2240
From: Fax 248 994 2241
David Cornwell and Damon Roth, Cornwell Engineering Group
Date: Arcadis Project No.:
April 6, 2018 20616001.0000
Subject:
Addendum No. 2 to the Assessment of Current Practices and Gap Analysis

Technical Memorandum, March 2017
Update to Internal Corrosion Control
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.
UPDATE TO ORIGINAL ANALYSIS
Updated Data Sources

The analysis of lead levels in the distribution system that was presented in the original memo was based
primarily on data from three sources:
• Residential dataset (MDEQ)
• Sentinel dataset (MDEQ)
• Sequential dataset (USEPA)
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
Water Quality Parameters (WQPs)

In addition to lead sampling programs that were conducted by MDEQ and USEPA, the City of Flint
continued to collect operational data both at the plant and at 10 WQP monitoring sites in the distribution
system on a regular basis. Prior to January 1, 2017, data were collected from each of the 10 WQP
monitoring sites on a twice-weekly basis. After January 1, 2017, in conjunction with the expansion of the
RTCR program and sample scheduling changes, data were collected at least once each week from the 10
WQP locations.
Updated Data Analysis
Lead Levels
The original study analyzed the available lead data in a number of ways, including:
 Seasonal Lead Distribution
 Before/After Flushing Event
 Spatial Lead Distribution
 Service Line Material

Out of these analyses, the most relevant comparison is the seasonal lead distribution. The other analyses
looked at the distribution of the maximum lead observations relative to the analytical parameter; since lead
levels have largely decreased across the City subsequent to the original analysis, the analyses based on
the maximum lead observations are largely unchanged. However, because the seasonal lead distribution
looks at changes over time, it is useful to repeat this analysis with the current, updated dataset.
The original analysis considered data across four seasons in 2016: winter (Jan-Feb), spring (Mar- May),
summer (Jun- Aug), and fall (Sept-Oct). The updated analysis includes 2017 data, and expands the
winter (Dec-Feb) and fall (Sept-Nov) sampling periods. To better understand the changes in lead levels at
specific locations over time, a subset was created from only those sites that were sampled during all
seasons in the data set. As of the original analysis, there were 126 locations with paired data in all four
2016 seasons. Through Summer 2017, the number of paired locations decreased to 82. The Fall 2017
sampling only included seven locations, three of which were common to the paired dataset. Therefore, Fall
2017 data was excluded from the paired data analysis. Table 1 summarizes the difference between the
sample populations of each dataset.
Table 1: Summary of Paired and Unpaired Locations by Season
Winter 2016 750 82
Spring 2016 820 82
Summer 2016 241 82
Fall 2016 196 82
Winter 2017 179 82
Spring 2017 175 82
arcadis.com Page:
3/7
MEMO
Summer 2017 159 82
Fall 2017 7 N/A
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.
Max Pb Level for Locations with Paired Data, by Season

(# of Samples)
Winter 2016 (82) Spring 2016 (82) Summer 2016 (82)
Fall 2016 (82) Winter 2017 (82) Spring 2017 (82)
Summer 2017 (82) Action Level
100%
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
Binned Distribution of Seasonal Max Pb - Paired Data

Winter 2016 Spring 2016 Summer 2016 Fall 2016 Winter 2017 Spring 2017 Summer 2017
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%
Figure 2: Binned distribution of maximum lead values at locations with paired data by season, using USEPA
1st sample lead values
Water Quality Parameters
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
Group Median UCL Median of Group Medians LCL
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)
NEW DATA ANALYSIS

In addition to the updated analyses discussed above, a new data analysis of the Confirming Lead
Elimination After Replacement (CLEAR) dataset was conducted since the original memo and is presented
below.
CLEAR Dataset Analysis

In April 2017, MDEQ initiated the CLEAR program. The goal of this program was to recruit at least 200
households to participate in tap water monitoring prior to and after replacement of the lead or galvanized
steel service line. The initial sampling round was to be conducted within 90 days prior to replacement of
the service line, with six subsequent sampling rounds to be conducted monthly following replacement of
the service line.
The initial sampling round included 235 households. Participating in the subsequent sampling rounds
ranged from 147 to 187 households, resulting in a final data set with 94 locations with paired data across
all seven sampling rounds. Figure 4 shows the percentile distributions of the paired data.
The CLEAR dataset shows that overall, even prior to replacement of the lead or galvanized steel service
lines, the 90th percentile of lead data from the participating households was less than the 15 µg/L action
level for lead. The percentile distribution of lead levels also decreased over time following the service line
replacement. However, some locations did see an increase in the lead concentration after the service line
replacement (Rounds 2 and 3). This is consistent with the mobilization of particulate lead during the
service line replacement event. These elevated lead levels did appear to subsequently decrease as time
passed (Rounds 4 – 7).
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.
28550 Cabot Drive
Suite 500
Novi, Michigan 48377
Tel 248 994 2240
Fax 248 994 2241
www.arcadis.com

Appendix A: Assessment of Current Practices and Gap Analysis Technical Memorandum

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Appendix A: Assessment of Current Practices and Gap Analysis Technical Memorandum

Uploaded by

Copyright:

Available Formats

APPENDIX A

ASSESSMENT OF CURRENT PRACTICES AND GAP ANALYSIS

Arcadis of Michigan, LLC

This document is intended only for the use of

System Overview .................................................................................................................................... 1

1.0 Introduction ........................................................................................................................................... 21

1.1 Project Background....................................................................................................................... 21

2.0 Water System Overview ....................................................................................................................... 24

2.1 Water Treatment Plant .................................................................................................................. 24

2.1.1 Supply ................................................................................................................................ 24

2.2 Water Distribution System ............................................................................................................ 25

2.2.1 Distribution System Storage .............................................................................................. 25

3.1 Industry Best Practices ................................................................................................................. 28

3.2.1 Coordinate with Regulatory Agencies ............................................................................... 37

4.0 Results .................................................................................................................................................. 44

4.1 Disinfectant Residual and Disinfection Byproducts ...................................................................... 44

4.1.1 Background ....................................................................................................................... 44

4.2 Main Breaks and Impacts of Pressure .......................................................................................... 57

4.2.1 Background ....................................................................................................................... 57

4.3 Pressure Management ................................................................................................................. 76

4.3.1 Background ....................................................................................................................... 76

4.4 Internal Corrosion Control ............................................................................................................. 79

4.4.1 Background ....................................................................................................................... 79

4.5 Distribution System Practices ....................................................................................................... 95

4.5.1 Cross-Connection Control ................................................................................................. 95

4.5.12 Water Loss Control .......................................................................................................... 124

4.6 Supporting Utility Programs ........................................................................................................ 132

4.6.1 Administration .................................................................................................................. 132

5.0 Summary and Immediate Recommendations..................................................................................... 139

5.1 Summary..................................................................................................................................... 139

5.1.1 Approach ......................................................................................................................... 139

5.2 Immediate Recommendations .................................................................................................... 149

6.0 References .......................................................................................................................................... 151

Table 4-8: Water Department Staffing Summary ...................................................................................... 134

ACRONYMS AND ABBREVIATIONS

GAC Granular Activated Carbon

OWQM Online Water Quality Monitoring

Table ES-1: Distribution System Pipe Characteristics

 52% as 6-in pipe (~300 miles)

 Disinfection residual (representative of distribution system water quality integrity),

Areas that Meet

Figure ES-1: Gap Analysis Categories

Category Description Source

Category Description Source

Category Description Source

Category Description Source

Includes design specifications for locations,

Category Description Source

Category Description Source

AWWA) & Economic and Engineering

Category Description Source

Table ES-3: Areas that Meet or Exceed Industry Standards

Areas that Meet or Exceed Industry Standards

Areas that Meet or Exceed Industry Standards

Table ES- 4: Areas that Need Minor Strengthening

Areas that Need Minor Strengthening

Areas that Need Minor Strengthening

Areas that Need Minor Strengthening

Table ES- 5: Areas that Need Major Strengthening

Areas that Need Major Strengthening

Areas that Need Major Strengthening

Areas that Need Major Strengthening

Areas that Need Major Strengthening