Wayne,
Per your request.
Eric Cline | Department Manager | State of Michigan
Michigan Department of Treasury | Local Government Financial Services Division | Fiscal Responsibility Section
430 W. Allegan Street, 3rd Floor | Lansing, MI 48922
Lansing Office (517) 335-2078 | Cell Phone (517) 243-8450 | Traverse City Office (231) 922-5228
E-mail cliner 1 @muchigan.gov
CONFIDENTIALITY NOTICE: This e-mail, and any attachments, is for the sole use of the intended recipient(s) and may contain
information that 1s confidential and protected from disclosure under the law. Any unauthorized review, use, disclosure, or distribution
is prohibited. If you are not the intended recipient, please contact the sender by reply e-mail, and delete/destroy all copies of the
original message and attachments Thank you.
xxxEND_PAGE:treasury01_b35_5648_6098_252
<r
.
4a,
“ey
“ia,
Operational Evaluation Report
City of Flint
Trihalomethane
Formation Concern
February 27, 2015
Lockwood, Andrews
& Newnam, Inc.
A LEO A DALY COMPANY
xxxEND_PAGE:treasury01_b35_5648_6098_253
CITY OF FLINT a
Operational Evaluation Report
February 27, 2015
EXECUTIVE SUMMARY
Environmental Protection Agency (EPA) and Michigan Department of Environmental Quality
(MDEQ) regulations require that public water suppliers test drinking water quarterly
throughout the distribution system for disinfectant by-products (DBP’s). Two categories of
DBP’s, tri-halomethanes (THM) and halo-acetic acids (HAAS), are regulated and must be
tested for. The City of Flint began operation of their water treatment plant (WTP) full time
with the Flint River as the source on April 25, 2014. Since that time, four quarters of samples
taken have resulted in an annual average violation for total THM. Prior to the first violation
(Nov, 2014), the City hired Lockwood, Andrews & Newnam, Inc. (LAN) to complete this
Operational Evaluation Report (OER) in conformance with EPA guidelines with the goal to
determine the cause(s) of high levels of THM and evaluate possible solutions.
The EPA promulgated the Stage 2 Disinfectants and Disinfection By-Products Rule (DBPR) in
January 2006 which set maximum contaminant levels (MCLs) for total trihalomethanes
(TTHM) and HAAS based on an annual running average, tested quarterly, for a given
sampling location. The City of Flint reports levels from 8 sampling test locations. Of the four
quarterly sampling cycles since Flint began operating the WTP full time, HAAS levels have
been acceptable but TTHM levels were high at 4 sampling sites following the third sampling
cycle. Average THM levels exceeded the MCL at 3 sites following the fourth sampling cycle.
A number of issues have been identified as possibly contributing to the high THM levels
measured.
Inefficient ozone system functionality which has resulted in increased chlorine feed.
Upstream source influences in terms of increased chlorine demand.
Bypass stream around softening contributed to chlorine demand and increased total
organic carbon (TOC) levels in the effluent.
Unlined cast iron pipes in the distribution system contributing to chlorine demand.
High water age in the distribution system due to:
a. Broken valves causing less than ideal flow patterns.
b. Inefficient pump station pressure zones
c. Water storage volumes in excess of that needed for today’s demands
d. Oversized water mains
e. Low water demands
6. High chlorine demand in filters.
7. High THM formation potential (THMFP) in source water.
8. Less than optimal removal of THM precursors
ar YEP
A graphical representation of how the factors above relate to the timing of THM compliance
sampling is shown as Table 1. Compliance sampling dates are hatched. Each row in Table 1
describes a factor than can lead to increased THM levels and the table defines when each of
those factors applied. Note the convergence of nearly all factors around the second sampling
period on August 21, 2014 to create what appears to be a worst case scenario. The table also
shows that the factors listed as those that the City can control have been addressed prior to
the February sampling period. Monthly operating report data up to February 1,2015 is
depicted on the Table.
Page 1 of 22 ion Lockwood, Andrews
A LEO A DALY COMPANY
xxxEND_PAGE:treasury01_b35_5648_6098_254
CITY OF FLINT
Operational Evaluation Report
February 27, 2015
s10j2e} waysds!"
SIHSHBYICIEYD JOANY YUN]
JONUOD Jeys JULY JapuN si0}eY
ep duydues aouerjduo+
sivak ainyny ut yeadas 0} paydadxa pue a3ua1NI20 Jo sayLq
@UALNIIO JO Sayeq
Andrews
Inc.
A LEO A DALY COMPANY
&Newnam,
lon &
Page 2 of 22
xxxEND_PAGE:treasury01_b35_5648_6098_255
CITY OF FLINT
Operational Evaluation Report
February 27, 2015
ACTION PLAN
The City of Flint has signed an agreement with the Karegnondi Water Authority (KWA) to
purchase raw water drawn from Lake Huron. The KWA system is currently under
construction and expected to be operational by late 2016. The water supply from Lake Huron
will have entirely different water quality characteristics from the Flint River and those
characteristics are expected to yield drastically reduced DPB fermation. With that, non-
structural options to help reduce THM levels are much preferred over solutions requiring new
construction. Therefore, two categories of actions have been devised: Stage 1 being actions
that can be completed relatively quickly without major construction and Stage 2 consisting of
either long term actions or solutions requiring major construction. The City is actively
working to complete Stage 1 actions as soon as possible. Stage 2 actions are to be
implemented only if Stage 1 actions are ineffective in adequately reducing TTHM levels and
therefore Stage 2 is contingent upon the outcome of Stage 1. As of the date of this report,
status updates for action items are shown in red.
Stage 1 — Immediate Actions
® Hire third party water quality expert to complete independent ‘water audit’
o The City hired Veolia Water to review all water quality related operations,
procedures, actions taken and planned responses. Recommendations from
Veolia are expected by the first week in March 2015.
e Obtain an in house THM analyzer to allow regular operational monitoring of THM
levels
o THM analyzer was installed 2/17/15.
e Hire ozone system manufacturer to troubleshoot ozone system
o Manufacturer and controls programmers performed on site evaluations in
January 2015.
e Bench scale jar testing
o Match existing process and assess possible areas of improvement
= Existing process was simulated and an evaluation of existing chemical
feed dosages has been completed by LAN.
= Existing process TOC profile was developed by Veolia.
o Simulate potential modifications to treatment process
= Soda ash softening evaluation completed and PAC feed testing
completed by LAN.
° Exaloate coagulation and flocculation polymer aid feeds to assist with TOC
remova
= Evaluations of polymer aids completed by LAN and PVS Technologies.
e WTP operational changes
o Discontinue softening bypass stream to reduce chlorine demand
= Operational directive has been set to soften no less than 80% of flow.
© Disinfection of filter beds to reduce chlorine demand
= Utility Service Group contracted by City and condition assessment
completed. Controls improvements have been completed.
© Begin coagulation and flocculation polymer aid feeds to assist with TOC
removal if bench scale test results are positive
= Jar testing completed to date has not indicated a useful benefit to
feeding coagulation/flocculation polymer aids. Increased ferric doses
have been implemented at the WTP based on positive jar test results.
Page 3 of 22 an Sxclcercne Aires
| &Newnam, inc. __
& LEO A DALY COMPANY
xxxEND_PAGE:treasury01_b35_5648_6098_256
CITY OF FLINT
Operational Evaluation Report
February 27, 2015
e Increase water main flushing efforts to minimize stagnant water
0 Flushing efforts are ongoing as weather permits.
e Water system modeling to identify areas with high water age and potential solutions
o The water model has been improved and preliminary results, including
system wide water age, have been produced. Water demand updates and
reconciliation with operator’s data are scheduled for March 2015 to complete
water modeling analysis.
o Cedar Street Pump Station potential recirculation
= Water model analysis to be completed in March 2015.
o West Side Pump Station potential recirculation
= Water model analysis to be completed in March 2015.
o Storage tank volume use
« Operating levels of West Side and Cedar Street reservoirs have been
lowered to reduce water age
0 Possible broken closed valve locations
" Model has been updated with known broken valve locations. Model
results are being evaluated for indications of other possible broken
valves. City has also initiated valve assessment program.
o Locations in need of flushing
« High water age areas have been identified in the water model. Further
evaluation forthcoming to determine most effective flushing points.
Stage 2 — Contingent Actions
e Fix ozone system
o Repairs have been made to gauges and programming and the system 15
producing proper ozone and functioning under manual operation. Further
minor repairs are planned for the 1* quarter 2015 to allow automatic
operation,
e Start feeding coagulation and flocculation polymer aids to lower TOC, if not
completed in Stage 1
o Polymers evaluated by LAN did not demonstrate notable benefit.
o PVS Technologies evaluated a proprietary polymer that showed little benefit.
e Convert to lime and soda ash softening
o Cost effective analysis to be developed, if necessary, based on routine
operation THM level monitoring.
e Change disinfectant to chloramine or chlorine dioxide until KWA
o Cost effective analysis to be developed, if necessary, based on routine
operation THM level monitoring.
e Install pre-oxidant feed at intake to optimize ozone disinfection
o Permanganate feed at intake was evaluated by Veolia the week of 2/16/15
and recommendations are expected by the first week in March 2015.
e Replace filter media with granular activated carbon (GAC) media
o Cost effective analysis to be developed, if necessary, based on routine
operation THM level monitoring.
© Implement advanced treatment for THM precursor removal
o Cost effective analysis to be developed, if necessary, based on routine
operation THM level monitoring.
® Increased main flushing based on water modeling results
o Water model analysis to be completed in March 2015.
Page 4 of 22 on Lockwood, Andrews
I wna,
A LEO A DALY COMPANY
xxxEND_PAGE:treasury01_b35_5648_6098_257
CITY OF FLINT
Operational Evaluation Report
February 27, 2015
e Continue valve replacements with water model assistance
o Water model analysis to be completed in March 2015.
e Emphasize cast iron pipes on water main replacement priority list
o Flint has bid replacement of over 2 miles of 24” cast iron pipe along Dupont
and Bishop Streets to be completed this coming construction season The
water main section is considered a critical transmission main, and is expected
to contribute to decreases in water age when complete
THM samples have been taken and tested 5 times since the City began using the Flint River
for supply. Four sets of samples were taken for official regulatory compliance and one set of
samples was taken by the City for operational monitoring. Samples were taken in May
2014, August 2014, November 2014, January 2015, and February 2015. THM levels at all
sample sites have declined from August 2014 to February 2015. The average of all sample
sites in August was 142.1 ug/l and the most recent average of samples taken in February
2015 was 19.8 ug/l. The MCL defined by the EPA is 80 ug/l.
& LEO A DALY COMPANY
Page 5 of 22 an pee Andrews
| jewnam,
xxxEND_PAGE:treasury01_b35_5648_6098_258
CITY OF FLINT
Operational Evaluation Report
February 27, 2015
[. BACKGROUND
The City of Detroit Water and Sewer Department (DWSD) has historically provided drinking
water for the City of Flint and Genesee County. In the late 1990’s growing concern regarding
the reliability of the DWSD supply prompted the City of Flint to upgrade their existing water
treatment plant (WTP). Those improvements, defined as Phase I, were completed in 2005
and were intended to allow the Flint WTP to operate, using the Flint River as the source, for
an extended period of time in the event that supply from the DWSD was temporarily
interrupted. Additionally, the Phase | improvements set the stage for Flint to break free from
dependence on the DWSD supply and water charges over which they had no control.
A. WATER SUPPLY TRANSITION
1. Detroit Water and Sewer Department (DWSD)
Until recently the Genesee County and Flint region had been provided drinking water
by the DWSD. However, due to excessive cost increases and reliability issues with
the DWSD system other options had to be explored.
2. Karegnondi Water Authority (KWA)
In 2010 the Karegnondi Water Authority (KWA) was formed for the purpose of
developing a new water supply from Lake Huron to serve the region in lieu of the
DWSD supply and the City of Flint elected to join. The KWA expects the new system
which is currently being constructed to become operational by the fall of 2016.
3. Flint River — Interim Period
With a renewing water supply agreement between Flint and the DWSD being
terminated by the DWSD (effective April 30, 2014) and the KWA system not
expected to be operational until late 2016, the City of Flint decided to initiate
operation of the existing WTP full time utilizing the Flint River as the interim water
source. A variety of WIP improvements were necessary for the Flint plant to become
a full time plant. For purposes of this report, Phase II improvements to the Flint WTP
are improvements intended to allow the plant to operate full time with either the Flint
River as the source or the KWA supply as the source.
B. TTHM VIOLATIONS
The EPA and MDEQ method of determining if TTHM sample results exceed the MCL uses
a locational running annual average (LRAA). Flint’s first TTHM violation was cited by the
MDEQ following the third cycle of sampling completed in November 2014. Of the 8
sampling sites, 4 were in violation. At that time the MDEQ used the following calculation
for determining if the MCL had been violated:
(2 x current quarter value + previous 2 quarter values) / 4 = Operational Evaluation Value
Flint has now completed tests for 4 quarters and a straight annual running average
applies. Based on samples taken on February 17, 2015 the number of sites in violation of
the THM MCL limit has decreased to three. The City also conducted a round of sampling
on January 27, 2015 in order to internally monitor the progress of actions taken to
address the THM issue. It is worth noting that each round of sampling since August of
2014 has indicated a significant drop (improvement) in THM levels. Test results are
fabolied in Table 2. HAAS sample results are shown in Table 3, of which Flint has had
na violations.
Page 6 of 22 an Lockwood, Andrews
| SNewnam, tc. __
A LEG A DALY COMPANY
xxxEND_PAGE:treasury01_b35_5648_6098_259
CITY OF FLINT
Operational Evaluation Report
February 27, 2015
Sample Location | sovtig sath wetid | were ed LRAA
WIP Ta 56 86 33 16 = Ta
A Se ise 162.4 | 145.3 58.6 35 16.2 95.6
Bae aut tay 116 | 1120 | 36.2 23 19.9 | 69.9
Tiquor Palace Hwy 96.5 | 127.2 | 33.3 23 16.8 | 685
a a6 omunna | 106.4 | 181.3 | 33.9 24 18.1 84.9
ef cout Fie ig 75.1 196.2 | 93.6 35 245 97.4
oe aM 82.2 1142.4 50.1 33 28.5 68.3
eae vale 88.2 144.4 53.6 29 19.2 76.4
is pal 79.2 118.3 | 41.1 21 14.9 63.4
TTHM MCL = 80 ug/l
TABLE 3 — HAAS TEST RESULTS (ug/L)
Sample Location
1) 3719 Davison
McDonalds
2) 822 S. Dort Hwy
BP Gas Sta.
3) 3302 $. Dort Hwy
Liquor Palace
4) 3606 Corunna
Taco Bell
1° Qrt Qrt | 3™ Qrt
5/21/14 | B/21/14 | 11/21/14
36 (taken June 14,
Inter. 4rth Qrt
1/27/15 | 2/17/15
2014) [| na
pe [ae |
30 |
:
Rite Aid
8) 6204 N. Saginaw
N. Flint Auto
HAAS MCL = 60 ug/l
5) 2501 Flushing
Univ. Market
6) 3216 MLK
Salem Housing bi Na
7) 5018 Clia 49 N
Cape
a
a
C. WATER TREATMENT PLANT RECENT IMPROVEMENTS & STATUS
1. Phase | WTP Improvements
Since 1965, the Flint WTP has remained a secondary or backup supply system to the
DWSD primary supply. Typically the secondary supply for a public water system is
expected to be needed only during emergency situations and normally is designed for
short term operation such as providing the average daily demand for a few days.
Conversely, Phase | improvements were designed with the intent to upgrade the Flint
WTP in order to allow for an extended short term period (6 weeks) because of the
perceived high risk that the DWSD supply would fail and remain out of service for an
Page 7 of 22
Lockwood, Andrews
& Newnam, Inc.
A LEQ A DALY COMPANY
lon
xxxEND_PAGE:treasury01_b35_5648_6098_260
CITY OF FLINT
Operational Evaluation Report
February 27, 2015
extended duration. Regardless, the Flint WTP was still intended to serve as a standby
plant and as such the Phase | improvements lacked redundancies that would be
required for a primary supply WTP.
2. Past Pilot Study & Testing
During design of the Phase | improvements a treatability study was completed by
Alvord, Burdick & Howson, LLC (AB&H) in 2002. The Treatability Study evaluated
the current treatment processes that are in place at the Flint WTP today with the Flint
River as the source. The report recommended the following:
0 R O D O
: Point of Dosage
Treatment Purpose Application (mg/l)
Sodium permanganate | Zebra mussel control Intake 0.3
Ozone Taste & odor removal, disinfection _[_Diffusor basin 1.5
Ferric chloride Coagulation Rapid mix 40
Coag aid polymer Turbidity & TOC removal Rapid mix 2.0
Floc aid polymer Turbidity & TOC removal Floc basin 0.05
Lime Softenin Softening basin 175
Soda ash Softenin Softening basin 52
Carbon dioxide pH adjustment Recarb basin 37
| Media filters Filtration N/A Na
Chlorine Disinfection Filter effluent 1.0
Of the recommended items, zebra mussel control, coagulant and flocculation
polymer aids, and soda ash feed have not been incorporated into the treatment
process.
3. Phase I} WTP Improvements for Full Time Operation
Phase 11 WTP improvements are those needed to convert the Flint WTP from a back-
up supply to a primary supply plant. A number of improvements have already been
constructed as they were necessary to operate full time when treating water from the
Flint River. The improvements under the title of Phase II that have been completed or
are nearly complete include installation of the future raw water feed connection point
and valving for the KWA supply, upgrades to the lime sludge lagoon, the lime sludge
lagoon decant and disposal system, decant pump station and force main, installation
of mid-point chlorination before filtration, and upgrade of the electric feed sub-
station.
Additional improvements to the Flint WTP that are to be completed to become part of
the normal treatment process using water supplied by the KWA are:
e New oxygen and nitrogen storage facilities for the ozone system (under
construction)
e New coagulant feed system
e Electrical
o Pump Station #4 upgrades (under construction)
o SCADA and controls upgrades
o Filter transfer pump station feeders
e Pump replacements and VFD installation in the low and high service pump
station (under construction)
e Filter transfer pump station to Dort Reservoir
« Facility security improvements
Page 8 of 22 ion Logkwond, Andrews
A LEO A DALY COMPANY
xxxEND_PAGE:treasury01_b35_5648_6098_261
CITY OF FLINT
Operational Evalualion Report
February 27, 2015
SOURCE WATER EVALUATION
A. DATA ANALYSIS
Based on past data collected and the 2002 Treatability Study by AB&H, the Flint River
water quality varies seasonally with higher hardness and alkalinity experienced in the
winter. Higher magnesium concentrations are also experienced in the winter, adding
difficulty to the settling process due to neutrally buoyant floc. General water quality
average characteristics recorded for the 2002 Treatability Study as compared with average
characteristics recorded in 2014 are shown in Table 5 below.
TABLE 5 — FLINT RIVER WATER QUALITY CHARACTERISTICS
Period | Turbidity | TOC | Alk. PAG Total Col. | THMFP.
NTU Mg/l | Mg/ CaCO3 Count/day Mei
O01 870-1230
Apr—Oct 7.9 215 (7300 max) 410
10.3 1900-9000
: 5/22/14 (48,300 max)
The Flint River characteristics do not appear to have changed significantly over the past
10+ years. Note that further investigation by City staff revealed a sewer leak upstream of
the plant that may have contributed to the total Coliform count. The leak was
subsequently repaired.
B. CONCLUSIONS
Considering the minor changes in Flint River water quality, much of the information
contained in the 2002 Treatability Study by AB&H remains relevant today. Data from that
report assumed to be consistent today include the following:
e Flint River is influenced by groundwater from a dolomitic aquifer
e Hardness varies seasonally with higher hardness and alkalinity in the winter
e Hardness, alkalinity, magnesium concentrations tend to be reduced by run-off
In development of the 2002 Treatability Study, processes were simulated which resulted
in low THM levels. Therefore, information contained in that report will be used to assist
with establishing a baseline jar testing procedure as discussed further in Section III.
Page 9 of 22 1a yee prckaoad, Ancrew
ALEQA 3 iEG A DALY COMPANY COMPANY
xxxEND_PAGE:treasury01_b35_5648_6098_262
CITY OF FLINT
Operational Evaluation Report
February 27, 2015
1. TREATMENT PROCESS EVALUATION
A. EXISTING PROCESS DESCRIPTION
The existing WTP consists of an intake with screening from the Flint River, low lift
pumping, ozonation, rapid mix, flocculation, settling, softening, recarbonation, filtration,
storage and high service pumping. A process diagram is shown as Figure 1.
1. Intake
A 72" diameter pipe draws water from the Flint River through 2 traveling screens to
the low lift pump structure. No chemicals are currently fed for Zebra mussel control
or pre-oxidation as recommended by the 2002 Treatability Study. Manual removal of
zebra mussels is more economical than installation of chemical feed equipment
considering the short term need.
2. Ozone
There are 2 ozone generators designed to provide adequate ozone for a WTP flow of
up to 36 mgd. There are 3 ozone contact basins. The ozone generators were
designed to produce 900 Ibs/day at 10% concentration and up to 1300 Ibs/day at 6%
concentration each. Prior to recent repairs, readings indicated a production rate of
approximately 700 Ibs/day at 4% concentration. It is possible that before the recent
improvements the ozone feed might not have been optimized. !n fact, it is known
that less than optimal ozonation previously led to increased chlorine feed which
would have contributed to THM formation.
3. Rapid Mix
East and West rapid mix chambers allow chemical feed prior to the flocculation
basins, Each rapid mix chamber is equipped with a 5 hp mixer.
4. Coagulation / Flocculation
The WTP contains two equally sized flocculation basins, east and west, and each
basin provides tapered or gradually slowed mixing from inlet to outlet. There are
fifteen 2 hp mixers for each basin with VFDs to control mixing speed. The 2002
Treatability Study recommended feeding both coagulation and flocculation polymer
aids. Neither polymer aid is being used today because turbidity and TOC removals
have been sufficient to meet regulatory requirements.
5. Settling
Primary clarification takes place within 3 basins containing plate settlers. The settlers
are operating as designed.
6. Softening
Again, there are two basins for softening: east and west. Each basin is 120’ in
diameter and contains a solids contact softening unit. Each softening basin/unit has a
design capacity of 18 mgd. The east clarifier has an effluent weir imbalance that the
City intends to fix when low demands allow for construction. Low lift pumping
limitations, flow control to the basins, control restrictions on residuals removal, and
fluctuating demands have made it difficult for WTP staff to stabilize the softening
process. Softening is accomplished by feeding lime. The decision was made by the
City not to feed soda ash in order to remove non-carbonate hardness because
acceptable hardness levels could be achieved with lime feed only and softening is
short term until Lake Huron water becomes available. Lime and soda ash softening is
a possible consideration to assist with TOC removal and thus reduce THM formation.
A LEO A DALY COMPANY
Page 10 of 22 jan Lockwood, Andrewa
xxxEND_PAGE:treasury01_b35_5648_6098_263
CITY OF FLINT
Operational Evaluation Report
February 27, 2015
7. Recarbonation
Recarbonation for pH adjustment is accomplished in east and west recarbonation
basins between and to the north of the softening basins. Carbon dioxide storage and
feed equipment is located west of the recarbonation basins.
8. Filtration
Filtration is accomplished with 12 dual media filters, equally sized and designed to
filter 3.0 mgd each. Media consists of 12” of sand and 18” of anthracite. The filters
have been operated intermittently over the years due to the standby nature of the
WTP and until recently, chlorine injection took place downstream of the filters. It is
possible some microbial growth had developed in the filters leading to increased
chlorine demand. The City recently hired a contractor to upgrade the electrical
controls for the filters and that work has been completed.
9. Disinfection
Disinfection is provided by ozonation and by feeding chlorine. Ozonation occurs at
the front end of the WTP. Chlorine is fed prior to filtration and prior to finish water
storage / high service pumping. The intermediate chlorine injection location was
recently constructed under the Phase II, Segment 1 contract.
10. Clear Well & Pumping
The pump building sits adjacent to a 3 MG clear well and contains both low and
high service pumps.
JAR TESTS / EXPERIMENTS
1. Approach
There are several well practiced methods by which DBPs can be reduced. First, the
disinfectant can be changed to an alternate that has a lower tendency to form DBPs.
Second, additional treatment systems such as activated carbon or air stripping
(depending on the nature of the precursors) can be added to remove DBP precursors.
Lastly, the existing treatment processes can be optimized to remove as much DBP
precursor as possible. Of these options, optimizing existing treatment processes is the
only strategy that does not require the construction of new and expensive facilities. It
is anticipated that Flint will be receiving Lake Huron water in approximately two
years and this water will have a completely different chemistry from the Flint River.
Major process changes instituted to address THM levels using Flint River water are
likely to be unnecessary for Lake Huron water and may even be inappropriate.
Therefore, those options which require addition of new treatment processes are
undesirable at this time. In recognition of this upcoming change in water source,
efforts for this study have concentrated on improving the existing processes, rather
than adding new ones. New treatment processes will only be recommended if
operational changes to the existing treatment train prove ineffective.
Recent sample test results suggest that most of the DBPs are formed in the distribution
system rather than within the treatment plant. Therefore, the most logical approach is
to reduce the DBP formation potential (DBPFP) rather than simply lowering the levels
of DBPs leaving the plant. During bench scale testing, formation potential (FP) levels
were the primary indicator of success or failure of proposed process modifications.
2. Protocol
Bench scale pilot testing is intended to reflect actual plant operating and hydraulic
conditions so the bench scale treatment units were sized based on various
dimensionless factors to ensure the pilot treatment matched the actual system. Bench
Page 11 of 22 on Lockwood, Andrews
| lewnam,
A LEO A DALY COMPANY
xxxEND_PAGE:treasury01_b35_5648_6098_264
CITY OF FLINT
Operational Evaluation Report
February 27, 2015
scale ozonation was not practical due to time and cost limitations. Therefore, water
samples were withdrawn from the plant ozone basin effluent. These samples were
transported to the laboratory and dispensed into square testing jars. The jars were
used to simulate rapid mix, three-stage flocculation, and settling. Rapid mix and
flocculation conditions were matched to the plant based on “Gt” values. The “G”
value is a measure of the mixing intensity and is a function of mix time, viscosity of
the liquid, and mixing power applied to the water. “Gt” then, is a size scaling factor
where time has been accounted for. Settling time was scaled to match the shorter
settling depth of the testing jars. After settling, samples were decanted from the test
jars. The decanted samples were then lime softened; softening conditions were
similarly matched on the basis of “Gt”. Carbon dioxide was sparged into the samples
to reduce the pH. The water was then vacuum filtered through filter paper, sized to
simulate the plant's dual media filters. The samples were dosed with excess chlorine
and allowed to react for seven days at 25° C before testing for DBPs to determine the
formation potential.
The following conditions were applied during testing to properly match small scale
testing to actual plant processes.
TABLE 6 — BENCH SCALE TEST MIXING INTENSITIES
G
Plant
Rapid Mix
Flacculation, Stage 1
The primary variables during testing were chemical additions and chemical dosages.
Specific chemicals and dosages used for initial testing conditions were selected to
reflect current plant usage and the recommendations of the 2002 Treatability Study:
TABLE 7 — BENCH SCALE TEST CHEMICAL FEED RATES
Chemical Current Usage { 2002 Study |-. Test Values
QOzonation 4.66 mg 1.5 mg/l
Ferric Chloride 7.7 mep/l Fe3+ 40 mg/l Fe3+ 7.7 — 80 mep/l Fe3+
Coagulant Aid Polymer
Flocculation Aid Polymer
Powdered Activated Carbon
Not used 2.0 mg
Not used 0.05 mel | O- 0.05 mg/l
Not used N/A 20 — 100 mg
120 mep/l 175mel | 120-175 mg
Not used 52 mg 0-52 mg
Soda Ash | O-—52mei___
Cationic Softening Polymer 3.13 me/l Not used
Anionic Softening Polymer i g
Fluoride
Carbon Dioxide
Chlorine
32 mg/l 37 mg/l Fed A awe BH of
[10mg
A LEO A DALY COMPANY
Page 12 of 22 ian Lockwood, Andrews
xxxEND_PAGE:treasury01_b35_5648_6098_265
CITY OF FLINT
Operational Evaluation Report
3. Considerations
The 2002 Treatability Study did not note significant formation of DBPs. This may be a
function of different Flint River water chemistry at that time. However, recognizing
the considerable differences in chemical usage and dosages between that study and
current operations, those differences in chemical use and dosage are an obvious
starting point for optimizing treatment to prevent DBP limit exceedance.
Although it is believed that optimization of current treatment can correct the DBP
issue, should optimization of present treatment prove insufficient, alternate residual
disinfectants (chloramines and chlorine dioxide) will be investigated as additional
treatment measures,
4. LAN Test Results
Two rounds of jar testing were completed by LAN during the weeks of December 15,
2014 and January 26, 2015,Detailed test data is included in Appendix A. Testing
results can be summarized as follows:
e Increased dosages of ferric chloride resulted in higher reduction of THMFP.
e The currently utilized feed rate of lime at 120 mg/l is appropriate
¢ Softening with soda ash in addition to lime resulted in minor additional THMFP
reduction — in the range of 0% - 10%.
e The benefits of using a cationic polymer during softening at a dosage range of
0.31 - 3.13 mg/l to help reduce THM’s are unclear
¢ The benefits of using an anionic polymer during softening at a dosage range of
0.09 ~ 0.88 mg/l to help reduce THM’s are unclear
¢ Feeding oe was ineffective in reducing THMFP within the dosage range of 20
— 100 mg/l.
5. Testing by Others
In addition to jar testing completed by LAN, the chemical supplier who provides
ferric chloride for the City, PVS Technologies, ran tests using their recommended
flocculant polymer aid. Plus, Veolia Water completed jars testing of their own the
week of February 16” to analyze other process details and current WTP parameters.
Experiments completed by PVS Technologies showed very little TOC removal beyond
that obtained with straight ferric chloride feed.
6. Conclusions
Increasing the dose rate of ferric chloride is an operational change that can easily be
implemented without the need for any additional equipment. Test results show that
over 40% THMFP removal can be obtained with a dosage of 60 mg/l Fe3+ or higher.
Increased dosing of ferric chloride would be most ideal coupled with regular raw
water TOC monitoring so that TOC levels would dictate the appropriate ferric
chloride feed rate.
Softening with soda ash in addition to lime is another option the City should consider
if increased ferric chloride doses are not adequate to maintain THM levels under the
MCL, particularly during warmer months. Again, monitoring of TOC in the raw water
could provide useful information of when lime/soda ash softening is necessary.
Page 13 of 22 on Lociowood, Andrews
| lewnam,
A LEO A DALY COMPANY
xxxEND_PAGE:treasury01_b35_5648_6098_266
CITY OF FLINT
Operational Evaluation Reporl
February 27, 2015
IV. DISTRIBUTION SYSTEM EVALUATION
EPA guidance for the distribution evaluation portion of an OER is focused on identification
and isolation of a specific portion of the distribution system that led to the exceedance. The
circumstances of Flint’s apparent pending TTHM exeedances are unusual in that a new
supply has been implemented which clearly corresponds to the high TTHM sample results.
Although the new source is one element in increased TTHM levels, value remains in
evaluating the distribution system since water age is also a critical factor. Additionally, there
may be distribution improvements that can be made to help alleviate the problem.
A. INFRASTRUCTURE
1. Piping
According to the most recent MDEQ Sanitary Survey, the distribution system is
estimated to contain 70% cast iron, 20% ductile iron, 2% concrete and 8% steel
water mains. Unlined cast iron pipe can become pitted, allowing colonization sites
for microorganisms leading to chlorine demand. Additionally, much of the piping in
the system is aged and in poor condition. Increased chlorine demand could be
resulting from biofilm in older pipes and from main breaks/repairs, The extent of
contribution is not known but any water main replacement project will decrease
chlorine demand somewhat if constructed properly. Unfortunately, water main
breaks may also assist with decreasing water age by providing unintentional flushing.
All things considered, it is impossible to quantify the impact existing piping has on
THM formation.
The City utilizes City Point software and GPS equipment to document main breaks
and prioritize replacements. However, main break information is more pertinent to
rusty water complaints and has little relevance to THM levels. Areas that have been
targeted for main replacements include the transmission main from the WTP west to
Dupont and south to the West Side reservoir, Fenton Road, Atherton Road, Dort
Highway, Averil Street, and Boulevard Drive.
2. Storage
There are 4 finish water storage tanks and 1 raw water tank as tabulated below:
TABLE 8 — STORAGE TANKS
T Volume |_ Operating | - Absolute - |
ype (MG) [LWL | HWL | Bottom | OF
[DortReservor i
Dort Reservoir | Ground | Raw _| 2 2
WIP Tank Elevated | Finished 883.0 | 896.0 | 863.0
WIP Clear Well — PS #4 Finished 15” [708.5
Cedar Street Reservoir Finished 737.2
Finished 761.8 | 779.
The MDEQ typically recommends providing a minimum finish water storage volume
of 1/3 the maximum daily demand (MDD). According to the 2013 MDEQ Sanitary
Survey, the 5 and 10 year MDDs are 21.57 mgd and 30.05 mgd respectively. A
commen rule of thumb for clear well storage volume at a WTP is 10% of the design
flow rate. Another general guideline for reliability is to provide total storage to allow
for 2 X the average daily demand plus fire flow demand. For this analysis, fire flow is
assumed to be 2,500 gpm for either one industrial fire (2,500 gom) or a combination
of one residential (1,000 gpm) and one commercial (1,500 gpm) fire at a 4 hour
duration which results in a total volume of 600,000 gallons. These go-by
Page 14 of 22 on Second, butrews
[Q fo a DALY COMPANY
20
3
= Saas
12 =
12’ ()
xxxEND_PAGE:treasury01_b35_5648_6098_267
CITY OF FLINT
Operalional Evaluation Report
February 27, 2015
approximations are summarized below with the applicable flow rates and are
compared to the existing storage volumes currently being utilized.
Recommended Volume In
Common Practice Flow Rate Volume by ROT Use
Clear well 10% of Design Flow 18 MGD 1.8 MG 3.0 MG
Finish Storage = 1/3 MDD 30.05 MGD 10.0 MG 37 MG
Total Storage = 2 * ADD + FFD 13.87 MGD 28.3 MG 57 MG
Based on the values above, it appears the storage volume used in the Flint system
could be decreased without negatively effecting reliability. The appropriate volumes
of individual tanks will be further evaluated using the water system model and
discussed in Section V (C).
All reservoirs have baffling to minimize stagnant water. Also, all tanks have been
maintained and are in reasonable condition, The Westside reservoir has an exposed
roof that is in need of rehabilitation, but its current condition has no influence on
THM formation.
3. Pump Stations
All pump stations are in good condition but pumps are generally oversized. As an
independent consideration, oversized pumps are not a contributing factor to high
THM levels. Controf of pumps and pressure zones are discussed in detail below.
B. OPERATIONS AND MAINTENANCE
1. Pump Station & Storage Operations
Pump stations and storage tank levels are controlled as shown in Table 9.
TABLE 9 — PUMP STATION CONTROLS
Control Point
PS No. 4 Raw Operator
PS No. 4 Finish Match plant flow [ - | :
Westside System pressure (elevated tank) 22.57 | 33°
Cedar Street System pressure (elevated tank) 22.5" 33’
[Torrey Road Distr. — Brown/Bradle; <45 psi > 45 psi
Cedar Street and Westside pump stations are operated as needed and they are
alternated. Typically, Cedar Street is run in the morning and Westside is run in the
evening and reservoirs for each are filled during low demand periods at night.
Westside, Cedar Street and Torrey Road pump stations are used to boost system
pressure when high demands warrant it, but there are not well defined pressure zones
within the system. Therefore, the possibility exists that water is being recirculated
allowing for increased water age.
2. Booster Disinfection Practices
Booster disinfection is provided at the Cedar Street and Westside pump stations.
When sustained residuals are not provided by chlorine feed at the WTP, sodium
hypochlorite is applied at the reservoirs while being filled.
Page 15 of 22 Ion Lockwood, Andrews
ee
A LEO A GALY COMPANY
xxxEND_PAGE:treasury01_b35_5648_6098_268
CITY OF FLINT
Operational Evaluation Report
February 27, 2015
3. Changes in System Demands
Water demands in the City have been declining since the 1960's as the population
has dropped. As a result, many of the water system components are oversized
including storage tanks and water mains which both increase the time for water to
reach the user.