Director Wyant,
Attached is a draft of the letter we intend to send to Senator Ananich by close of business today, December 23. You will
note that we made some changes based on information we obtained at the December 18 meeting. We keep this
document confidential. We will copy you on the version we send to the Senator.
Take care. Doug.
xxxEND_PAGE:deq19_b02_0037_0259_001
Auditor General
Wee af dhe Aagslie
201.N. Washington Squaze,
SPR SGES cca
Sixth Hoor* Lansing, Michigan 48913 ° Phone: (517) 334-8050 ° www.audgen.michigan.gov
December 23, 2015
The Honorable Jim Ananich
Senate Minority Leader
State Capitol, Room S-105
Lansing, Michigan
Dear Senator Ananich:
taber 20, 2015 letter to our
king Water and Municipal
specific to lead
additional questions
Enclosed are answers to the questi
office regarding the audit we are
Assistance (ODWMA), Departmen
contamination in the City of Flint's
you posed in yo
of the Office
If
contact our office“
Doug Ringler
Auditor General
xxxEND_PAGE:deq19_b02_0037_0259_002
Questions and Answers
Q1: How does ODWMA ensure the data it receives is accurate?
A: With regard to the United States Environmental Protection A
and Copper Rule (LCR) monitoring requirements, DEO relie
controls to ensure the accuracy of test results:
cy (EPA) Lead
e following key
e State-owned laboratories test water sample
e State-owned laboratories send test result
e The City of Flint Water Treatment
sample sites are classified as tie
The current Flint WTP LCR sampling
1. DEO informs the Flint WTP o
sample size.
ater lead and copper
sites for sampling.
receives water lead and copper sample results, which include
ollowing information: date collected, date received, address
ere collected, type of residence (e.g., single family or apartment),
d sample point (e.g., kitchen sink or bathroom sink).
DEQ tracks, and follows up if necessary, the number of samples
collected by the Flint WTP to help ensure that the required minimum
number of samples are collected by the monitoring period deadline.
. Single-family or multiple-family residence with lead service line, lead solder copper piping
constructed after 1982, or lead plumbing.
Page 1
xxxEND_PAGE:deq19_b02_0037_0259_003
12. The Flint WTP submits lead and copper report to DEO that certifies
whether sample sites meet tier 1 criteria.
13. DEQ prepares the LCR 90th percentile calculation report.
During our review, we noted two potential improvements for the Flint WTP
sampling process (see Question 5 of the additional questions answered by the
OAG regarding tier 1 sample validity):
ly tier 1 sample
the LCR
deral Regulations
e DEO could verify that the sampling pool was limite
sites to ensure that the Flint WTP is in complia
(Title 40, Part 141, section 86(a)(3) of the C
[CFR]).
tifications.
e DEQ could independently verify the validity*6f sample
data
Q2: What accountability measures are in plac WMA staff who fail to
verification protocols?
WTP and, therefore,
lint WTP's LCR data
col for lead and copper
P certifies samples
see steps 7 and 12 in
above). We did not
y that submitted
A: DEQ doés not provide any direct over
does not have any accountability measures
verification protocols. DEQ's data verificat!
water sampling is limited.4a. verification that
submitted to the State-
the Flint WTP LCR samp’
identify any instances in Ww
samples were certified by
Q3: t ODWMA staff who lie or
MA staff must adhere to the rules and
Civil Service Commission. If any
present information to the EPA, they
Director, =
extending
points and
uty Director, ODWMA Chief, and other key ODWMA staff)
ik to January 1, 2013. We did so to identify the key decision
ersations that occurred leading up to and through the situation in
ew was also intended to determine whether State, Flint, or other
pted to conceal key test results or other information.
one e-mail exchange between DEO and the EPA that appears to be a
ant contributor to the concern that DEO misrepresented information to
the EPA. The EPA requested clarification on February 26, 2015 regarding the
type of optimized corrosion control treatment the Flint WTP was using. DEO
responded on February 27, 2015 that the city had an optimized corrosion
control program in place, but DEO did not provide any program details. DEO
informed us that the Flint WTP corrosion control program included performing
Page 2
xxxEND_PAGE:deq19_b02_0037_0259_004
lead and copper monitoring for two consecutive six-month periods to determine
whether corrosion control treatment would be necessary in the future.
However, it appears the EPA interpreted corrosion control program to mean that
corrosion control treatment was being performed.
On April 23, 2015, the EPA again inquired as to what the Flint WTP was doing
for corrosion control treatment. DEQ responded on April 24, 2015 that the Flint
WTP was not practicing corrosion control treatment.
Based on our review of this and other e-mails, we have n
believe that DEO willfully misrepresented the informati
ic reason to
e EPA.
Q4: What policies do DEO and ODWMA have in place to.
the chain of command?
A: We did not note any instances of major i
policies, laws, regulations or specific di
the course of our review. DEQ doe
place to escalate major infractions p
our review of DEO correspondence co
chain of command related to the Flint situa’
policy is for staff to notify the proper level
determine necessary actios
stated that its informal
ement of infractions to
Page 3
xxxEND_PAGE:deq19_b02_0037_0259_005
Additional Questions
Answered by the OAG
Application of the LCR
Q1: How did the Flint WTP become the primary water supplier for the;City of Flint?
A: Upon notification of the City of Flint's plans to switch t
Authority (KWA) in April 2013, the Detroit Water and
(DWSD) submitted a letter to the City of Flint statin
agreement to provide water services on April 17
aregnondi Water
ge Department
ould terminate its
According to DEQ management, the Flint WT,
DWSD to maintain it as the City of Flint w
negotiations were unsuccessful, the Cit
request of its intent to operate the Fli
Although the Flint City Council vote
KWA pipeline, the vote was silent on
drinking water source.
ackup source of water for
emergency purposes p SD WTP to the Flint WTP.
On a quarterly basis, th
Flint WTP, which met
test the water's effect on
Q2.
how to apply the LCR prior to
nth monitoring periods of the Flint WTP
experiences applying the LCR monitoring
propriately applied the LCR
ods with acceptable lead levels. However, a water system that
corrosion control, and which has treatment in place, should
operate and maintain optimal corrosion control treatment.
ff explained that they did not treat the switch to Flint River water as a
new system, but as anew source. DEQ further stated that because the Flint
River was a new water source and there was a change in chemicals needed to
treat the new source, a corrosion control study was needed to determine the
impact on the water distribution system. Therefore, it was DEQ's interpretation
that two rounds of six-month monitoring were still needed to evaluate the water
quality and determine optimal corrosion control treatment.
Page 4
xxxEND_PAGE:deq19_b02_0037_0259_006
Q4:
The Flint water system had optimal corrosion control treatment when the DWSD
WTP was the water supplier. Based on our review of notes from a July 21,
2015 EPA and DEO conference call on DEQ's implementation of the LCR
regarding whether the Flint WTP should have continued to maintain corrosion
control treatment, it appeared that the EPA did not agree with DEQ's
interpretation of the LCR. Region 5 EPA staff explained that they would talk to
the EPA headquarters about the interpretation of regulations and believes that
systems that have been deemed optimized need to "maintain" corrosion control.
The Region agreed to provide supporting regulatory citation
about maintaining corrosion control.
On November 3, 2015, the EPA issued a memorand g that the LCR had
differing possible interpretations; however, the EP. iti
important for large water systems to take the st
appropriate corrosion control treatment is mai
that public health is protected. Based on t
corrosion control treatment should have
Should DEQ have required the Flint WT|
control treatment after the first round of
lead action level of 5 parts per billion (ppb)?
ig results were above the
A: Yes. According to DEQ'
end of the monitoring p
acceptable lead level, D
optimized corrosion cont
thin six months after the
results exceeded the
WTP to start pursuing
The LCR states that the lea i if the lead level, as
i er than 15 ppb. If the
e water treatment. However, for water
corrosion control treatment, they can be
© consecutive six-month periods below 5 ppb.
ve notified the Flint WTP to start pursuing optimized
ent. However, DEQ waited until the second round of
“completed (June 30, 2015) to assess whether water sample
corrosio
sampling
results im
rify that only tier 1 sample sites were selected by the Flint WTP in the
Is of six-month samples?
A: DEQ did not verify that only tier 1 sample sites were selected. DEO relies on
the Flint WTP's certification of sample sites and does not perform any
independent verification of those certifications.
Page 5
xxxEND_PAGE:deq19_b02_0037_0259_007
In a November 19, 2015 Flint Journal article, the Flint WTP indicated that it did
not have the ability to ensure that all sites were tier 1. In fact, water samples
came from the random distribution of 175 testing bottles without regard for
whether the homes were at risk for high lead levels. DEQ issued a formal
memorandum on November 9, 2015 requesting that the Flint WTP verify the
classification of all prior sample items. The results are due back from the Flint
WTP on December 30, 2015.
Q6: DEO dropped two water sampling sites from its second six-mgi
(January 1, 2015 through June 30, 2015). Was this app
mple
A: Yes, it was appropriate for DEO to drop these two
regulation 40 CFA 141.86(a) states:
ling sites. Federal
. each water system shall compl
of its distribution system in order t
This regulation also requi
consist of only tier 1 s
monitoring requirements:
argeted sampling pool
mber is available to meet
According to federal regul
water sample if it determin taken rom a site that did not
meet the site selection crit
hat does not meet the tier 1 requirements
y residence. The second sample site
7: i ne night before drawing a sample an appropriate sample
equires that samples be a first draw of water after six hours of
he LCR does not indicate whether or not the water line should be
0 collecting the sample. In the sample instructions, DEQ required
to ensure that sampled faucets were not stagnant for an excessive
'f time beyond the targeted six hours (e.g., rarely used faucets or when a
wner has been gone for an extended period of time.)
The LCR requires six hours of stagnation; however, it does not preclude DEO