An audit of the unit within Michigan’s Department of Environmental Quality that’s responsible for making sure water systems are following drinking water regulations declares the state’s oversight is “not sufficient.”
The report was released Friday morning by the Michigan Office of Auditor General.
The office confirmed earlier reports that MDEQ did not follow the federal Lead and Copper Rule after Flint switched drinking water sources in 2014 in two ways. First, MDEQ did not require Flint to properly treat it’s water to help prevent corrosion of heavy metals, like lead, from getting into people’s tap water. Second, MDEQ failed again to require proper treatment in March 2015, after a series of tests showed an increase in the number of homes with higher lead levels – above a federal threshold that requires corrosion control treatment. Instead, MDEQ waited until August 2015, to ask Flint to begin this treatment.
Outside of Flint, the audit found other problems with MDEQ’s oversight.
MDEQ does not verify that water samples for lead come from the right homes
In order to figure out if a water system’s lead levels are okay, systems have to test certain homes that are considered to be at the highest risk of lead in the water. These are homes with lead service lines or with lead solder, if those homes exist in the individual water system.
Water systems are responsible for certifying that the homes they test meet this criteria. But the audit found MDEQ doesn’t verify that, and essentially just takes a water system’s word for it.
In Flint, MDEQ asked the city last November to verify 324 homes historically used for sampling met the sampling criteria. Flint reported that all the homes did meet the criteria. MDEQ has reviewed only 46 of those 324 sample sites and found only 13% of the 46 sites actually met the criteria under the federal law.
“To help ensure that community water supplies sampled appropriate sites, DEQ should proactively and routinely validate sample sites selected by the water supplies,” the audit says.
MDEQ doesn’t have a procedure for determining the population served by a water system
This is important because water systems that serve different sized populations have to follow different sets of rules.
In Flint’s case, MDEQ required the city to get 100 samples during an initial round of testing in 2014 because the city had over 100,000 people. Flint was supposed to collect 100 samples in a round of testing done in 2015 as well. But when the city was only able to collect 69 samples, MDEQ decided Flint didn’t need 100 samples after all, citing U.S. census data that showed Flint’s population has dipped below 100,000 people.
The the audit found MDEQ had no written procedures for determining population size. Outside of Flint, it found MDEQ used inconsistent sources of information to determine population, including “undocumented estimates of occupancy and verbal responses from the community water supplies.”
MDEQ didn’t follow its own policy for water system visits and detailed surveys
Once every three years, MDEQ is supposed to complete a detailed “sanitary survey” for each water system. These in-depth reports include an on-site review of a water system, including the water treatment, distribution system, pumps, monitoring, reporting and “data verification for existing or potential health hazards.” These surveys determine how well a system is able to “produce, treat, and distribute” enough water and that it meets all drinking water requirements.
The audit found that MDEQ didn’t finish these surveys for 10% of the community water supply systems on time. Of those 137 surveys that were late, nearly 20% were a year or more past due; one survey was more than 5 years past due.
Another MDEQ policy requires state staff to visit each water system in person, between one and four times a year. The visits are supposed to give MDEQ staff a chance to follow up on any concerns outlined in the sanitary surveys and to become more familiar with the water systems.
The audit found MDEQ did not visit 15% of the water systems within the required timeframe. Of those 214 systems where an MDEQ visit was past due, nearly 20% were late by a year or more.
In a written response to these two audit findings, MDEQ noted that the U.S. EPA does not require site visits and that, at 90%, the state agency beat the federal agency’s target (79%) for getting sanitary surveys completed on time.
MDEQ fees do not cover the costs of oversight
The audit also found that the fees collected from water systems, along with state and federal funds, were not enough to cover the cost of overseeing the systems. In fiscal year 2014 the difference was nearly $2 million. In 2015, it was $1.6 million.
Audit also faults federal Lead and Copper Rule
“Through this audit, we became aware of many instances in which sole reliance on the LCR may not serve the best interest of Michigan citizens.”
- No requirement to monitor human exposure to lead
- No requirement to monitor at hospitals or schools or other facilities whose occupants are more susceptible for health effects from lead.
- Resident-collected water samples could cause variations in results
- 90th percentile calculation is the sole means to determine lead action levels “Regardless of the 90th percentile lead level reported for LCR compliance purposes, DEQ should further assess the sample results to identify warning signs of potential system wide problems. For example, for the Flint WTP, the number of samples with lead levels of 5 pbb or higher increased for 17 of 100 (17%) samples in 2014 to 28 (41% of 69 samples) in 2015. This increase could indicate that the City of Flint’s switch from water treated with a phosphate corrosion inhibitor to water with no added corrosion inhibitor caused excessive corrosion throughout Flint’s distribution system that needed immediate attention.
- LCR lacks guidance on necessary actions if DEQ becomes aware that a prior 90th percentile calculation was based on inaccurate information “Inaccurate samples sites used in 90th percentile calculations could delay or prevent DEQ from taking necessary actions to protect the drinking water supply. Therefore, DEQ should recalculate 90th percentile calculations when it determines that sample site or other information was inaccurate.
- LCR allows implementation of corrosion control treatment to occur over several years
- Partial replacement of lead service lines is ineffective
MDEQ declined an interview request for this story, instead issuing this written statement from Keith Creagh, the interim director of MDEQ:
“The DEQ appreciates the OAG’s thorough review of the Public Water Supply Programs within the Office of Drinking Water and Municipal Assistance. The department is committed to developing and implementing process and program improvements to address the findings in the report.”
