Our sampling of 252 homes demonstrates a high lead in water risk: Flint should be failing to meet the EPA Lead and Copper Rule

Over the weekend, we analyzed all samples shipped to Virginia Tech from Flint to date. Flint residents have already returned an astonishing 84% of the sample kits we sent out (252 out of 300 samples). We will continue to analyze water samples as they are returned. However, mathematically, even if the remaining 48 samples returned have non-detectable lead, our conclusion will not change — FLINT HAS A VERY SERIOUS LEAD IN WATER PROBLEM.

Forty percent (40.1%) of the first draw samples are over 5 parts per billion (ppb). That is, 101 out of 252 water samples from Flint homes had first draw lead more than 5 ppb. Even more worrisome, given that we could not target “worst case” homes with lead plumbing that are required for EPA sampling, Flint’s 90%’ile lead value is 25 ppb in our survey. This is over the EPA allowed level of 15 ppb that is applied to high risk homes. This is a serious concern indeed. Several samples exceeded 100 ppb, and one sample collected after 45 seconds of flushing exceeded 1000 ppb.

We now advise Flint consumers to heed EPA information that advises consumers on how to avoid adverse health effects from exposure to excessive lead in drinking water. The main concern is related to water used for drinking or cooking. With the exception of one home that we sampled which had astronomical levels of lead, the levels of lead detected in Flint were safe for bathing, showering, toilet flushing and watering lawns/gardens.

Until further notice, we recommend that Flint tap water only be used for cooking or drinking if one of the following steps are implemented:

  • Treat Flint tap water with a filter certified to remove lead (look for certification by the National Sanitation Foundation (NSF) that it removed lead on the label), or
  • Flush your lines continuously at the kitchen tap, for 5 minutes at a high flow rate (i.e. open your faucet all the way), to clean most of the lead out of your pipes and the lead service line, before collecting a volume of water for cooking or drinking. Please note that the water needs to be flushed 5 minutes every time before you collect water for cooking or drinking. For convenience, you can store water in the refrigerator in containers, to reduce the need to wait for potable water each time you need it.

We do not issue this warning lightly, and note that our concern is based on several lines of evidence. First, scientifically, we predicted based on past research that the Flint River water chemistry would create a serious lead in water problem. Second, we confirmed the very high corrosivity of the Flint River water for lead in our laboratory testing at Virginia Tech. Third, for some reason that no one has yet explained to us, the Flint River water was introduced into the pipe distribution system without any measures (or even a plan) to reduce its corrosivity. We are therefore very perplexed by recent MDEQ assertions that the situation in Flint is normal. Finally, we have the results of our survey of 252 homes conducted with the assistance of Flint consumers. Because of the very serious and permanent health damage that arises from lead exposure, we feel that this problem requires immediate public health warnings and intervention– we provide that for Flint consumers in this report.

Another mystery that must be examined very carefully in the days and weeks ahead: How is it possible, that Flint “passed” the official EPA Lead and Copper Rule sampling overseen by MDEQ? In our experience, following the EPA site selection criteria targeting homes with the highest risk for lead, the MDEQ sampling should have found much worse results than our sampling. Instead, MDEQ is asserting that the lead levels in Flint are much lower. Hence, we call on the U.S. EPA and others, to conduct a detailed audit of the 2014 and 2015 LCR sampling round overseen by MDEQ in Flint, to determine if it was conducted consistent with requirements of the law.

Primary Author: Dr. Marc Edwards

Samples Analysis: Dr. Jeffrey Parks, Anurag Mantha

Acknowledgements: Siddhartha Roy

Our Flint Sampling Trip (Aug 17-19 2015)

Objective: In response to the possible public health problems and infrastructure degradation that is occuring in the City of Flint, we organized a field trip for extensive sampling to better assess water quality concerns. This trip was planned and executed immediately after we finished mailing out 300 lead sampling kits to Flint residents.

Personnel: Our entire team of undergraduate/graduate students and scientists contributed in the time period leading up to the trip. Dr. Marc Edwards, Rebekah Martin, Dr. David “Otto” Schwake, Colin Richards, and Min Tang were the team selected to drive to Flint from Blacksburg VA. Emily Garner spent over 80 hours preparing a comprehensive biological sampling kit to look for OPPPs and bacteria that cause corrosion, while Dr. Jeff Parks and Anurag Mantha led up efforts to equip the FLINTWATERSTUDY van as a mobile lab.

Day 1 (Aug 17 2015)

Our trip to Flint MI from Blacksburg VA took eleven hours after which we made our first stop at Mrs LeeAnne Walters’ house to pick up sample kits supplied by Genesee County Health Department for measuring Escherichia coli and Coliforms. E.coli and coliforms are common (fecal indicator) bacteria and our analyses showed no presence of these in the water. We also tested for chlorine residuals in her tap water and were introduced to her family.

Ms. LeeAnne Walters shows Dr. Edwards a used Granular Activated Carbon (GAC) filter that was full of rust seven days into use
Ms. Walters shows Dr. Edwards a used filter that was full of rust seven days into use

Day 2 (Aug 18 2015)

We spent about five hours sampling at 10 sites, most of which are used by the city for their distribution system monitoring. These included the Flint River, a McDonald’s franchisee next to the water treatment plant, and another eight distribution monitoring sites typically used by the city. We took 1.5 L samples for microbial analyses and 500 mL samples to analyze the water temperature, dissolved oxygen and chlorine residual. These samples were processed immediately in our mobile lab. The purpose of this approach was to gather data on water quality at different points along Flint’s distribution system, and to verify similar data that was being collected by the city on the same day.

We then shipped the biological samples overnight to our Virginia Tech labs in Blacksburg, and submitted the E. coli and coliform samples to the Health Department for analysis in the afternoon. We later had dinner with Flint residents and community leaders, and interviewed them about their experiences with Flint water (to be uploaded).

Day 3 (Aug 19 2015)

We divided ourselves into three groups to maximize sampling. The first group sampled four businesses still on Detroit water, the second sampled four homes that are on Flint water, and the final group sampled two homes where residents had reported health issues. Chlorine levels were also monitored for one place with Detroit water and another with Flint water. We stayed up late to sample what happened to the water at night. The Detroit water had consistent chlorine residual of about 0.5 mg/L whereas the Flint waters had no residuals at 3 am.

We wrapped up sampling around noon, said our goodbyes to Ms. Walters and her family, and set off for the long journey back home.


Figure 4. The sampling sites with Detroit water (A-D) and Flint water (101-104, 11, 13). The water from one resident had white precipitates which needs further investigation.
The sampling sites with Detroit water (A-D) and Flint water (101-104, 11, 13). The water from one resident had white precipitates which needs further investigation.

Primary Authors: Min Tang and Colin Richards

Acknowledgements: Siddhartha Roy, Dr. Marc Edwards

Why is it possible that Flint River water cannot be treated to meet Federal Standards?

The key problem is that water from the Flint River is highly corrosive to iron and lead, and, unfortunately, these pipe materials are widely used throughout Flint. Herein, we discuss the higher corrosivity of Flint River water to iron, and the associated problem maintaining chlorine disinfectant in the system.

Why is Flint River water more corrosive compared to Detroit water? Flint River water has about 8 times more chloride (Cl) in it than Detroit water. Chloride is generally considered to be very corrosive to iron. For instance, chloride present in road salts applied in the winter causes iron in cars and bridges to rust. Detroit also adds a corrosion inhibitor chemical (orthophosphate) to their water that helps to reduce corrosion of metals such as iron and lead. So, current Flint water is not only more corrosive, but there is also no corrosion inhibitor present.

Iron corrosion can cause a serious problem for meeting Federal standards using Flint River Water, because iron corrosion consumes chlorine. Chlorine is added to the water to prevent growth of microorganisms that cause disease, and maintaining a chlorine residual is the best way to protect public health against human pathogens.

To illustrate how iron corrosion is problematic for maintaining a chlorine residual, we collected a sample of Flint water. We put some of that sample into a clean glass container, and some more of it into an identical container with a piece of iron to simulate the effect of water on iron pipes in Flint. We then measured the decay of chlorine over time (Figure 1). The initial level of chlorine was 1.15 milligrams per liter (mg/L), and it stayed pretty high when it was in a glass container. It only dropped to 0.95 mg/L over 12 hours (blue line). Thus, if Flint had a glass (or plastic or concrete) pipe system, chlorine would stay high as it was transported to homes. But when we did the same test with iron present, the chlorine dropped faster due to the corrosion, and was all gone after only 12 hours (red line). After 6 days of doing the test, the chlorine dropped even faster, and was below the minimum required chlorine residual of 0.2 mg/L in just about 1 hour in our test system (green line). The key point is that Flint River water is corroding iron pipes, and that will cause the chlorine to disappear very quickly.

Figure 1: Decaying of free chlorine in Flint River water with and without iron
Figure 1: Decaying of free chlorine in Flint River water with and without iron

When we did the exact same test using water from Detroit (collected on the outskirts of Flint), the chlorine in the container without the iron stayed high at about 0.6 mg/L over a 12 hour test (blue line below). When iron was present, chlorine dropped, but much slower than in the Flint River water (red line and green line). Ironically, Detroit water started with only about 60% of the chlorine initially present in Flint water, but ended up with much more chlorine after 12 hours because there was less iron corrosion. What is more, comparing Day 1 to Day 6, Flint water ate up more chlorine the longer the test was run, whereas Detroit water ate up less chlorine the longer the test was run.

Figure 2: Decaying of free chlorine in Detroit water with and without iron

Figure 3 shows a picture of Flint water and Detroit water after reacting in the glass containers described after the first 5 days of the test. This also illustrates why residents have been complaining of “red” or discolored water after the switch to the Flint River source.

Figure 3: Higher release of iron is evident in the Flint water glass reactor containing iron than that of Detroit water
Figure 3: Higher release of iron is evident in the Flint water glass reactor containing iron than that with Detroit water

At present, as a rough estimate which we will elaborate on later, it looks like Detroit water is about 5 times less corrosive to iron pipe than Flint River water. This also probably means that the iron pipes in the city of Flint system will fail 5 times faster using the Flint River water rather than the Detroit water. In fact, this is probably already occurring as evidenced by increased rates of water main leaks and breaks. While an economic analysis cannot yet be done based on our limited data, it is possible (and even likely) that the economic damage to the Flint pipe system due to corrosion is going to cost the city tens of millions of dollars more in pipe repair costs in the coming years compared to what they would have paid if they had stayed on Detroit water.

Conclusion: The high rates of iron corrosion from using Flint River water as a drinking water source are damaging the Flint distribution system. The corrosion is also causing chlorine to disappear quickly, which may make it more likely for harmful bacteria to grow in the water. Furthermore, it is possible that with the existing unlined iron pipe system in Flint, and the relatively low water demand (due to declining population, loss of GM – which used a lot of water – as a water customer, and high rates), that it will very difficult to meet Federal standards for minimum chlorine levels no matter what is done to treat the water.

Primary Author: Dr. Marc A. Edwards

Acknowledgements: Siddhartha Roy

Opportunistic Pathogens (OPs): #1 cause of waterborne diseases in the United States

We have long known about the dangers associated with waterborne pathogens. Fecal contamination of drinking water sources can lead to gastrointestinal illness and deaths caused by a variety of waterborne pathogens. We control for these risks by 1) selecting clean water sources for treatment, 2) treating the water to remove particles and bacteria by filters, and 3) disinfecting water with chlorine. Treatment plants have to prove they are meeting very high standards in protecting the public from this danger, by frequently measuring filter efficiency and chlorine levels at the treatment plant. In addition, utilities also have to double-check their control of this threat, by sampling their water pipelines for indicator bacteria which are present at very high concentrations in feces, such as coliforms and E. coli, and also maintaining significant levels of chlorine as water is transported to homes.

In recent decades, we have also come to recognize that fecal contamination is not the only source of waterborne pathogens in our drinking water. These other bacteria do not come from fecal contamination, but rather they grow in pipes and homes themselves, and are commonly referred to as Opportunistic Pathogens, or OPs for short. The presence of OPs in drinking water is a danger that is not directly addressed by existing Federal Regulations, despite the fact that an OP, Legionella pneumophila is now known to be the most frequently reported causative agent of waterborne disease outbreaks (and deaths) in the United States.

The impacts that OPs can have on human health are numerous, and they generally do not come from drinking the water. For example, Legionella pneumophila (the bacterium which causes Legionnaire’s disease) and Mycobacterium avium live in pipes, and human exposure occurs when consumers breath tiny water droplets in the air from showers or washing hands. Staphylococcus aureus can lead to skin infections when it contacts open cuts or existing irritations. Pseudomonas aeruginosa can lead to a variety of infections: lung and blood stream infections, skin irritations, and infections of the eye or ear.

Legionella bacteria (Image courtesy: CDC)
Legionella bacteria (Image courtesy: CDC)

Most OPs are not regulated or routinely measured in drinking water. Legionella is only regulated in water as it leaves the treatment plant, which is the location that is least likely to have high levels of Legionella, and the regulation does not do enough for public health protection. However, the conditions which are associated with a typical building’s plumbing, can create ideal conditions for these bacteria to take hold, such as warm water, so Legionella can be present at very high levels in homes even when it is not detected at the treatment plant.

Residents of Flint have also reported many symptoms (see, here, here and here) which are consistent with those associated with some infections caused by waterborne OPs, such as skin irritations. However, it is unclear whether OPs may be contributing to this problem due to a lack of water testing, and difficulties in linking human health problems to water exposure. There is some indication that in March 2015, the Health Department was exploring possible links between drinking water and cases of Legionnaire’s disease in Genesse County.

Additional water testing targeting OPs, particularly of samples collected at the point of use within homes, could offer valuable information regarding the microbial water quality of the drinking water available to residents of Flint.

Primary Author: Emily Garner

Acknowledgements: Dr. Marc Edwards, Dr. Brandi Clark and William Rhoads