This post is entirely focused on effective, fast, and economical solutions to the problems facing the people and businesses of Flint.

It is not intended to start a conversation about blame, those conversations are ongoing across the web, and I am hoping that the guilty will face harsh criminal and civil charges. Comments here which focus on blame will be deleted, because I want to talk about problem scope and solutions.

Scope of the problem:

From Wikipedia’s Flint page
“2010 census … There were 40,472 households of which 34.3% had children under the age of 18 living with them … The average household size was 2.45 and the average family size was 3.13…”

The greatest immediate threat is to households with young children, who are most impacted by lead exposure, probably about 10K households, for which single-tap filtration is untenable, since even washing(?) children in the city water puts them at risk. Of course all of Flint’s households and businesses are at risk, and not just from lead, since the poor water quality water is also thought to be causing a large spike in Legionella. Wikipedia on the water crisis

Recent testing:

The most recent sampling of lead levels in Flint water showed levels above the filterable limit: “…26 of more than 4000 samples … collected since late December..” Those 26 samples (26 homes?) measured between 153 and 4000 ppb, “…no concentrated area with spiked lead levels in Flint in the most recent round of testing…” Lansing State Journal

“…According to health officials, water lead levels at 3,900 other sites were considered safe…” CBS News

With about 40,000 homes in Flint, the sampling was pretty impressive. (No really, they managed ~100 samples/day, with almost no resources, while facing active opposition.) It seems they’ve sampled about 10% of homes since mid-December, if the news reports are correct. That suggests that only about 260 homes in Flint are egregiously poisonous. <sarcasm> It took them about a month to find the first 26, so within a year they should be able to find all the other poisoned residents. </sarcasm>

Statistically the sampling results make little sense. Why were there no neighborhood spikes, and why were no samples found with lead levels between 15 ppb and 153 ppb? The most reasonable explanation might be that those 26 samples came from homes that had been plumbed internally with lead, but since Flint acknowledges that the water distribution system has a lot of lead piping, and we’ve no intermediate samples, the study becomes a little suspect. I don’t know if the study has been or will be published on the internet, if anyone has a link, or better understanding, please comment. Or perhaps the reports are accidentally(?) misleading, conflating the post-filtering and pre-filtering numbers.

Safe levels of lead:

The EPA current standard is that no level of lead is ‘safe’, but levels above 15 ppb require alleviation.

NFS/ANSI 53 (NFS53) is a standard for filter efficacy that covers many contaminants including lead. Oversimplified, the lead standard requires that for the filter’s claimed life (typically a few thousand gallons) it reduces lead solids and dissolved lead by ~93%. The test requires that a maximum ‘Influent challenge’ of 0.15 mg/L (equal to 150 ppb) should be reduced to (at most) an effluent of 0.010 mg/L (10 ppb). The test does NOT specify the filter’s behavior for influent levels above 150 ppb. This is why the headlines from January 30 claimed the filters would not be good enough, there were samples well above the challenge level. Detroit News

One thing that my research has yet to clarify is whether claiming compliance with NFS53 implies your filter removes ALL NSF53 listed contaminants, or only those you explicitly claim. I raise this because frequently tech datasheets for filters claim NSF53, but then only list protection for a few contaminants. Few list lead explicitly. The NSF provides a list of all certified filters that do pass that lead test, though it’s not as helpful as one might hope: NFS lead certified filters

Issues with the current filters:

The single stage faucet-attached filters being distributed in Flint are compliant with NFS/ANSI 53, but in a house with a lead level of 4000 ppb, you’d need three of them in series to lower the level to a reasonable 4 ppb, and the first filter would probably fail by clogging or bypassing within a fraction of its published lifespan. The bottom line is that nobody can really say what happens when even a good filter faces lead levels much higher than that the NSF expected to appear in the input stream.

Costs and values:

Estimates to repair water infrastructure have ranged from $60 million to $1.5 billion, excluding costs for medical remediation of victims.

One expectation is that given time, and CORRECT water treatment, much of the lead and iron plumbing infrastructure will recover and stop leaching out, by becoming recoated (with calcium?) That’s not expected to be particularly expensive treatment, a few hundred a day or less for the entire city system. (Too tired to re-find the references.)

The median value of a home in Flint is $33,000 according to Best Places, though I suspect it’s really much lower lately. One interviewee on a recent Rachel Maddow segment estimated the current value of homes in the nicer neighborhoods of Flint to be closer to $10,000.

So … 40K homes times $30K value = 1.2 billion in 2015, but maybe only 400 million in 2015. No I’m not proposing that the feds buy all of Flint and plow it underground. No matter what they were paid, the citizens of Flint would be screwed.

On the other hand, high end, multistage, whole house filters cost less than a few thousand dollars, and another thousand a year to maintain. So that’s $50-80 million the first year, to fit every house in Flint, then a decreasing bill of $40 million/year. as (and if) the water quality improves, and the filters don’t need such frequent replacement.

I’ve got lots more to puzzle out over this, but fitting houses with decent filters seems faster, safer, and cheaper than ripping up most of the streets and yards of Flint as the residents are slowly poisoned.


This is a job for Amazon and the thousands of underemployed plumbers, pipe fitters, and machinists in the rust belt. Who knows Jeff Bezos? Rachel help me out?

Too tired, back soon…

More Links:
Filter references
NSF/ANSI 53 lead standards:
The current standard doesn’t seem to be published online, but here is a draft of the lead section: