The Mercurial Trail, Part 2: The Spill Spreads

Tuesday, February 17, 2015, 8:52 am
Alice Dreger

Image: fall leaves along the Red Cedar River

This article picks up from "The Mercurial Trail, Part 1: The Spill."

A spill of mercury, like the kind that occurred at the East Lansing Wastewater Treatment Plant in November 2013, demonstrates why mercury is named after the Roman God known for speed, and why it used to be called “quicksilver.” When you spill mercury, the little silver balls that result race around the room as if there is no friction. Tracking down all the bits can be tricky.

Whomever tried to clean up the mercury spill Friday, November 22, 2013, at our wastewater plant may have thought they’d gotten rid of it all with the shop vacs and duct tape they employed, along with their attempts to wash some of it down the sink.

But what these actions did was not so much contain the spill as present multiple opportunities for it to spread further—via the shop vacs, the duct tape, the plumbing, and more.

What happened when some of the mercury was flushed down the maintenance room sink? Like all wastewater in the city should, it presumably ended up as waste “influent” in the wastewater plant. In other words, it mixed into all the other wastewater of the city.

Data from the plant suggests it did just that—and that, in fact, it caused a spike in mercury readings of wastewater at the plant.

This spike went unnoticed for four months, only being noticed after the spill was reported.

The City is required to do regular readings on mercury and other contaminants in wastewater influent (the waste that comes to the plant) and effluent (the processed water that, in the case of East Lansing, gets dumped into the Red Cedar River).

About four months after the spill, looking at the plant’s data records about a week after the report of it to the DEQ, the superintendent of the plant, Catherine Garnham, suddenly noticed a spike in mercury had occurred in the “raw influent” data around the time of the spill.

While the readings of mercury levels in raw influent usually hovered around an average of 40-50 ng/L—and had never since May 2009 exceeded 77.3 ng/L—the first reading after the spill suddenly shot to 159 ng/L of mercury. This was on December 10, two-and-a-half weeks after the spill.

Once Garnham noticed this—looking back four months after the spill—she and her boss, Director of Public Works Todd Sneathen, appear in email communications to have been relieved to also see that the standard processing at the wastewater treatment plant seemed to have worked as it should in the face of this spike: it seemed to have cleaned out most of that mercury as part of normal business.

While that influent number had been unusually high after the spill, the first “final effluent” reading after the spill (also on December 10) shows a typically low level of mercury of 1.1 ng/L. In other words, presuming that the first reading after the spill is representative of the period two-and-a-half weeks earlier when the spill occurred, the mercury that got washed down the shop drain does not seem to have resulted in extraordinary contamination of the Red Cedar River, because the wastewater plant did a good job at removing mercury before it dumped water into the Red Cedar. (The contaminated sink itself is another story, one we will return to in a later installment.)

But could some of the mercury have gotten to the river through another route? It is possible, given the proximity of the plant to the Red Cedar River, that some of the contamination left the building on land, through human actions, and then was simply washed down into the river through rain and snowmelt.

When thinking about this, there’s the original source of the mercury to consider: the broken manometer, the device that had spilled its mercury contents. Readers of our last installment may recall that, after the accident, the device had been left in a metal tub that had filled with water outside the building. There it was found four months later by a Health Department investigator.

The cold weather probably kept the manometer frozen in the tub at least most of the time it lay in the tub outside the shop, but we can’t know for sure if the device continued to leak mercury into the tub between November and March, nor whether any mercury might have leaked out from the tub into the environment, including into the Red Cedar.

Then there’s the possibility that some mercury had ended up in outside dumpsters from the cleanup, and had been subsequently washed by rain from the dumpsters into the surrounding environment:

What dumpsters? According to records obtained through the Freedom of Information Act (FOIA), the worker who eventually reported the spill told the Department of Environmental Quality (DEQ) that during the “cleanup,” some of the mercury had ended up in a dumpster or dumpsters outside the shop. How exactly that happened isn’t made clear from the records—were the mercury-laden contents of the shop vacs dumped into the dumpster, or was the duct tape used for mercury cleanup thrown in there?

That we don’t know, but this we do, from internal communications: once upper management knew about the spill and how it was mishandled, they became concerned about dumpsters as a possible source of contamination of humans.

In March, 2014, when he finally learned of what had happened at the plant four months earlier, Todd Sneathen, then Director of Public Works, personally contacted Keith Granger, the CEO of Granger waste management—a contractor who works with the city. Sneathen emailed Granger to let him know that the City would pay for any Granger employees to get medical tests they might need because of possible exposure to mercury from the spill.

Granger sent back the names of three employees who were at risk, including two drivers and a mechanic. (Wastewater plant workers were also sent for testing.)

So is it possible Granger workers had come by to the plant just after the spill and had unknowingly collected some of the mercury as ordinary, non-hazardous waste? The fact that there was worry about Granger employees would suggest that may have happened, or that it was at least a serious concern. (We’ll return to the dumpster trail later in the series.)

What then of those shop vacs that had been used to clean up the spill? Here the story gets even more complicated.

The original report, as the DEQ wrote it down from the worker’s call in March 2014, said that three shop vacs had been used to try to clean up the mercury.

But the Health Department report says that Sneathen told investigators “a clean up was attempted using two vacuums that were in the shop” (emphasis added). Only two? Were there only two, or was Sneathen just not aware there were three?

The Health Department’s report continues, “Both vacuums were canister shop vacuums, both were in the work shop” when the Health Department showed up, four months after the spill.

It’s unclear from this as well as later documentation whether this means that workers had continued to use these vacuums during the interim period. But it appears they probably did.

At that point in its on-site investigation, the Health Department “bagged in plastic bags” the vacuum cleaners. In doing so, they noticed that “One of the vacuums did not have its hose attachment. When asked, it was explained that the hose was currently attached to a third vacuum which was currently in an E. Lansing WWTP [Wastewater Treatment Plant] truck.”

So at least one vacuum hose used in the clean up had moved on to other uses. Was this the hose that had—we learned almost a year after the spill—been used to “blow out” vent coils at the Hannah Community Center?

Again, we don’t know. Here’s what we do know: Once the health department learned of the stray hose, “The vacuum was retrieved and tested.”

Investigators looking in the truck noticed then something more: “loose beads [of mercury] found on driver’s seat of the truck.”

For as many as four months, City workers had apparently been driving beads of mercury around the city.


Read what happened next in  The Mercurial Trail, Part 3: The Call.


UPDATE: This article was corrected on February 17, 2015, at 1:45 pm to note that the time elapsed between the spill and the report of it was four months, not five.