Agencies Probe Salty Water in Dunkard Creek Fish Kill

Experts and scientists are trying to determine the cause of the fish kill on Dunkard Creek

The State Journal
15 October 2009
By Pam Kasey

MORGANTOWN -- What is the source of the salty water in Dunkard Creek?

That fundamental question must be answered now that nearly all the fish and mussels are dead on more than 30 miles of the creek and agencies in two states are receiving results for hundreds of lab samples.

Investigators say a toxic bloom of golden algae in September is responsible for the massive kill along the West Virginia-Pennsylvania border. And they say high conductivity — saltiness — supported the toxic bloom.

Conductivity is a measure of water’s ability to pass an electric current. It rises with higher concentrations of the kinds of salts introduced through coal mining and oil and gas production.

It was extra high on Dunkard Creek recently.

In 2005 and 2006, conductivity documented in monthly samples from the creek ranged up to about 13,000 units, according to the West Virginia Department of Environmental Protection, but during the fish kill it was recorded at up to 34,000 units.

Both coal and coalbed methane currently are produced in the Dunkard Creek watershed, and both require the management of salty water.

Could it Be Coal?

Coal companies working in the region have to deal with a maze of interconnected mine voids, some many decades old, that continually fill up with water known as a “mine pool.”

On the parts of the creek affected by the fish kill, CONSOL Energy pumps, treats and discharges water from the pool to mine its coal, notably at the Blacksville No. 2 mine.

The highest conductivity levels DEP found during the fish kill were below the Blacksville No. 2 discharge.

And preliminary results from four samples DEP took within Blacksville No. 2 during the fish kill indicate conductivities ranging from 27,000 to 45,000, according to DEP’s Mike Carico.

That’s not typical for water from this mine pool. Cariso said it’s well above conductivities characteristic of a deep mine in this region, adding it also is above those of similar mines nearby.

Mine pool expert Paul Ziemkiewicz, director of the West Virginia Water Research Institute at West Virginia University, agreed, suggesting that the water is more characteristic of coalbed methane water and natural gas well drilling brine.

Coalbed Methane

CONSOL’s CNX Gas produces coalbed methane in the headwaters of Dunkard Creek.

The company received a U.S. Environmental Protection Agency permit in 2005 to inject the associated water into CONSOL’s depleted Blacksville No. 1 mine, also in the watershed.

Although CNX did not provide numbers on the chemistry of its coalbed methane water, spokesman Tom Hoffman agreed with conductivities ranging from 35,000 to more than 100,000 units.

An item of speculation throughout the fish kill, and one that the Pittsburgh Post-Gazette wrote on Oct. 13 is being confirmed by the Pennsylvania Department of Environmental Protection, is that the Blacksville No. 1 injection well is leaking high-conductivity water into the adjacent and lower Blacksville No. 2 mine.

Hoffman argues against that, pointing to what he characterized as an 800-foot-thick coal barrier between the two mines. However, he conceded that mine maps published by WVU show a segment as narrow as 150 feet.

Hoffman said CONSOL’s own laboratory data on relevant chemistry from within Blacksville No. 2 show no change from before injection began until now.

Carico’s data also may be inconsistent with a breakthrough between the two mines, or at least a sudden, recent breakthrough.

He found high conductivity not only near Blacksville No. 1 but in all four samples from throughout the mine.

Gas Well Drilling Brine

Hoffman said CNX never illegally dumped brine from its Marcellus Shale operations at the Blacksville No. 1 injection site.

That brine goes to municipal treatment plants at Clairton and McKeesport, Pa., he said.

A manager at McKeesport confirmed that CNX has been trucking about 40,000 gallons of water there six days a week “at least since July.” The State Journal was unable to further confirm the disposition of CNX’s brine in time for this story.

Hurting, Helping the Probe

Complicating the matter for investigators is the fact that CNX violated its EPA permit over a period that Hoffman said ended in August 2008.

Violations included failure to provide adequate operator staffing and security at the Blacksville No. 1 injection site, according to the agency’s draft administrative order, and Hoffman said CNX cannot be certain what someone might have put into the well during that period.

But CONSOL and CNX are cooperating and assisting with this investigation, he said.

CONSOL stopped discharging from Blacksville No. 2 on Sept. 18, and CNX stopped injecting coalbed methane water into Blacksville No. 1 on Sept. 28.

In the companies’ defense, Hoffman said the first location where dead fish were reported Sept. 1 is miles below the Blacksville No. 2 treatment discharge.

“There may be other things going on in that creek besides what we’re doing,” he said.

“This is obviously very, very complicated,” he said. “We have taken the view as a company that the important thing right now is to get to the bottom of this and to find out how this occurred — and really only then can we figure out what to do or what we might need to do in terms of our practices.”