This weekend I spoke with Dr. Avner Vengosh, one of the researchers from Duke University that published results of a study looking at wastewater quality from "fracking" operations in Pennsylvania. Their study, "Impacts of Shale Gas Wastewater Disposal on Water Quality in Western Pennsylvania", was published this month in the journal Environmental Science & Technology.

The study was widely covered in the news and on various blogs, but there are some nuances that deserve a little more attention. The summary from Bloomberg:

Naturally occurring radiation brought to the surface by gas drillers has been detected in a Pennsylvania creek that flows into the Allegheny River, illustrating the risks of wastewater disposal from the boom in hydraulic fracturing.

Sediment in Blacklick Creek contained radium in concentrations 200 times above normal, or background levels, according to the study, published today in the journal Environmental Science and Technology. The radium, along with salts such as bromide, came from the Josephine Brine Treatment Facility about 45 miles (72 kilometers) east of Pittsburgh, a plant that treats wastewater from oil and gas drilling.

After reading the paper, I came away with three big takeaways:

  1. The research team found evidence of flow back and produced water from Marcellus shale gas operations based on chemicals and isotopic ratios associated with the Marcellus shale. These are the chemical and isotopic 'fingerprints' that allowed the scientists to identify Marcellus oil and gas wastewater in the mix of normal wastewater at the Josephine Brine Treatment Facility. The team found elevated levels of chloride and bromide, along with the strontium, oxygen, radium, and hydrogen isotopic compositions. It should be noted that radium (a radioactive element) is found naturally in the Marcellus shale brine and is technically called a NORM: naturally occurring radioactive material.
  2. Wastewater treatment plants are effective at removing 90% of barium and radium such that the effluent is well below the industrial waste discharge limits. The water exiting the wastewater treatment plant (effluent) only has 10% of the radioactive material left in it, which is within allowable limits but still poses a problem (see next point). The discharge limit for radioactive material is 2.2 Bq/l - well above what the team found for 226Ra (0.11 Bq/l-0.19 Bq/l) and for 228Ra (0.04 Bq/l-0.13 Bq/l) - note the volume basis. Also concerning is the highly saline effluent with elevated toxic metals like barium and strontium.
  3. Accumulation of radium in sediment exceeds U.S. regulations as does the solid/sludge precipitate and requires special disposal (cannot dump on soil or in a standard municipal landfill). During the treatment process, chemicals are added to the water to bond with chemicals, which precipitate out. The resulting solid/sludge is then sent for disposal. The Duke team notes that at 900 Bq/kg (note the mass basis), the sludge would need special waste processing. As for the sediment, even though the volumetric amount of radium leaving in the wastewater effluent is within current regulations, it does not remain in the liquid phase and is adsorbed in the river sediment near the discharge site. It then builds up and can be consumed by bottom feeders (and eventually other wildlife) or accumulate in freshwater plants. Sediment samples were recorded at 2,072 Bq/kg (228Ra) and 8,732 Bq/kg (226Ra) - up to 200 times the background levels measured upstream of the treatment facility.

The paper recommends advanced treatment technologies to "prevent discharge of contaminants (Ra and Br) to the environment in areas of shale gas development and hydraulic fracturing". Dr. Vengosh elaborated (via email):

Our data show that halogens (chloride and bromide are not removed at all) and Ra was removed but not entirely. Removal of halogens would require different treatment technology, such as desalination. The technologies to treat such high level of salinity and radioactivity complex water do exist, the questions are the cost, implementation, and monitoring.

I asked Dr. Vengosh if industrial discharge limits should be revised to anticipate mixing and sedimentation. Here is his answer:

Yes! In our previous study on the impact of effluents from coal ash ponds in NC (Ruhl et al., 2012, see attached) we showed that arsenic is attached to particulate matter and redissolved under reducing conditions at the lake bottom sediments. Thus in spite of the relatively low content of As [arsenic] and Ra [radium] in the effluents their accumulation in river or lake sediments could cause a long-term environmental hazard.

A follow up question from me: "is this unique to this type of stream? Or put another way, with different stream and flow conditions, would the amount of accumulated Ra to be lower and within regulations?". His reply:

I do not think so, we have results from a pond (not published yet) with similar Ra accumulation results. So radium (and other toxic metals like arsenic in oxic conditions) would tend to be attached to any suspended matter or sediments. The rate of the stream flow might cause a larger and perhaps more diluted zone of Ra accumulation, but the same process is expected.

The major impacts of radium accumulation appear to be localized to less than 200 meters downstream from the treatment facility, and not traveling in significant concentrations downstream and into other watersheds. However, significant amounts of shale-related water is sent through centralized waste treatment facilities and discharged to local streams.

These results indicate how quickly the industry has moved relative to the regulatory and scientific communities and the need for stricter regulations based on scientific data. Increased water reuse (or waterless fracking) would address this issue (previous research estimates that 70% of flowback and produced fluids are currently reused in the Marcellus Shale region) as would revised effluent limits for salinity, toxic metals, and naturally occurring radioactive materials.