A Red Flag On Disclosure Of Hydraulic Fracturing Chemicals

It’s not often that a new regulatory idea becomes so popular that one or more states per month climb on the bandwagon. But that is precisely what has happened with the push to disclose which chemicals are pumped into the ground to stimulate oil and natural gas production during the process known as hydraulic fracturing, or “fracking.”

A year ago, only three states (Arkansas, Montana and Wyoming) required oil and gas producers to tell the public what chemicals they were using. Two other states (Colorado and Texas) were actively developing such rules. Today, just twelve months later, statutes or regulations mandating “frack” chemical disclosure are on the books in no fewer than 18 states, and proposals are pending or under consideration in several others.

FracFocus, an online registry that compiles information on hydraulic fracturing chemicals both for states where disclosure is voluntary and required, has been up and running for just 20 months, but already it houses approximately 800,000 records that include ingredients data. As of December 5, 2012, this data represented 33,606 wells. The amount of information on the site continues to grow rapidly.

It is impressive that so much information has been made available in such a short time. Still, people have begun to wonder whether the disclosure rules are accomplishing what was intended. The question is important because rules that aren’t working need to be changed. A good regulatory system is based on a process of continual improvement, not a naive idea that the rulebook can be written in a way that will never need changing.

Unfortunately, judging from early press reports, there are quite a few bugs in the system. To be fair, the reporting requirements are quite new and still being implemented — and analysis of the data has barely begun. But  problems are emerging. The issue receiving the most media attention is the sheer number of trade secret claims.

Following precedents established by federal occupational safety and “community right to know” laws, all of the state disclosure laws contain exemptions for trade secrets. A key question (if one accepts the idea that some sort of special handling of trade secrets information is inevitable) is whether the bar for trade secret protection is set high enough, and if so, whether trade secret claims are being adequately policed. It may be that the answer to both questions is no. Certainly a lot of trade secret claims are being made.

Bloomberg News reported last week that there have been about 19,000 trade secret “or similar” claims in Texas this year and that, nationwide, companies have withheld one out of every five chemicals used (Bloomberg apparently means one in five of the number of chemicals used per well rather than one in five of the total number of discrete chemicals used by the industry).

And Energy Wire reported in September that on a well-by-well basis at least one chemical was kept secret in 65 per cent of disclosures. Although these statistics don’t by themselves prove that the trade secret provisions in disclosure rules are being abused, the numbers are a red flag.

After reading the Bloomberg article, EDF staff took a close look at data submitted to FracFocus for just one state (Texas) and one month (July of this year). Here’s what we found:

1. The data is “dirty.” A total of 29 percent of reported Chemical Abstract Service (CAS) numbers – unique identifiers needed to look up what is known about a chemical in scientific literature – were wrong. Meaning that the identifiers reported do not exist in the official CAS registry.
2. By coincidence, the total percentage of discrete chemicals that companies claimed as trade secrets was also 29 per cent. The percentage claimed by one company was right in line with what they predicted prior to enactment of the rules – 15 percent. But another company seems to have claimed 26 percent of its chemicals as trade secrets, although it told EDF before the rules were enacted that fewer than 5 percent of chemicals in its catalogue would need trade secret protection. Like I said before, this doesn’t prove or disprove company compliance, but it does suggest that hard questions need to be asked.
3. Based on a very small subset of the July records (only ten out of 868 wells with reports filed that month), the rate of trade secret claims appears to be much higher for chemicals that are not subject to Occupational Safety and Health Administration MSDS (Material Safety Data Sheet) disclosure requirements than for chemicals that are.
4. Based on that same small sample of ten wells, the rate of trade secret claims traceable to “third party” suppliers who may be too far up the supply chain to be governed by the current rule was dramatically higher than the rate of trade secret claims by those reporting on their “own” chemicals. In addition to pointing toward potential rule amendments at the state level, this underscores why it is important that EPA act under TSCA, the federal Toxic Substances Control Act, to require manufacturers and other entities at the beginning of the supply chain to report what they know about the use and health effects of particular hydraulic fracturing chemicals. EDF, together with EarthJustice and other groups, has petitioned EPA to take such action.

Some shortcomings are already being addressed by the Ground Water Protection Council and the Interstate Oil and Gas Compact Commission, which oversee FracFocus, the online registry. For example, FracFocus is transitioning from a collection of files to a true database with search features and automated “pop-ups” that will warn if data being entered appears to be wrong. Information in the system is also being made more readily accessible to enforcement agencies.

In its current form, the FracFocus chemical registry does not allow data to be aggregated and analyzed other than for single wells. Recent press stories, and reviews such as EDF’s look at the data for one state over one month, are only possible at this point thanks to SkyTruth and Pivot Upstream Group, which have expended great effort to “scrape” the FracFocus data. As John Amos of Skytruth said last month when releasing the raw information, the objective was to capture the dataset “as it is – warts and all – in hopes that other enterprising researchers and analysts would be able to help clean it up and produce useful analyses to inform the public. …Shortcomings can now be systematically analyzed and quantified by anyone with the skills and interest to do so.

Although it would be better to work directly with the data filed by industry than with a “scraped” version (to avoid questions about errors that might have occurred during the scraping), Amos is right. It is time to begin identifying the shortcomings.