MEASURING MERCURY: EPA pollution data worth a look, flaws and all

Investigative Reporters and Editors, Inc. The IRE Journal, Jan/Feb 2008 by Overberg, Paul

Last summer, USA Today and Gannett News Service joined forces to analyze the Toxics Release Inventory data from the U.S. Environmental Protection Agency.

The TRI database details how many pounds of more than 600 chemicals are released into the air, water, and various waste and recycling streams each year from tens of thousands of facilities. Those include manufacturers, mines, oil and chemical terminals and even federal facilities. Small firms and those that only use small amounts of listed chemicals are exempt.

We approached this as a test for collaboration in data collection, analysis and publication. The project culminated in a three-part series called "Toxic Legacy," which included a close look at mercury pollution, along with other known and emerging environmental threats. Our coverage included a Google Maps mashup of five years of TRI data for 2,400 facilities and a state-by-state digest of fish consumption advisories and links to each state's fish advisory program Web site.

An over-under bet

This was the second time USA Today had collaborated on a TRI investigation. In 1992, when the EPA released the first year of TRI data, it arrived on 9-track computer tapes the size of dinner plates. We shipped custom reports to Gannett papers on greenbar computer paper. (Current and archival TRI data is available from the IRE and NICAR Database Library, www.ire.org/datalibrary.)

News technology and the TRI itself have changed dramatically, but some things haven't. There are still challenges in using TRI data, but it remains a priceless and underused tool.

The TRI was launched through the Emergency Planning and Community Right-to-Know Act of 1986 (EPCRA) and expanded by the Pollution Prevention Act of 1990. The data represents self-reported estimates from private industry and government-run facilities.

There's definitely underreporting-and overreporting. The TRI is like an election where the polls never close. Facilities can revise their data years later, and sometimes are forced to do so after inspections. Extensive information about the reporting rules and industry guidelines are online at www.epa.gov/tri.

Environmental groups point out that the TRI understates emissions because smaller facilities are exempt, and the EPA allows some facilities to simply estimate emissions. The EPA has exempted some facilities from filing a detailed report if they don't use or release much of a chemical. (California quickly re-imposed the old requirement. Twelve states are suing the EPA to overturn the new rule.)

And to the dismay of many journalists, the TRI can't say anything about human exposure levels, or what culprit released the mercury found, for example, in a given lake. Similarly, there's debate about just how significant U.S. industrial emissions of mercury are in the global scheme of things. Other sources-volcanoes and industrial emissions from Asia, Mexico and Canada-also affect U.S. mercury levels.

The mercury example

We chose mercury in part because the TRI made the biggest source of mercury emissions easy to analyze. Other reasons had more to do with storytelling:

Mercury's toxicity is well-known and well-documented In humans, mercury acts as a nerve toxin and affects fetal development.

It has many industrial uses, and traces remain in enough household products that it's familiar to many consumers.

Industrial sources are widely scattered, although many of the biggest lie in the Ohio River Valley. And it's mostly released into the air, where it is then widely dispersed.

Regulation varies. The EPA advises consumers to recycle fluorescent light bulbs to prevent traces of mercury from escaping. And if you spill more than two tablespoons, you're required to call the EPA's National Response Center immediately. But in the TRI, big mercury sources don't even have to measure emissions directly. Some have found that their estimates are hundreds of pounds shy of what they actually pour into the air each year.

Finally, everyone's hands are dirty. Americans use more and more electricity every year. Coal-fired power plants exist in most states and their output is directed throughout the grid.

In the environment, airborne mercury poses the biggest problem once it falls to the ground with rain and reaches bottom sediments. There, bacteria convert it so it can be taken up in the food chain, accumulating along the way. Almost every state has issued advisories to limit consumption of certain fish from certain bodies of water because of mercury contamination.

Our series showed how overall airborne emissions of mercury have dropped since 2000 but those from utilities have actually risen slightly. This is due to the growth in electricity generated from coal and the slow pace of additional controls. After years of planning, an EPA cap-and-trade plan to cut mercury emissions has begun, but it's under attack in federal court as doing too little too slowly.

We compared airborne mercury data reported for 2001 when it was first estimated and as it's now reported on the EPA's online system (www.epa.gov/tri). The difference: Among 839 facilities, 20 reported underestimating by a collective 3,474 pounds of emissions; 21 reported overestimating by a collective 1,274 pounds. The difference made up an extra ton of mercury, or 1.6% of the national total.