Wednesday, January 21, 2009

An agenda for the Toxic Release Inventory

Today Obama signed a new Executive Order revitalizing FOIA, and spoke about transparency being one of the touchstones of his Presidency. As a sort of minor reciprocal gesture of faith, I've decided to put up an agenda of changes that I think should be made to the Toxic Release Inventory (TRI), one of U.S. EPA's most successful data programs. At any time during the last eight years, I wouldn't have talked about these ideas in any public forum until each piece was ready to be proposed formally, knowing that talking about the agenda beforehand would only give industry a heads-up to prepare to scuttle it. But perhaps being willing to talk about it beforehand will help everything go more smoothly. Needless to say, I'm also eager to hear about other people's ideas.

I have no formal authority to propose these kinds of changes at all, and am not writing this under any organizational affiliation. They are simply my personal ideas. However, I've worked extensively on the Toxic Release Inventory outside EPA since 1990, and I consider myself to be an expert on it. There are a number of organizations which I plan to work with to propose these things formally, in the event that no one else does.

I'll try to provide a little bit of background, in the unlikely event that anyone reads this who doesn't know all about TRI already. If you want to find out basic information about the Toxic Release Inventory, you can go to EPA's Web page on it, but in short it's a database, legislatively mandated to be publicly accessible in electronic form, that requires most kinds of large, fixed polluters to report their releases and transfers of toxic chemicals, as well as the toxic chemicals in waste that they generate. Unlike EPA's hazardous waste databases, it also requires that these reports be of the amount of chemical, not the amount of chemical plus inert filler, and that the reports be in common units. Facilities have to report releases of chemicals to air, land, water, and underground. These characteristics make the database EPA's most useful one for general toxic pollution issues. TRI was the model for many similar databases in other countries; these databases are generally called PRTRs, Pollutant Release and Transfer Registers.

The Toxic Release Inventory has had notable expansions during its now decades-long span -- waste data was added by the Pollution Prevention Act in 1991, Federal facilities added in 1994, hundreds of chemicals added in 1995, new industries outside manufacturing had to report starting in 1998, and data on persistent bioaccumulative chemicals was expanded in 2000. And for the last eight years, that's where it has stood -- other than a disgraceful reduction in TRI data in 2006, when the "Form A" option for facilities to avoid reporting was expanded, despite widespread opposition from the public, lawmakers, and scientists. Suffice it to say that although EPA's civil service employees generally understand the value of TRI, the administration running the agency at that time was actively hostile.

So these ideas have been around, in some cases, for many years, but there was no point in proposing them until the administration changed. Now that it has, here they are:

1. Add chemicals that cause global anthropogenic climate change (global warming)

This would be primarily CO2, methane, and nitrous oxide -- the Montreal Protocol chemicals, which also contribute to global warming, are already in TRI. I'm not a lawyer, but I would think that this would be possible under current legal authority because chemicals can be added to TRI if they cause a sufficient adverse effect on the environment, which GHGs are now known to do. The statute does refer to the adverse effect as being due to toxicity, so this interpretation is not certain. But if new authority is needed, it's going to be gotten anyways -- the Obama administration plans to do something about global warming, and you can hardly do anything about it without tracking emissions.

I can understand that people who want to track GHGs may want a new database designed for that particular purpose. (Currently EPA and DOE have a patchwork of databases, none of which were really designed for that, and which are going to have to be replaced.) TRI has certain limitations as a database tracking GHGs; what, for example, would be done about transportation sources, like fleets of trucks? But it seems to me that these chemicals should be added to TRI in any case, even if another database is created to specifically track them, as a cross-check. The information would also then benefit from TRI's well-developed distribution paths and user base. And TRI would come closer to matching other countries' PRTRs, several of which added GHGs quite a few years ago.

2. Change range reporting and speciation reporting of highly toxic chemicals

Since the beginning of TRI, people have faced the problem that TRI most easily allows comparison of raw pounds of chemicals -- but some chemicals are much more toxic than others, and some releases affect more people than others. These problems are attacked by a (in my opinion) excellent program within EPA called RSEI. But whenever there is a data release using data from RSEI -- most recently, the USA Today report -- the same thing happens, a facility reports a release range of 11-499 pounds of a highly toxic chemical like chromium compounds or diisocyanates, and this rightly is converted by EPA into the midpoint of the range, 250 pounds. (I blogged about a case of this here). 250 pounds is generally a very large release, for these chemicals, and it makes facilities pop up as major polluters and wastes everyone's time as people have to track down what is really going on. Most often, it turns out that the facility either thinks that its releases are towards the bottom of the range, or they don't release the more dangerous form of a chemical with two or more forms -- hexavalent vs trivalent chromium, for instance.

EPA could help both the reporting facilities and the public by changing the reporting rules. First, for those highly toxic chemicals that have more than one common form, EPA could require that each form be reported separately. This is already done for Dioxin. Second, EPA could prevent release and transfer range reporting from being used for these chemicals, or at least warn people who report through the TRI-ME reporting software what they are about to do.

3. Fix the Form A.

I won't write about attempts to roll back Form A reporting to what it used to be; that's already being done by various people. But whether the expanded Form A is rolled back or not, EPA should change the way in which it treats already-reported Form A data. A Form A should be treated as a range report, in the same way as TRI release and transfer ranges are treated. If someone sends in a current Form A for a non-PBT chemical, they are reporting that they release not more than 2000 pounds of the chemical, and generate in waste not more than 5000 pounds. This is not zero pounds. In fact, it's a range: 0-2000 pounds for releases, and EPA should use its well-established procedure for handling release ranges and take the midpoint, converting this to a quantity within the database of 1000 pounds. This would preserve the TRI reporter's ability to not have to take the time to fill out the additional information required for a Form R, but it would also give the public a more accurate estimation, based on best available information, of what is going on. If a TRI reporter didn't want the public to assume that the best guess was 1000 pounds, they always have the option of choosing to fill out a Form R and reporting any number of pounds, including zero.

This procedure is currently used for TRI data provided by RTK NET (described here), so I know that it's both possible and easy to do. The release media (air, land, or water) can't be determined, but a general Form A pseudo-release-medium can be assigned. No change in how facilities report would be required in order to do this; it could be done simply as an internal change in how EPA handles the data.

4. Fix Pollution Prevention Act reporting.

Pollution Prevention Act data were added to TRI in 1991, and they've never been able to be done exactly right. The waste data are currently reported as amounts recycled on and off site, burned for energy recovery on and off site, treated (destroyed) on and off site, and released or disposed of on and off site in various ways. But all of these numbers were supposed to add up to a single number, the quantity of the chemical in waste generated by the facility. That's because reporting these numbers was supposed to encourage source reduction, the practice of changing processes to reduce the amount of waste generated. But the regulation was sabotaged by ideologues at the Office of Management and Budget, and EPA was prevented from defining what a waste was, or something similar. So it could only require that facilities report the components that add up to the overall waste number, without really referring to it in a coherent way.

I don't know whether any actual change in reporting is required for this to happen -- perhaps just better guidance from EPA? Re-opening the issue with a new OMB? Perhaps having facilities total up their Section 8 waste quantities and report the total would help them realize what it's supposed to be for.

5. Re-open cooperation with other countries' PRTRs

I'm not sure where this stands, organizationally, and it's not the kind of thing I'd know about directly. (Although I have worked for the Commission for Environmental Cooperation on putting together a Web site to display combined U.S., Canadian, and Mexican PRTR data.) But it seems to me that EPA could have more of a sense of where chemicals within TRI stand within an international comparison. Just about the entire First World has some sort of PRTR -- does EPA track transfers from one country to another using multiple PRTRs? Could they, perhaps, report on how U.S. emissions of particular chemicals compare to those of other countries? I'd guess that more is being done on this score than I know about, but anything that EPA can do to make this kind of data more truly global would be highly valuable.

(I should probably note that I once tried to put together a proposal to make a free, publicly accessible database-backed Web site that would allow people to search all the PRTRs at once. I couldn't find a funder for it -- most charitable foundations and NGOs that I deal with focus on the U.S. That was a few years back, and I don't know whether anyone is doing that now, but it would still be a good idea.)

That's probably enough for now. I invite anyone who wants to comment on this to feel free to.


  1. Nice list.

    In addition to the GHG's, maybe EPA should also include the full set of the criteria pollutants in TRI. CO, SO2, NOx, Ozone, Particulate Matter are not now included. Would it be feasible to add all of them?

    Maybe also include mobile corporate sources (airplanes, ships, trucks, rail, car fleets, buses). These would be especially important if the chemical list were expanded to include the criteria pollutants.

  2. Paul Orum tried to post the comment below but it was rejected somehow, which is really too bad since I thought I had set this blog to accept any comments, even anonymous ones. But he Emailed it to me and so here it is:

    "I think you are correct about Section 8, Pollution Prevention Reporting, and the solution in my opinion is to total the different waste streams on the form. That will expose everyone to the material waste incentive of that firm to reduce pollution at the source - before it becomes pollution.

    I would add: organize facility identification reporting across all environmental laws, such that facilities identify themselves, completely, only once. That will provide reporting efficiencies both within TRI and across all environmental laws.

    And, take a hard look at section 7, waste treatment methods, to see if it can be made simpler."

  3. Thanks for the comments, Michael and Paul. I've invited a few other people to comment, and couple more will probably come in eventually -- Catherine Miller mentioned that the criteria pollutants have indirect effects on global warming, so perhaps they should be included with the GHGs if for no other reason. My sense is that the reason they aren't in TRI is mostly historical; data on the criteria pollutants was one of the only kinds of regular quantity-based facility-specific national pollution reporting that I know of that existed before TRI did. Paul, there's an ongoing effect, going years back, to organize facility identification reporting under the Facility Registry System, FRS.

    I'm not really optimistic about EPA's ability to do this. They have to get into the data collections and get the data right when it's collected, as opposed to correcting the information later once the communication with facilities has already taken place and the data are already in various systems. The collect-and-correct approach goes back to FINDS, and there we're talking about, maybe, something that goes decades back. When EPA tried to set FRS up, as I remember, it got into an awful muddle between the states, EPA, OMB, and EPA's varying statutory authorities to collect data. So it's not something that I would be particularly interested in spending energy on.

    However, I should note that EPA has hived off lat/long and RCRA ID data collection within TRI to FRS. I don't know whether anyone has looked at how successful this has been in getting actual lat/longs and RCRA IDs for all the TRI facilities. Something to look into.

  4. I've heard back from someone else on this by phone, and I should emphasize that adding the GHGs to TRI should not pre-empt the creation of a special registry database for GHGs. A specialized database for GHGs might require all sorts of things that TRI doesn't handle: verification of release quantities, quarterly reporting, reporting of chemicals in products (i.e. gasoline) instead of in wastes, etc. It's just that TRI might be a valuable check on and / or additional public distribution path for these data. If a facility is measuring these numbers for a GHG registry, it shouldn't take much effort to them copy them into a TRI submission. And it's conceivable that reports on GHGs might appear within TRI that don't appear in the GHG registry, which would indicate that the criteria for reporting to the GHG registry might need to be expanded.