Date: May 28, 1999 To: Heidi King, Office of Management and Budget (OMB) Sandy Zavolta, Environmental Protection Agency(EPA) From: Kevin Bromberg, Office of Advocacy, Small Business Administration Subject: IUR Draft Proposal - Second Comment This is a continuation of the series of comments that the Office of Advocacy plans to provide EPA and OMB about the draft IUR rule. The first comment was provided yesterday (May 27). A. Chemical Universe for IUR Full Reports - HPV/Toxicity Screen Should be Used EPA has not demonstrated that the data from all chemicals in excess of 100,000 pounds have "practical utility" for use in a risk assessment or risk prioritization, leading potentially to risk management decisions by EPA. We are also concerned that the agency may not have adequate resources to analyze this data before it is collected again. Despite significant expenditure of industry resources on UEIP for forty eight chemicals, only four chemicals have even reached the RM1 stage after several years of the program implementation. We are concerned about whether EPA will have the resources to screen many thousands of chemicals during the four years following 2001. In order to prioritize the 2001 data, we suggest that EPA limit the initial chemicals to the HPV set of chemicals that are subject to a toxicity screen. Those HPV chemicals, with known human and ecotoxicity values, that are below certain EPA-defined values, would be exempt from the 2001 data collection. After the 2001 update, EPA may reestablish its priorities, and include new chemicals in the 2005 update. We recommend that EPA use one of the existing toxicity scoring systems, used by TSCA program, or one of the others that were listed in Appendix B to the June 1997 TRI Environmental Indicators Methodology. There is an additional reason why non-HPV chemicals should be excluded from 2001: a lack of toxicity data upon which to perform screening. According to the EPA website, 43% of the HPV chemicals have no toxicity data. Presumably, greater than 43% of the non-HPV chemicals have no toxicity data. Unless the toxicity data becomes available, no valid risk assessment can occur. As I understand the preamble, the petroleum stream chemicals have been exempted, at this time, for very similar reasons - lack of information about toxicity. The IUR data will not have "practical utility", unless the toxicity data will become available in some reasonable time frame. The agency describes the high data turnover rate. The IUR data could become well outdated by the time toxicity data for a given chemical is available. Why not look for exposure data when the toxicity might become available? This also means that the agency needs to coordinate its plans for acquiring exposure data, with its plans for acquiring toxicity data. For the HPV chemicals, that coordination has occurred: there is a plan to complete the missing toxicity testing. Thus, HPV chemicals that are more than minimally toxic can be clearly established as a valid priority for the agency, particularly because the toxicity data are or will also become available in the near term. Furthermore, EPA emphasizes that it needs data from all chemicals above 100,000 pounds in order to select the highest priority chemicals. However, there is no stated basis for EPA's assumption that all these particular chemicals need to be prioritized in one year, let alone by using IUR data. Given the lack of detail in the IUR data, why does EPA believe that modeled data would result in a meaningful prioritization, given the modeling uncertainties? Alternatively, why shouldn't EPA select several hundred chemicals at a time for a more comprehensive survey (like UEIP), based on the information it already has? Is the IUR data and risk modeling really likely to yield such accurate risk estimates? EPA is trading the questionable ability to better prioritize chemicals for the clear improved capability to do better chemical risk assessments with more detailed data. This does not appear to be a good policy choice. B. PAIR Data In the draft preamble at page 29, EPA describes the PAIR database and analysis that will be available in the public record. It indicates that 360 companies submitted 1668 reports on 458 separate chemicals in possible support of Section 4 test rules. I (and presumably Heidi) would like copies of the analysis to review. How well did the PAIR screen work to support the test rules? Did the test rules issue? How do the elements of PAIR compare to IUR? Is PAIR a more effective screening mechanism than IUR? C. Questions on the RM1 Review We are still interested in the answers to the criteria used in the RM1 review, and other written questions we posed earlier this week in preparation for the Wednesday meeting.