Accident Epidemiology Project - PowerPoint PPT Presentation


PPT – Accident Epidemiology Project PowerPoint presentation | free to download - id: 123587-ZTBiN


The Adobe Flash plugin is needed to view this content

Get the plugin now

View by Category
About This Presentation

Accident Epidemiology Project


First tranche of data covered the period 1994-1999, and became available for analysis in 2000. ... Second tranche of data began arriving in 2004-2005. ... – PowerPoint PPT presentation

Number of Views:100
Avg rating:3.0/5.0
Slides: 15
Provided by: ValuedGate2312


Write a Comment
User Comments (0)
Transcript and Presenter's Notes

Title: Accident Epidemiology Project

Accident Epidemiology Project
  • Paul R. Kleindorfer
  • The Wharton Center for Risk Management and
    Decision Processes
  • The University of Pennsylvania
  • Robert A. Lowe
  • Center for Policy and Research in Emergency
  • Oregon Health Science University
  • Irv Rosenthal
  • The Wharton Center for Risk Management and
    Decision Processes
  • The University of Pennsylvania
  • Center Advisory Committee Meeting 06/16/06

Accident Epidemiology Project
  • Rule 112 (r) of the Clean Air Act Amendments
    requires filing of accident history data for
    15000 facilities in the U.S. The resulting RMP
    and accident history data can be used, with other
    data, to test a number of hypotheses regarding
    the precursors and consequences of accidents in
    the U.S. Chemical Industry.
  • Results can provide valuable insights about
    environmental policy, but also about process
    safety (OSHA) and security issues (DHS). Today
    we focus on the EPA/OEM issues.
  • This research was carried out with the support
    and participation of U.S. EPA/OEM

Overview of Research
  • First tranche of data covered the period
    1994-1999, and became available for analysis in
    2000. This gave rise to a number of studies at
    the Risk Center.
  • Second tranche of data began arriving in
    2004-2005. We are currently analyzing this, both
    on its own and as a comparative study with the
    results of the 1994-1999 data.
  • In addition to statistical work, we have also
    been studying the use of the accident
    epidemiology framework as a means of evaluating
    management systems for predicting and controlling

(No Transcript)
The big picture for 2004-2005 data-I
  • Many facilities appear to have failed to re-file
    as mandated
  • There appear to be significant differences
    between the data reported for the 1995 -1999 and
    the 2000 - 2004 periods regarding two key RMP
  • Hazardousness
  • - Accident rates

The big picture for 2004-2005 data -II
  • By several measures, facilities reporting in the
    2004 RMP data collection round were at a somewhat
    higher level of hazardousness than those the
    1999 round
  • Accident rates reported for 2nd period were lower
    than those reported in the 1st round
  • Drop may be real or reflect change in industry
    interpretation of accident consequences that
    determine RMP reportabality i.e., what
    constitutes a worker injury
  • Substantive data concerns that could affect the
    ultimate study findings are remain under study

Choosing The Population for Tracking the Impact
of The RMP Regulation
  • Unresolved issues remain as to character of the
    100 facilities that apparently failed to meet
    2004 registration requirements and investigation
    of this issue will continue
  • Our study of the cultural, facility and business
    factors influencing the occurrence, character and
    consequences of RMP reportable accidents will
    focus on the facilities that filed in both
    1999-2000 and 2004-2005, a Cohort of 10,446

Did Cohort hazardousness change over time?
  • Hazardousness increased (as measured by
    maximum inventories of RMP regulated substances
    relative to threshold levels)
  • covered by CAA Title V increased
  • covered by EPCRA 302 increased
  • covered by OSHA PSM did not change
  • Trivial change in maximum RMP program level
  • Number of chemicals did not change
  • Need to account for level of Economic

Number of accidents by yearour cohort (N 10,
Number of accidents by periodour cohort (N
Did facilities interpretation of criteria for
reportable RMP accidents change, or did accidents
actually decrease?
  • How would we know if reported accident
    reduction was due to facility reinterpretation of
    threshold reporting criteria?
  • The number of accidents with consequences that
    couldnt be easily hidden would not decrease
    substantially therefore, the of accidents with
    such serious consequences such as public
    evacuations, medical treatment of community
    members, etc. would increase.
  • What if accidents actually decreased?
  • Number of accidents with serious consequences
    would decrease at least as much (proportionately)
    as total number of accidents therefore, the of
    accidents with serious consequences would remain
    the same or decrease

Other questions being studied
  • How do the consequences of reported accidents
    compare across the two five-year periods? (what
    we discussed above was just related to frequency
    of reported accidents)
  • What have been the trends in worker and public
    responder injuries? In property losses?
  • Have there been fundamental changes in accident
    rates or consequences for major sectors of the
    chemical industry when comparing 1994-1999 to

A BIG question for EPA/OEM/OMBHas RMP
Regulation been Cost/effective?
  • Cost of implementation?
  • Value of Reduction in accidents/injuries?
  • Value of Reduction in worst case footprints?
  • Improvements in communication and response
    patterns for emergency responders?
  • Improvements in the interaction of EPA and
    facility managers with communities and NGOs?
  • Improvements in Management Systems for major
    accident prevention?

Some Questions for the Day
  • How can this on-going research be best
    communicated to industry and other stakeholders
    interested in process safety and environmental
  • There is a question about the role of reporting
    behavior as one of the reasons for the apparent
    downward trends in the RMP data.
  • What is your general perspective on the
    incentives on companies regarding external
    reporting, using RMP as an example?
  • E.g., Have the incentives for underreporting,
    noted in earlier research, been off set by
    pressures for greater transparency and better
    governance since Sarbanes-Oxley?
  • The RMP data suggest that hazardousness has not
    decreased. Assuming that this is the case, and
    given the pressures to improve this (both for
    security reasons, as well as for chemical
    safety), why hasnt this occurred. In
    particular, what are the strategic costs and
    benefits of reducing the potential for reasonable
    worst-cast consequences?