How many worst-case release scenarios must be analyzed and documented in the RMP?

As a part of the hazard assessment requirements in 40 CFR Part 68, Subpart B, I must analyze worst-case release scenarios and document certain analyses in my risk management plan (RMP). How many worst-case release scenarios must be analyzed, and how many analyses must be specifically documented in the RMP?

In order to accurately determine which worst-case release scenarios should be specifically documented in the RMP, the owner or operator may need to analyze the worst-case release scenario (i.e., determine the scenario with the greatest distance to an endpoint) for every covered process. If any particular covered process contains more than one regulated substance in excess of a threshold quantity, worst-case release scenarios involving each of those substances may need to be analyzed.

In order for any process to be eligible for Program 1 requirements, the owner or operator must demonstrate that the worst-case release from that process will not affect any public receptor (40 CFR §68.10(b)(2)). A worst-case release scenario analysis must therefore be reported in the RMP for each Program 1 process (40 CFR §68.25(a)(1)). If a Program 1 process contains more than one regulated substance in excess of the appropriate threshold quantity, the documented analysis should be for the release scenario involving the substance that would result in the greatest distance to an endpoint.

For Program 2 and 3 processes, a single worst-case release scenario analysis may be documented to represent the worst of all scenarios (i.e., the scenario resulting in the greatest distance to an endpoint) from covered processes containing regulated toxic substances, and a single worst-case release scenario analysis may be documented to represent the worst of all scenarios involving regulated flammable substances. Additional worst-case release scenarios must, however, be analyzed and reported for Program 2 and 3 processes if worst-case releases from other covered processes potentially affect public receptors different from those affected by the "worst" worst-case release scenarios (40 CFR §68.25(a)(2)(iii)).