Washington state firefighters are urging Democratic senators to halt development of a novel dispersion modeling approach for assessing risks from toxic gas releases, arguing the federal agency and industry collaboration underestimates risks to the public in filings with EPA and other agencies, though an industry group says the method is based on sound science and may still be revised.
In an Aug. 2 letter to Sen. Maria Cantwell (D-WA) and other senators, Washington Fire Chiefs Executive Director Wayne Senter argues that chemical and railroad industries are pushing the new assessment approach -- developed in conjunction with the Departments of Homeland Security (DHS) and Defense (DOD) -- that downplays risks from potential disasters by dramatically reducing the distance toxic gases are projected to travel.
“The industry-initiated but federal agency-enabled and Congressionally-funded effort . . . has all along explicitly aimed at significantly modifying in a risk-minimizing direction the downwind toxic cloud estimates in all of the major national emergency response guidance documents, including . . . chemical facility submissions to the US EPA's Risk Management Program,” Senter writes.
Sens. Tom Carper (D-DE) and Claire McCaskill (D-MO) are copied on the letter, as is Mark Light, executive director of the International Association of Fire Chiefs.
Senter's letter to Cantwell continues months of pushback from the Washington state firefighters association and an environmentalist. This spring, Senter sought to enlist the help of the U.S. Chemical Safety Board (CSB) in opposing the novel assessment method, and an environmentalist unsuccessfully sought a staff opinion on the method.
In his group's April 6 letter to CSB, attached to the Cantwell letter, Senter says facilities have already used the approach to comply with reporting RMP requirements, even though it has not been peer reviewed and is “utterly nonsense.”
Senter contends that the new approach yields dramatically reduced estimates of the reach of a toxic chemical release than past Chlorine Institute (CI) guidance, saying that a toxic plume once estimated to travel 15 miles would span only 1,184 feet or 0.2 miles under the new approach.
“Many fire chiefs find this new industry information utterly lacking in credibility for use in a real release event, and some emergency managers say they will refuse to use it for pre-planning,” Senter told CSB. “We ask you assistance in publicly challenging the current ill-advised industry push for an astonishing risk minimization on paper, by demanding full transparency in the assumptions and models and field test data and a full discussion with the emergency response community and the public and with the federal agencies being directly targeted for re-calculation of their long-standing toxic gas risk guidance; DOT/PHMSA, NOAA, DOT and EPA.”
RMP Rule
The risks posed by potential release of toxic chemicals were the driving force behind the Obama EPA's Jan. 12 final rule overhauling the agency's RMP facility accident prevention rule with new requirements for certain facilities to conduct third-party audits, hazard analysis, and streamline disclosure of toxic chemical holdings.
EPA Administrator Scott Pruitt, while Oklahoma's state attorney general, opposed the proposed rule, faulting disclosure provisions as worsening terror threats.
Now leading EPA, Pruitt has delayed the update rule nearly two years and is weighing significant revisions, despite opposition from environmental, labor and first-responder groups. But if industry groups and federal agencies are able to implement the new toxic release assessment method, it could further ease RMP and other requirements on industry.
A 2014 DHS fact sheet says that it is leading a collaborative project with government, industry and academia that will use large-scale outdoor chlorine release trials to fill data gaps on toxic inhalation hazards from chemical releases. It says the new modeling method seeks to incorporate real-world variables, such as gases' reactions with soil and vegetation, into dispersion estimates. It also says the project seeks to fill information gaps for toxic inhalation hazards that have never been tested or validated at scales representative of releases from rail cars or storage tanks. An industry source says this differs from previous methods that have been based on computer modeling that uses conservative assumptions and have overestimated risks.
But Senter says DHS and DOD have supported testing of the new method, and that some facilities have already used the approach to comply with EPA's RMP rule requirements for submitting an analysis of potential off-site consequences of a facility release. “Federal taxpayers should not continue to fund a complacency-inducing and unreviewed campaign which is dangerous for the emergency responders and the communities at risk for [Toxic Inhalation Hazard] releases,” he says.
Senter says industry has long touted the approach, including in CI's 2015 update to its “Pamphlet 74 Guidance on Estimating the Area Affected by a Chlorine Release.” He says the revised guide backs a risk estimation approach that dramatically reduces the estimated distance toxic chlorine gas released from a rail car accident would travel from 14.8 miles in previous iterations to 0.2 miles.
“Congress should not allow the risk minimization campaign to stay mainly in the shadows, as currently enabled by the secretive US DHS and US DOD agencies keeping locked away from the at-risk public the most important information underlying the risky changes quietly being infiltrated into industry and federal emergency guidance.”
He notes that while early DHS test results using the method suggested that chlorine and other toxic gases may not travel as far downwind as previously believed, recently released DHS data show the releases travel far.
“The new data, unlike results from the previous 2010-2015 smaller scale tests which had been designed mainly to show clouds 'held up' at the release point in various ways, show a long chlorine gas cloud plume, dangerously far downwind. This measurement data decisively undermines the chlorine industry's current risk-minimizing efforts and can help protect the safety-conservative estimates” emergency responders rely on, he says.
He urges the three senators to “ensure that the new sobering field test data is publicized vigorously and directly to the whole Congress, the emergency response community and the public.”
But a CI official says in an email that the 2015 update to the group's guidance brings new science to distance estimates that previously were conducted using conservative and hypothetical computer modeling. The approach stems from a project DHS' Chemical Security Analysis Center began in 2010 and CI joined in 2015. While Senter argues more recent testing undermines the approach, the CI source says that the 2015 update to Pamphlet 74 incorporates test results conducted in 2010, and that it will be updated again to account for more recent testing when those results are published.
“CI members’ primary concern as it relates to dispersion modeling is accuracy. Accurate models assist them in their planning and response efforts in the unlikely event of a release,” the source says, adding that CI members would respond to emergencies alongside local responders. “CI members want models that use the best available science and that is what Pamphlet 74 does.”
EPA Query
In a March 21 letter to James Belke, of EPA's Office of Land and Emergency Management, Fred Millar, an environmental consultant who used to work for Friends of the Earth, queried the agency on the validity of the new gas model and asked whether use of the model is appropriate for complying with RMP reporting requirements.
Citing 1999 EPA guidance on conducting RMP off-site consequence analysis, Millar acknowledges that facilities may use industry approaches to estimate risks from toxic releases, but he argues that EPA guidance suggests that the agency should have a role in assessing the validity of methodologies used to comply with the rule.
Millar asks EPA's Belke a variety of questions, including whether the agency has been provided with models and assumptions underlying the new approach, how widely-accepted the approach is by industry and how commonly it is being used, and whether EPA has standards for assessing whether the model is appropriate for RMP compliance.
Millar tells Inside OSHA Online that EPA has not provided answers to his letter.
In the email to Inside OSHA Online, CI, the industry group, says that DHS selected chlorine to test a new dispersion method because prior modeling tools had been proven to overestimate how far toxic gases would travel in case of a release. While CI members have provided chlorine for testing the experiments were conducted by DHS and its academic partners, the email says.
The source says that CI's primary interest in dispersion modeling is accuracy for emergency planning and suggests that firefighters are misinterpreting the institute's guide. The source says the updated guide presents modeling results in terms of dosage -- a factor of concentration multiplied time.
“To the casual reader, because these results are depicted graphically, this can appear that the footprints represent the predicted size of the cloud,” the source says. “The footprints, as explained in Appendix B, specifically section B.1, are dosages. This measure expresses risk in clear, clinical terms required to assist emergency responders.”
The source also says the CI members and DHS officials briefed EPA staff on the new modeling approach in 2015 before CI updated its new guidance. While EPA officials asked questions during the meeting, they have not followed up to request additional information or to seek further discussions, the CI official says. -- Dave Reynolds(dreynolds@iwpnews.com)