OMB Approves Emergency ICR Renewal for CSB Incident Reporting

On Monday, the OMB’s Office of Information and Regulatory Affairs (OIRA) announced that it had approved an emergency renewal of an information collection request from the Chemical Safety Board for “Accidental Release Reporting”. The emergency approval was needed because the ‘current’ approval expired on April 30^th, 2023 and CSB did not publish the 30-day ICR notice until May 1^st. This 30-day emergency approval allows CSB to continue to legally collect the incident reporting information. They have already met OIRA’s submission deadline for ‘regular’ renewal of the ICR.

The CSB has a history of problems (see the ‘Commentary’ section) with this ICR process, but since this is their only ICR, they do not have a lot of institutional experience with the process. Hopefully they will do better the next time this ICR is up for renewal.

CSB Guidance on Reporting Accidental Chemical Releases

Yesterday, the Chemical Safety and Hazard Investigation Board (CSB) published their promised guidance document on the reporting of accidental chemical releases under 40 CFR 1604. As with any governmental ‘guidance’ document there is a lot of superfluous verbiage setting the background of the need of, and legal justification for, the base regulation. CSB does a better job than most with making this readable, but the first seven pages of the document can be skipped by all but the most ardent infophiles.

Pages 8 and 9 of the document provide a nicely done summary of what CSB expects when an accidental chemical release occurs. An interesting pull quote from page 9 reflects an unusual amount of bureaucratic naivete about corporate interests in reporting chemical incidents:

“Under the CSB’s Accidental Release Reporting Rule, it is always safer for an owner/operator to report, rather than fail to report. Thus, it is the CSB’s position that if an owner/operator is unsure whether the incident should be reported, the owner/operator should report, rather than risk violating the rule by failing to report. There is no sanction or enforcement action associated with reporting an accidental release, which in retrospect, did not have to be reported. The opposite, however, is not true. Failure to report an accidental release when required by this rule could lead to an enforcement action brought by the EPA.”

The remainder of the document is formatted in a frequently asked questions (FAQ) format. CSB does go beyond just quoting relevant parts of the regulation. For example, at the end of the reply to FAQ 2.11 on pages 10 and 11, CSB concludes the discussion by reminding folks that the intent of the rule is to provide CSB with the necessary information to determine if an investigation is warranted and then states:

“The CSB has and will continue to investigate matters involving the accidental releases of chemicals, petrochemicals, and hydrocarbons of all types, provided that a fatality, serious injury, or substantial property damage is caused by the accidental release at issue.”

Any owner or facility manager for a facility that produces, uses, or handles hazardous chemicals of any sort, ought to have a copy of this document handy, and look at it periodically. Every EH&S manager needs to have read and understood this guidance before a chemical release incident takes place, probably before close of business yesterday.

Gas Leak or Glass Leak? Is it a Reportable Release ?

An interesting set of articles about a weekend fire at a glass manufacturing facility in Wichita, TX. Initial reports (here, here, and here) explained that a ‘gas leak’ was the cause of a fire and explosion at the facility that lead to nine employees being sent to the hospital. The local Sheriff was quoted as providing the information about gas leak near a burner leading to the explosion.

A statement from the company is quoted here as saying that the fire was the result of “a leak allowing hot, molten glass to escape from the vessel.” That quote goes on to explain that the evacuated employees were being treated for heat stress. It would be easy to miss-hear “glass leak” as a “gas leak” in a verbal report about the incident.

This incident does raise an interesting point related to the CSB’s chemical release reporting rule. If the fire had been due to a natural gas leak, would it have triggered a reporting mandate under 40 CFR 1604.3? That section requires reporting for “any accidental release resulting in a fatality, serious injury, or substantial property damage.”

A couple of key definitions from §1604.2 bear on the answer to that question:

• Accidental release means an unanticipated emission of a regulated substance or other extremely hazardous substance into the ambient air from a stationary source.

• Extremely hazardous substance means any substance which may cause death, serious injury, or substantial property damage, including but not limited to, any ‘‘regulated substance’’ at or below any threshold quantity set by the Environmental Protection Agency (EPA) Administrator under 42 USC 7412(r)(5).

• Serious injury means any injury or illness that results in death or inpatient hospitalization.

Natural gas (presumably the fuel source for glass smelting process) would certainly classify as a ‘extremely hazardous substance’, so we would have had the requisite ‘accidental release’ if any of the employees taken to the hospital for treatment were admitted for treatment or observation.

Interestingly, an argument could be made that molten glass could be considered an ‘extremely hazardous substance’ upon release. And while no news reports claim that any of the evacuated employees were injured as a result of contact with the leaked molten glass, it would seem that the injuries could be considered to be as a result of that release. So, maybe this was a reportable release in any case.

Process Upsets and Chemical Release Reporting

I ran into a brief article on ChemicalWorld.com that caused me to thing about process upsets and the CSB’s chemical release reporting requirements. According to the article and employee was exposed to sulfur dioxide at a North Alabama chemical plant early last month. Two days after his exposure he was admitted to the hospital for breathing problems and then died from those problems last week, better than a month after the incident.

I did a quick check of the local news outlets and could find no reports about the incident. There were two articles about the employee’s death (here and here). Both articles were brief and contained second-hand accounts of the incident from the employees wife. The wife is quoted as describing the incident this way:

“There were 3 of them out there, outside, working on the towers. There are two towers that the chemicals flow through. Supposedly the excess chemicals flow into what they called the pit. When they went out that night, the smell from the pit was different than what it had been in the past.”

Rather than a classic ‘chemical release’ this sounds like some sort of chemical process upset that released unusual fumes. If the fumes were sulfur dioxide, then it would not have taken a large quantity of those fumes to cause damage to the employees lungs if he were close to the source of the fumes.

Now, according to the definitions in 40 CFR 1604.2, such fumes of sulfur dioxide would be an ‘accidental release’; “an unanticipated emission of a regulated substance or other extremely hazardous substance into the ambient air from a stationary source.” On the date of the release, it would not have been considered a reportable release under §1604.3 because it was not an “accidental release resulting in a fatality, serious injury, or substantial property damage.” Once the employee was admitted to the hospital for injuries caused by those fumes, the accidental release became a reportable release.

Interestingly, in this case, if the company had reported the release when the employee when the employee was admitted to the hospital 24 to 36 hours after the incident (the timeline is not clear in the news reports) they would already be significantly late in meeting the 8-hour reporting requirement in §1604.3. The CSB’s discussion of that reporting deadline in the preamble to the final rule last year, makes it clear that they are concerned with the clock starting at the time of the incident, not the time when the owner becomes aware that an incident has occurred.

This probably means that chemical facilities need to seriously consider submitting accidental release reports when process upsets cause even minor releases of hazardous chemicals when personnel may have been exposed to such fumes.

OMB and CSB’s New Reporting Rule – 4-9-20

Yesterday the OMB’s Office of Information and Regulatory Affairs (OIRA) published a notice that it had assigned an OMB Control Number (3301-0001) for an information collection request (ICR) submitted by the Chemical Safety and Hazard Investigation Board (CSB) in support of the CSB’s rule on Accidental Release Reporting. OIRA also reported that they had filed comments on the ICR on the rulemaking docket (not yet posted as of this writing; it could take a day or two).

The ‘Terms of Clearance’ in yesterday’s notice specifically states that: “This OMB action is not an approval to conduct or sponsor an information collection under the Paperwork Reduction Act of 1995.” Thus while the CSB’s final rule has been published, OIRA has not yet approved the CSB’s collection of reports under that rule.

Commentary

There are a couple of oddities with yesterday’s notice. First, OIRA reports that the ICR paperwork was submitted to them on April 2^nd, 2020. All of the data submitted was based upon the publication of the December 2019 notice of proposed rulemaking, not the final rule. Now CSB was required to submit that information as part of their NPRM processing. In the normal course of events the OIRA comments on that request would have been reviewed and acted upon as necessary by CSB as part of the preparation of the final rule. Unfortunately, the abbreviated time frame between the NPRM and the publication of the final tule would probably not have allowed for normal OIRA review and reply, but apparently the CSB never gave OIRA that chance.

I do not think that this was deliberate malfeasance on the part of CSB. This was, after all, the first ICR submission made by the Board. I suspect that CSB attempted to submit the initial ICR when they published their NPRM, but somehow failed to properly submit the documents and never made an actual connection at OIRA.

Apparently, the CSB recently realized that they had not followed proper procedures in the initial ICR submission process (probably when someone pointed out the lack of an OMB approved control number) and worked with OIRA to correct that issue. Unfortunately, this probably means that CSB will have to publish a 30-day ICR notice to complete the OIRA process.

In the meantime, until OIRA approves the ICR, the CSB cannot legally require (regardless of the publication of the final rule and the effective date – March 23^rd, 2020 – having passed) anyone to provide information to the CSB under that rule.