Safety
While the US Environmental Protection Agency (EPA) is loosening compliance demonstration for gasoline terminals, a separate proposal still awaiting a final decision points the other way.
Its proposed amendments to the air toxics standard for marine tank vessel loading operations would bring flare monitoring at loading terminals up to the level required at petroleum refineries.
EPA proposed the changes to 40 CFR part 63, subpart Y, on Monday, 2 March 2026, and published them in the Federal Register on Wednesday, 4 March 2026 (docket EPA-HQ-OAR-2025-0207).
Comments closed on Monday, 20 April 2026. A final rule had not been located at the time of writing [unverified – check EPA docket for final action].
The standard dates from September 1995, and EPA completed its most recent risk and technology review in April 2011.
This time EPA identified around 190 facilities subject to the standard, up from 152 in the 2011 review. Around 25 are also subject to the reasonably available control technology (RACT) standards.
At present, flares at marine loading facilities need only pilot flame monitoring. EPA proposes enhanced monitoring modelled on its 2015 rule for petroleum refineries.
That would cover the pilot or flame, visible emissions, maximum flare tip velocity, flare gas flow and combustion zone net heating value (NHVcz).
Perimeter air-assisted flares would also be monitored for dilution net heating value (NHVdil).
EPA expects most facilities to be able to use a sampling demonstration to establish a fixed minimum net heating value, rather than installing continuous net heating value monitors.
That choice will determine whether sites need on-line calorimeters or gas chromatographs, or whether periodic sampling and laboratory analysis will suffice.
EPA found no technological developments for non-flare control devices that would cut HAP emissions further. It instead proposes periodic performance testing at least once every 60 calendar months for control devices.
Where a device relies on a VOC continuous emission monitoring system (CEMS), an annual relative accuracy test audit (RATA) would be required.
The proposal would also remove the maintenance allowance and the affirmative defence provisions, so the standards would apply at all times, including during startup, shutdown and malfunction. Compliance reports would move to electronic submission.
EPA is not adding equipment leak standards for liquid components or changing the 10,000 ppmv leak definition. It proposes that annual EPA Method 21 monitoring takes place during active loading.
Repairs would be due within 15 days of identification or before the next loading, whichever is later, and confirmed by re-monitoring.
Underwater or floating vapour lines, and components that would need inspection above two metres, could be checked using audible, visual and olfactory (AVO) methods instead of Method 21.
Leaks found this way would have to be repaired without Method 21 confirmation.
EPA estimates that, if finalised, the amendments would cut emissions by around 3,500 tons per year of VOCs and around 280 tons per year of HAPs.
That includes around 118 tons of hexane, 62 tons of benzene and 56 tons of methanol a year.
It puts industry costs at around US$1.6–1.7m a year, with an average cost-to-sales ratio of 0.13 per cent.
Set beside EPA’s gasoline distribution reconsideration, the proposal shows that the agency’s current approach is not uniform.
Where it has found a monitoring gap, such as flares regulated only by pilot flame detection, it has proposed to close it.
For terminal operators, the practical questions are which net heating value demonstration EPA accepts in the final rule and how long they will have to comply.
Moving from pilot flame monitoring to the full refinery-style set of parameters is not a quick change. The proposal does not change the requirement to monitor during loading, which is when the emissions occur.
PIN 27.5 Oct/Nov 26