AFPM Vice President of Regulatory Affairs David Friedman provided comments at an EPA public hearing in Wilmington, CA on July 16, 2014.

AFPM appreciates the opportunity today to discuss our views on the EPA’s proposed refinery Risk and Technology Review rule. The American Fuel & Petrochemical Manufacturers (“AFPM”) is a national trade association of more than 400 companies, including virtually all U.S. refiners and petrochemical manufacturers. AFPM members operate 122 U.S. refineries comprising approximately 98% of U.S. refining capacity.   

Air quality has improved significantly in the United States over the past three decades and our industry has been a significant contributor to this improvement.  Since 1990, our industry has invested $137 billion to reduce emissions from our plants and make cleaner fuels. On the air toxics side, we have also made great strides in reducing emissions. Total hazardous air pollutants from the refining sector have been reduced by nearly two-thirds since 1990, while fugitive HAPs have been reduced by 79 percent and point-source HAP emissions have been reduced by nearly 40 percent.  

Despite these significant reductions, EPA is proposing some requirements in this rule that simply are not justified by the level of refinery risk that the Agency found in its analysis. EPA in its previous Risk and Technology Review rulemaking in 2008 determined that the risks posed by the refining industry to be acceptable and therefore no further actions were required. Although signed by the EPA Administrator, due to a change in administrations, that rule was never published in the Federal Register and was subsequently withdrawn. 

After EPA conducted a massive and unprecedented industry data collection effort from refiners in 2011, the risk levels found by EPA in its most recent analysis, performed for this rule, are similar to those found in 2008, still within an acceptable range.  One quarter of the refineries had maximum individual risk levels below one in a million and two-thirds had risk levels below 10 in a million.  To give a perspective on risk levels,  the odds of being hit by lightning in a given year are one in 960,000 (National Weather Service). According to the National Safety Council the odds of dying in a passenger car in a single year are one in 21,132 and the odds of dying as a pedestrian in a lifetime are one in 623. 

Unfortunately, we are faced with a rule with significant costs but with little or no health or environmental benefits. EPA estimates that this rule will cost $240 million, but our members estimate that it will cost in excess of a billion dollars. Of even greater concern is that the health benefit gains are insignificant by any measure, meaning this rule will have financial impacts to the economy with essentially no measureable benefits. 

Every refinery will be required to install fenceline monitors at significant on-going cost. Yet, as I mentioned earlier, many of these refineries have very low levels of risk.  We don’t understand how imposing this one-size fits all approach to every refinery will improve public health, particularly when according to EPA analysis a number of these refineries pose no health risk to the surrounding communities.    

The EPA rule proposes a rolling annual average benzene concentration action level of 2.8 parts per billion (ppb) at the refinery fence line, measured every two weeks from 20-30 samples collected around the refinery fenceline perimeter. Exceeding this “threshold” would trigger a root cause analysis and corrective action and could be deemed a violation by EPA.  However, the threshold value is not an indicator of risk – the stated purpose of this monitoring is to check EPA’s modeling of fugitive emissions. We fail to see how it can achieve this goal. Further, EPA’s own studies in the mid-1990s confirmed no effect on personal exposure of living close to major fixed sources of benzene including oil refineries, chemical plants and storage tanks. 

We have concerns with the flaring provisions in this rule. The recently adopted NSPS Ja flare rules resulted in instrumentation costs on the order of millions of dollars per flare and will result in significant overall reductions in already very limited flaring emissions. This proposal would require significant additional costs over and above those from the Ja rule and yet has few synergies with the earlier Ja investments and results in very little incremental benefit.  Any proposed flare combustion efficiency standards should (1) only address steam-assisted flares, (2) focus on oversteaming and (3) limit the need for the wasteful addition of natural gas to the flares.   

Pressure relief valves (PRVs) are designed to protect plant personnel, the public and equipment, and EPA should not change the MACT floor work practice for PRVs. PRVs relieve very infrequently and for short periods of time with very little total HAP emissions. While monitoring such releases to ensure that PRVs have properly closed makes sense and most companies already do so, prohibiting atmospheric releases will require routing PRVs to additional new flares.  This will require massive investments both for these new flares and for flare headers to control releases of rare events.   

EPA has greatly overstated the risks that coking units pose, and the proposed depressurization requirements go well beyond the new requirements set by the recently promulgated refinery NSPS rule. EPA has greatly underestimated the costs of controlling these emissions, controls that result in virtually no benefit. 

One final note regarding process. EPA reached a settlement agreement earlier this year that requires the rule to be completed 11 months after the rule was proposed in May. This simply does not allow enough time for everyone to properly review risk information and for EPA to finalize a rule that is defensible.  In 2007-2008, EPA needed more than 18 months to complete the refinery RTR rule once it was proposed. Developing a good rule that properly determines risk and addresses solutions to lower that risk is more important than meeting an artificial deadline. We hope that the Agency and the plaintiffs will reconsider the tight timeframe and suggest that the rule be finalized at a later date in 2015. This will make for a better result for everyone.

Thank you again for this opportunity to speak and I am happy to answer any questions.