EPA and U.S. Department of Justice have been requiring monitoring on a plant-by-plant basis, through legal settlements – called consent decrees – when the EPA sues companies for Clean Air Act permit violations. Under these consent decrees, companies must install and operate benzene air monitors around their plants and publicly report results on their company websites. Like the 2015 EPA rule for refineries, these consent decrees require companies to investigate and clean up benzene emission sources whenever fenceline monitors show annual rolling average concentrations of benzene exceeding EPA’s action level (excluding any contribution from offsite or non-refinery sources). One plant in Western Pennsylvania, owned and operated by Shell Chemical Appalachia LLC, is required to conduct fenceline monitoring as a result of a lawsuit brought by the Environmental Integrity Project and Clean Air Council.

In early 2024, EPA finalized fenceline monitoring rules for a subset of chemical plants that produce commodity chemicals, elastomers, and resins while emitting hazardous air pollutants into surrounding communities. The rules will require chemical plants to conduct fenceline monitoring for six pollutants: benzene, 1,3-butadiene, chloroprene, ethylene dichloride, ethylene oxide, and vinyl chloride if they emit, manufacture, store, or otherwise use these chemicals. Plants will begin monitoring by June 2026, with the first year of data to be published in June 2027. However, the 2024 rules do not cover many chemical plants that emit the same pollutants.

The EPA's Petroleum Refinery Sector Rule requires petroleum refineries to monitor benzene concentrations at refinery fencelines and establishes procedures to address elevated benzene levels. EPA established an action level of 9 micrograms per cubic meter (µg/m³) which, if exceeded, requires refineries to conduct a root cause analysis and take corrective action to reduce benzene emissions.

How the action level is calculated for a particular facility is best summed up in the EPA’s Office of Inspector General’s report on the subject:

“A refinery must continuously measure concentrations at each of its monitors over two-week periods and report those concentrations to the EPA at the end of each quarter. These two-week periods are referred to as sampling periods. After each sampling period, a refinery must determine whether its annual average concentration for the most recent 26 two-week sampling periods is over 9 µg/m³. The concentration that a refinery uses to determine whether it is over the action level is referred to as the delta concentration, or delta c. The delta c accounts for benzene in the ambient air that may not be directly attributable to the refinery. The delta c is determined by subtracting the lowest benzene concentration measured by any of a refinery’s monitors from the highest benzene concentration measured by any of a refinery’s monitors for each two-week sampling period. If the refinery’s average delta c for the preceding 26 two-week sampling periods is above 9 µg/m³ the refinery has exceeded the action level.”

Example: If during a two-week sampling period, the highest monitor reading is 12 micrograms per cubic meter (µg/m³), and the lowest monitor reading is 0.8 µg/m³, the benzene Δc value (or delta value) would be 11.2 µg/m³. The annual rolling average Δc is then calculated using the 26 most recent sampling periods. When this annual rolling average Δc is greater than 9 µg/m³, the facility has exceeded EPA’s action level threshold.

The regulations also allow refineries with site-specific monitoring plans approved by EPA to make further downward adjustments to exclude benzene that comes from specific sources, including plants outside the refinery’s boundary and certain sources within the refinery that are exempt from the regulation (such as benzene storage tanks). These adjustments allow some refineries to avoid triggering the EPA action level that requires cleanup of the plant’s emission sources, even when their monitors report very high benzene levels. Unlike refineries, chemical plants do not have site-specific monitoring plans.

Refineries and chemical plants operate continuously, barring shutdowns for emergencies or maintenance, emitting benzene and other pollutants that impact air quality and threaten the health of workers and those who live, work, or otherwise spend time near the plant. Prolonged exposure to benzene can lead to blood disorders, threaten the immune system, and increase the risk of leukemia. Studies have also shown that shorter exposures during vulnerable periods, like pregnancy and early childhood, can lead to adverse health outcomes. The actual risk to public health will depend on how much and for how long benzene drifts into surrounding neighborhoods. The benzene data reported to EPA, on company websites, and available on this website are limited to concentrations measured at facility fencelines – actual concentrations in surrounding communities will differ based on variables like distance, wind patterns, and other potential emission sources. EPA has made clear that the nine microgram per cubic meter “action level” does not establish a new air quality standard; in fact, much lower concentrations can be dangerous to vulnerable populations.

Long-term Health Risk

Even relatively “low” benzene levels can harm health. The California EPA, for example, has determined that continuous or repeated eight-hour exposures (i.e., workplace exposures) to three or more micrograms per cubic meter of benzene over about eight to nine years could increase the risk of noncancer health effects, such as damage to blood cells and a weakened immune system. The risk from benzene exposure increases with prolonged and additional exposure, especially among sensitive populations like pregnant people and young children.

In addition to the noncancer health impacts of benzene exposure, the pollutant is also a known carcinogen that can increase the risk of developing leukemia. EPA estimates that breathing benzene at a concentration as low as 0.13 micrograms per cubic meter over a lifetime could result in up to one additional cancer diagnosis per one million people exposed. As benzene levels rise, those risks increase. By EPA’s estimate, the additional cancer risk at a concentration of 3 µg/m³ over a lifetime, for example, would be 6 to 23 in a million. The California EPA has published a higher unit risk that suggests that long-term exposure to as little as 0.03 micrograms per cubic meter can increase cancer risk by one in one million. At 3 µg/m³, the additional cancer risk based on California EPA’s estimate would be 87 in a million.

Short-term Health Risk

Exposure to high levels of benzene over shorter periods of time can also lead to adverse health effects. Short-term exposure can cause headaches, dizziness, and irritation to the eyes, skin, and respiratory tract. It could also decrease blood cell counts, including in vulnerable groups like pregnant people and young children. The U.S. Agency for Toxic Substances Disease Registry (ATSDR) estimates that exposure to benzene concentrations above 29 micrograms per cubic meter for 1-14 days can increase the risk of noncancer health effects like a weakened immune system.

Health Risk Levels

Potential health risks are estimated based on many factors, including the length of exposure, concentration, how people are exposed, type of health effect, age, and other vulnerabilities. There are many uncertainties and assumptions that go into calculating risk and different government agencies have established different thresholds that estimate when a chemical poses an increased health risk to human health. Some risk levels for benzene include:

California Environmental Protection Agency’s Office of Environmental Health Hazard Assessment (OEHHA) Reference Exposure Levels (RELs): An estimated concentration at or below which there would likely be no noncancer health effects anticipated for the specified exposure duration.

U.S. Agency for Toxic Substances Disease Registry (ATSDR) Minimal Risk Levels (MRLs): An estimate of daily exposure to a substance that would likely not increase the risk of a noncancer health effect over the specified exposure duration.

U.S. Environmental Protection Agency Cancer Risk Levels: EPA publishes cancer potency estimates known as “unit risks” (in units of risk per µg/m3), which can be converted into exposure concentrations associated with certain levels of risk. For benzene, EPA has published a range of unit risks which suggest that long-term exposure to 0.13 to 0.45 micrograms per cubic meter creates a cancer risk of one in one million. The California EPA has published a higher unit risk that suggests that long-term exposure to as little as 0.03 micrograms per cubic meter can increase cancer risk by one in one million.

U.S. Environmental Protection Agency Noncancer Risk: The reference concentration for inhalation exposure (RfC) is a concentration that, with continuous inhalation exposure, is likely to be without an increased risk of noncancer health impacts.

Refineries submit quarterly benzene reports to EPA and can be downloaded using EPA’s WebFIRE Report Search . Refineries are required to submit these reports to EPA within 45 days of the quarter’s end, though there may be up to a 30-day delay for the reports to appear on the website. View detailed instructions on downloading EPA’s refinery reports here .

Chemical plants post monitoring reports to their own websites biweekly and are not available on EPA’s WebFIRE report search. Those reports can be found here:


Some refineries or chemical plants section off parts of the facility and provide multiple reports (e.g. the refinery and the refinery dock). EIP has counted these “sub-refineries” as separate facilities, consistent with how the refineries report to EPA or chemical plants report on their websites.

EIP downloads the reports as they are made available (every three to four months for refineries and biweekly for chemical plants) and compiles the data in an online database. There may be a lag between when EPA makes data available and when this website is updated. However, our goal is to keep the database as current as possible.

Data Limitations

The benzene monitoring data is collected at facility perimeters and are not direct measures of benzene within a community, residential neighborhoods, or homes. How much benzene drifts into surrounding areas can vary, in part due to meteorological variables like wind.

The fenceline monitors are passive samplers that measure average benzene concentrations over a two-week period. Concentrations over shorter periods during the two-week periods could have been much higher or lower.

Because of the two-week sampling periods and delayed reporting, benzene concentrations shown do not reflect real-time benzene concentrations.

Previous Reports by EIP

Environmental Justice and Refinery Pollution: Benzene Monitoring Around Oil Refineries Showed More Communities at Risk in 2020, April 2021

Monitoring for Benzene at Refinery Fencelines, February 2020

The Environmental Integrity Project gathers this data from the Environmental Protection Agency and company reporting websites. If you have additional questions about this data, or have issues accessing the data via Xata's API, send an email to info@environmentalintegrity.org and someone will get back to you as soon as possible. If you have a question about the visuals on a particular page, please navigate to that page and copy paste the link from that page in your email message. That'll make quickly answering your question much easier!