Preliminary Critique of

Flint Hills’s (Koch) Limited Review of Benzene Air Data in Corpus Christi

Neil J. Carman, Ph.D.

1) Corpus Christi has one of the highest concentrations of oil refineries (6) in the nation second only to Los Angeles (9), and ahead of Houston (5), Salt Lake City (4) Texas City (3), Port Arthur (3), and Lake Charles (3). Several cities have two and many have one refinery. Environmental Protection Agency (EPA):

2) In 2007, Corpus Christi (Nueces County) ranked third in the U.S. behind Texas City (Galveston County) and Houston (Harris County) in refinery benzene air emissions among over 100 communities with oil refineries. The EPA’s Toxic Release Inventory (TRI) data is searchable for petroleum refining sector at:

• Texas City: 185,216 pounds benzene

• Houston:152,198

• Corpus Christi: 136,200

• Lake Charles113,921

3) Benzene is a high volume pollutant in the oil refining sector according to EPA, and a high volume pollutant in Corpus Christi. The EPA’s Toxic Release Inventory (TRI) data for Corpus Christi refinery benzene was ignored in the Flint Hills presentation.

In 2007, six Corpus Christi oil refineries reported 68.1 tons of benzene air pollution, and this large quantity poses a significant health risk to the community as a hazardous air pollutant under the Clean Air Act.

The Flint Hills benzene presentation ignored 68.1 tons (136,200 pounds) of benzene air pollution reported by oil refineries in the 2007 TRI. The EPA’s 2007 TRI demonstrates that large quantities of airborne benzene were released by Corpus Christi oil refineries in one year and large quantities are being released each year due to the refining capacities of the Corpus Christi refineries.

Summary of Corpus Christi Oil Refinery TRI Benzene Reported in 2007 – US Rank

•. Flint Hills Resources LP - West Plant - 60,982 pounds (44.8%) #4 US

•. Valero Refining Texas LP East Plant - 38,561 pounds (28.3%)#10

•. Citgo Refining & Chem Co LP East Plt - 17,036 pounds (12.5%)#35

`•. Valero Refining Texas LP West Plant - 11,474 pounds (8.4%)#52

•. Flint Hills Resources LP - East Plant - 7,912 pounds (5.8%)#65

•. Citgo Refining & Chem Co LP East Plt - 235 pounds (0.2%)#

Total 2007 Refinery Benzene Toxic Release Inventory = 136,200 pounds (68.1 tons)

The EPA’s TRI benzene data is not based on real measurements, but derived from estimates that can be fudged so benzene appears lower in the TRI than the true emissions values, which are unknown.

4) Benzene is recognized by EPA as a known human carcinogen and has been for over 30 years.

EPA, as a result, began to adopt National Emissions Standards for Hazardous Air Pollutants (NESHAPs) regulations to require refineries to control and reduce benzene emissions in the 1980s. Refineries are required to implement multiple NESHAPs rules to reduce benzene.

Benzene is listed as a Hazardous Air Pollutant (HAP) in the 1990 Clean Air Act amendments passed by the Congress and signed into law on November 15, 1990 by President George H.W. Bush.

Hazardous air pollutants (air toxics): the most toxic substances to human health

1990 Clean Air Act amendments Title III: Air Toxics are defined as:

“Toxic air pollutants are those pollutants which are hazardous to human health or the environment but are not specifically covered under another portion of the Clean Air Act. These pollutants are typically carcinogens, mutagens, and reproductive toxins. The Clean

Air Act Amendments of 1977 failed to result in substantial reductions of the emissions of these very threatening substances. In fact, over the history of the air toxics program only seven pollutants have been regulated.”

5) Benzene health effects summary by EPA of the cancer risk, acute effects, chronic (noncancer) effects, and reproductive/development effects.

According to EPA, cancer risk is described in the following excerpt:

“ * Increased incidence of leukemia (cancer of the tissues that form white blood cells) has been observed in humans occupationally exposed to benzene. (1,4)

* EPA has classified benzene as a Group A, known human carcinogen. (4)

* EPA uses mathematical models, based on human and animal studies, to estimate the probability of a person developing cancer from breathing air containing a specified concentration of a chemical. EPA calculated a range of 2.2 x 10-6 to 7.8 x 10-6 as the increase in the lifetime risk of an individual who is continuously exposed to 1 µg/m3 of benzene in the air over their lifetime. EPA estimates that, if an individual were to continuously breathe air containing benzene at an average of 0.13 to 0.45 µg/m3 (1.3 x 10-4 to 4.5 x 10-4 mg/m3) over his or her entire lifetime, that person would theoretically have no more than a one-in-a-million increased chance of developing cancer as a direct result of continuously breathing air containing this chemical. Similarly, EPA estimates that continuously breathing air containing 1.3 to 4.5 µg/m3 (1.3 x 10-3 to 4.5 x 10-3 mg/m3) would result in not greater than a one-in-a-hundred thousand increased chance of developing cancer, and air containing 13 to 45 µg/m3 (1.3 x 10-2 to 4.5 x 10-2 mg/m3) would result in not greater than a one-in-ten thousand increased chance of developing cancer. For a detailed discussion of confidence in the potency estimates, please see IRIS. (4)”

6) Benzene is defined as a Class “A” carcinogen. Class “A” carcinogens are the most well-documented followed by Class B1 and B2. Class “A” indicates there is sufficient evidence in human studies to support a cancer-causing effect.

Toxicological profile for benzene August 2007 by the Agency for Toxic Substances & Disease Registry:

7) The only safe threshold of exposure with certainty for a known carcinogen such as benzene is zero. Flint Hills mistakenly assumes safe levels of benzene above zero or minimum level of detectability exist in theory near refineries even as science confirms that benzene is a highly potent cancer-causing agent: one exposure to benzene may cause cancer.

Evidence from new a benzene study in 2004 cited in the New York Times (Broad Study Suggests a Lower Tolerance for Exposure to Benzene by Andrew Revkin), found effects in a large group of 250 factory shoe workers exposed to low concentrations of benzene:

A similar December 2, 2004 article in the New York Times quoted several scientists:

“The work also reinforces the idea that for some chemicals that are commonplace and hazardous, there will be no level that science will ever define as harmless, said Dr. Martyn T. Smith, an author of the paper and a toxicologist at the University of California, Berkeley.”

New evidence of adverse health effects at low benzene concentrations has been published since 2004. Two leading benzene toxicology researchers are Drs. Stephen Rappaport (School of Public Health, University of North Carolina, Chapel Hill SBRP) and Martyn T. Smith (School of Public Health, University of California, Berkeley SBRP), who have collaborating together on investigations of various aspects of human metabolism of benzene. Their research findings indicate that benzene guidelines and standards need to be lowered significantly, and, indeed, the only safe level of exposure to benzene may be zero.

According to an April 1, 2009 article (Research Brief 172, Superfund Research Program/SRP, New Understandings of Benzene Metabolism and Implications for Risk Assessments), recent research by Rappaport and Smith states:

“They have developed and applied biomarkers of exposure, namely, benzene in breath and urine, benzene metabolites in urine, and protein adducts of benzene metabolites in blood; and investigated biomarkers of effect, in the form of changes to gene expression and DNA. They measured many of these biomarkers in over 1000 benzene-exposed workers and controls in numerous studies, as part of collaborations with Drs. Lan and Rothman at the National Cancer Institute (NCI) and Dr. Qu at New York University. Among their many findings, these collaborators have shown that:

A. At levels below 1 ppm, benzene causes a lowering of circulating blood cells

B. Benzene is toxic to progenitor cells (the unspecialized "parent" cells from which all other blood cells develop) and particularly to early progenitor cells

C. Biomarkers of exposure to benzene increase with benzene air concentrations, but the exposure-biomarker relationships are non-linear, with biomarker levels decreasing per ppm of benzene exposure at levels above 1 ppm”

“Taken together, the results of research led by Drs. Rappaport and Smith suggest that the leukemia risk associated with exposures to environmentally relevant levels of benzene could be substantially greater than currently assumed for the general population, and even higher for subgroups with specific genetic susceptibilities. These findings introduce new complexities into the already significant challenges faced by environmental and public health practitioners charged with making decisions regarding regulatory actions and potential cleanup costs, estimated in the billions of dollars.”

For More Information Contact:

Martyn T. Smith

Environmental Health Sciences Division

School of Public Health

Berkeley, CA 94720-7356

Tel: 510-642-8770

Email:

Stephen M. Rappaport

Department of Environmental Health Sciences

School of Public Health

Berkeley, CA 94720-7360

Tel: 510-642-4355

Email:

Also cited are in the NIEHS article above are three new studies by Rappaport and Smith:

* Lan, Qing, Luoping Zhang, Min Shen, William J. Jo, Roel Vermeulen, Guilan Li, Christopher Vulpe, Sophia Lim, Xuefeng Ren, Stephen M. Rappaport, Sonja I. Berndt, Meredith Yeager, Jeff Yuenger, Richard B. Hayes, Martha S. Linet, Songnian Yin, Stephen Chanock, Martyn T. Smith, and Nathaniel Rothman. 2009. Large-scale evaluation of candidate genes Identifies associations between DNA repair and genomic maintenance and development of benzene hematotoxicity. Carcinogenesis. ( 30(1):50-58. doi:10.1093/carcin/bgn249 (

* Rappaport, Stephen M., Sungkyoon Kim, Qing Lan, Roel Vermeulen, Suramya Waidyanatha, Luoping Zhang, Guilan Li, Songnian Yin, Richard B. Hayes, Nathaniel Rothman, and Martyn T. Smith. 2009. Evidence that humans metabolize benzene via two pathways. Environmental Health Perspectives. doi:10.1289/ehp.0800510 ( (in press)

* Ren, Xuefeng, S. Lim, Martyn T. Smith, and Luoping Zhang. 2008. Werner syndrome protein, WRN, protects cells from DNA damage. Toxicological Sciences. ( 107(2):367-375. doi:10.1093/toxsci/kfn254 (

8) Aromatic air pollution excluded from review: Refinery emissions of large quantities of benzene-containing compounds may contribute to human benzene exposures

According to the 2007 TRI data, the refineries listed 3-4 dozen benzene-containing compounds called “aromatics” that are emitted in large quantities by the Corpus Christi oil refineries as gases or are absorbed on or agglomerated into tiny soot particles called PM2.5 fine particles. However, most of the benzene-compounds are not measured on the soot particles and are simply assumed to be “particulate matter” in the TCEQ’s criteria emissions inventory rather than as specific benzene-containing chemicals, especially highly toxic chemicals called polycyclic aromatic hydrocarbons (PAHs).

One PAH is a super carcinogen, even more potent than benzene, is benzo-a-pyrene (five fused benzene rings). The Corpus Christi refinery community health concern is that several aromatics (benzene compounds) may contribute to benzene in the human body when a person inhales them. More unidentified potential benzene exposure is possible.

Corpus Christi oil refineries reported releasing the following aromatic or benzene-compounds: Ethylbenzene, toluene, xylenes, cresols, styrene, cumene, phenol, biphenyl, 1,2,4-trimethylbenzene, polycyclic aromatic hydrocarbons (PAHs), dibenzofuran, dioxins, and furans. Polycyclic aromatic hydrocarbons (PAHs), such as benzo-a-pyrene, are more toxic and carcinogenic than benzene itself.

Six Corpus Christi Refineries reported in the 2007 Toxic Release Inventory:

2006 = 609,200 pounds Aromatics vs 153,100 pounds benzene

2007 = 487,600 pounds Aromatics vs 136,200 pounds benzene

Total: 1,096,800 pounds Aromatic / Benzene Compounds vs 289,300 pounds Benzene

Approximately ~80% aromatic or benzene compounds vs ~20% benzene were reported in the TRI for 2006 and 2007, but more will be hidden in the particulate matter such as PM2.5 fine particles. For example. PM2.5 particles contain large concentrations of carcinogenic Polycyclic Aromatic Hydrocarbons (PAHs). Note that there are several dozen known PAHs that may be present in the PM2.5 soot particles. Analysis is needed.

9) Benzene compounds are closely associated with fine PM2.5 particulate matter.

Benzene and benzene-containing compounds are known to be carried within PM2.5 and PM10 particulate matter emissions from Corpus Christi oil refineries. PM2.5 is largely considered to be agglomerated or stuck together PAH compounds and is one reason why PM2.5 fine particle pollution is widely regarded by public health scientists as made of highly toxic substances. The EPA first adopted in 1997 a PM2.5 National Ambient Air Quality Standard, both a 24-hour and annual standard, in recognition of the toxicity of the PM2.5 fine particulate matter.

Six Corpus Christi oil refineries reported in TCEQ’s 2006 Industrial Emissions Inventory a total of 1,837.0766 tons as PM2.5 fine particle pollution and 1,978.7711 tons as PM10 particles.

The PM2.5 fine particle fraction accounted for nearly 93% of the PM10 emissions. Calculated as pounds, the refineries emitted another 3,674,153 pounds of benzene-loaded air pollution as PM2.5 particles. Combining from 2006 the PM2.5 volume of 3,674,153 pounds + 609,200 pounds of TRI reported Aromatic / Benzene Compounds yields up to 4,283,353 pounds of potential benzene-compounds.

Conclusion: The 4,283,353 pounds of potential benzene-compounds is nearly 28 times the 153,100 pounds benzene reported in 2006. This clearly demonstrates that there is a great deal more benzene and benzene-containing compounds being emitted than as benzene alone. Benzene gaseous emissions are under estimated most likely as well.

10) Benzene Potential-To-Emit (PTE) in refinery permits.

Benzene PTE is listed in refinery air permits was ignored. Another area excluded in Flint Hills’ benzene analysis is TCEQ air permits for each refinery listing benzene as an hourly pollutant, showing maximum PTE that air modeling may consider. Based on 2007 TRI airborne benzene and TCEQ air permits listing hourly or annual benzene, it’s factually inaccurate and grossly misleading for Flint Hills to speculate that there could not be that much benzene in Corpus Christ (i.e., as found in the biomonitoring study). The 68.1 tons is a large quantity of benzene. Calculated in grams, the 2007 benzene TRI refinery airborne releases yields more than sixty-one million grams of benzene: 61,780,320 grams in 2007 and ~7,053 grams averaged per hour emitted as a total. A toxic hit of airborne benzene is far less than a single gram. TCEQ Air Permit Reviewer Reference Guide (APDG 5944) Potential to Emit Guidance (December 2008):

11) Emissions inventory (EI) errors in the hourly and annual benzene data pose a concern.

Erroneous emissions factors are still used to estimate benzene EI’s at refineries. In a July 9, 2008 letter sent by Mayor Bill White of Houston (Data Quality Act petition) to EPA, scientific concerns were presented over errors in the antiquated emissions factors for calculating volatile organic compound emissions (benzene), and EPA-approved emissions factors are crucial in calculating hourly and annual estimates ultimately used to calculate annual EI’s at oil refineries. enviro.blr.com/news.aspx?id=104607

The EPA responded April 7,2009 to Mayor White’s letter from the City of Houston and acknowledges errors exist in emission factors that underestimate the benzene EI:

Improper use of EPA’s Tank Model 4.09D in calculating benzene emissions. Refinery Tank benzene emissions are not measured, for example, and the benzene EI for heated tanks is most likely highly inaccurate. EPA needs to audit the heated tanks for benzene at six Corpus Christi refineries since the tank benzene EI is likely erroneous in under predicting benzene vapor losses due to EPA Tank Model 4.09D. The TCEQ’s website indicates use of EPA Tank Model 4.09D and provides instructions for applying it:

Refineries may be improperly modeling benzene emissions from heated storage tanks with EPA’s Tank Model 4.09D used in Texas. The TCEQ has yet to require refineries to use proper model inputs, including the correct tank temperature. Benzene tank losses may be higher than reported. Heated tanks may have temperatures well over 100 degrees Fahrenheit (F) even up to 130 degrees F. However, if a tank temperature input of 100 degrees F is used rather than 125 or 130 degrees F, benzene vapor losses will be greatly under calculated by several tons of benzene per year even if the weight percent of benzene is no greater than 0.2%. A concern is that most of the six oil refineries in Corpus Christi have several large heated tanks in their tank farms and are potentially under calculating benzene emissions.

12) Refinery excess emissions events are another source of benzene air pollution.

Failure to review excess emissions due to refinery malfunctions and upset conditions when benzene emissions may spike to higher concentrations. Six refineries reported 130+ excess emissions events 2004-09. Several had benzene releases. Two large spikes in air pollution were measured this spring by monitors near refineries but is unknown. Flint Hills ignored such spikes. TCEQ’s website for Reports of Air Emission Events in Texas and searches can be performed in the “Air Emission Event Report Database 2003-09 at:

13) The Flint Hills benzene review excluded limitations in ambient air monitoring networks and data collected.

Monitors only detect benzene where they are located, and holes and gaps exist in every monitoring network, as EPA and TCEQ know, where higher benzene concentrations are missed, if no monitors are there, or are not sampling. Refineries contain thousands of fugitive emission points and dozens of point sources of benzene. Ambient air monitors only measure a fraction of oil refinery benzene air emissions released by the multitude of facilities.

14) Monitors have other limitations ignored in Flint Hills’s benzene review. Monitors are at fixed locations and may not detect the same pollutant levels as a living, breathing person who resides near refineries. What the actual human exposure is to benzene is not answered by monitor data.

15) Hourly and annual benzene averages may not represent the worst-case benzene exposures to people living near the refineries, since peak benzene levels may be higher. It is disingenuous to suggest that monitor averages, whether hourly or longer, are equal to benzene exposures to people.

16) Air modeling of maximum hourly benzene impacts was not considered, in which higher benzene hourly concentrations may be predicted off-site, if refinery-wide benzene air modeling was conducted. Higher hourly benzene air modeling concentrations were excluded.

17) The human body is biologically unique by breathing in and metabolizing benzene versus an inert monitor. A human body does not mathematically average exposure to benzene even if a monitor averages. Comparing average monitor values to what humans may be exposed to has many limitations, since many variables need to be evaluated.