Fox, PE 745 White Pine Ave. Rockledge, FL 32955 321-626-6885 May 4, 2017 Rachael Koss Adams Broadwell Joseph Cardozo 601 Gateway Boulevard, Suite 1000 South San Francisco, CA 94080-7037 Dear Ms. K055: Per your request, I have reviewed the FluxSense Report1 that compares real-time measurements of Volatile Organic Compounds (VOCSF, Nitrogen Oxides Sulfur Oxides and Benzene, Toluene, Ethleenzene, and Xylenes (BTEX) at six refineries and one tank farm in the South Coast Air Quality Management District (SCAQMD). These six refineries include the Tesoro Carson re?nery, the largest refinery included in the study and the second largest in the SCAQMD, as well as a separate study at the Tesoro Carson tank farm. As I explain below, the FluxSense Report indicates that the Draft Environmental Impact Report (DEIR) for the Tesoro Los Angeles Refinery Integration and Compliance Project (Project)3 significantly underestirnated VOC emissions and health risks due to the Project. The DEIR estimated VOC and BTEX emissions using canned, generic emission factors that are not specific to the subject refineries. These emission factors have long been known to underestimate refinery emissions, based on real-time Flu xSense lnc., Emission Measurements of VOCs, N02 and 502 from Re?neries in the South Coast Air Basin Using Solar Occultation Flux and Other Remote Sensing Methods, Final Report, April 11, 2017 (FluxSense Report), Exhibit 1. 3 The VOCs measured in the FluxSense Report are total alkanes minus methane (FluxSense Report, 8, 31: "The main objective of this study was to quantify the total gas emissions of non-methane VOCs (alkanes and BTEX), N02, 50; and methane from six major refineries in the Los Angeles Thus, these measurements are consistent with the SCAQMD de?nition of VOCs as ozone precursors. SCAQMD Rule 102 defines VOCs as: ORGANIC COMPOUND (VOC) is any volatile compound of carbon, excluding methane, carbon monoxide, carbon dioxide, carbonic acid, metallic carbides or carbonates, ammonium carbonate, and exempt compounds.? Methane is typically treated separately from other VOC emissions because it has much lower ozone formation potential. 3 Environmental Audit, Inc., Tesoro Los Angeles Refinery Integration and Compliance Project, Draft Environmental Impact Report, Submitted to: South Coast Air Quality Management District, SCH No. 2014091020, March 2016 Available at: . 1rov home libra documents- su art-material lead-a en - ermit? ro'ects. monitoring at other refineries.4 When Project VOC and BTEX emissions are revised to use site-specific measurements, the Project results in highly significant air quality and health risks. The responses to comments on the DEIR and drafts sections of the Final Environmental Impact (FEIR), obtained through Public Record Act Requests (PRAs) thus far also do not acknowledge this new information. Thus, the DEIR should be revised to accommodate this new information and recirculated for public comment. The SCAQMD sponsored a series of measurement projects to study industrial emissions using Optical Remote Sensing (0R5) methods. The results of these studies at six refineries in the SCAQMD were recently reported.5 One of the refineries included in this study is the Tesoro Carson facility, designated as Refinery A in the study. Refinery A can be identified as Tesoro Carson from the reported crude capacity (257,300 bbl/ day)5 and by comparing the aerial photographs in the FluxSense Report with those in the The Tesoro Carson refinery was monitored for 15 days during the period August 28 to November 10, 2015,a the longest of any of the six refineries. The tank farm at this refinery was also monitored.9 This study did not include the Tesoro Wilmington refinery, its much smaller (257,300 bbl/ day vs. 104,500 bbl/ day) neighbor proposed to be more fully integrated under the Tesoro Los Angeles Refinery Integration and Compliance Project (Project).10 The FluxSense Report indicates the Carson refinery collaborated with the SCAQMD in conducting the The FluxSense study demonstrates that the increase in volatile organic compound (VOC) emissions from the Project will result in a See, for example, John K.E. Johansson and others, Emission Measurements of Alkenes, Alkanes, 801, and 02 from Stationary Sources in Southeast Texas over a 5 Year Period Using 50F and Mobile DOAS, journal of Geophysical Research: Atmospheres, February 27, 2014 (Johansson et al. 2014: Exhibit 5 FluxSense Report, Exhibit 1. See also AGU Fall Meeting, December 2016 Abstracts at: l1ttps:[ and ht s: a u.confe)c.com a frn16 meetin a .c i Pa 180782. 6 Compare crude capacity in FluxSense Report, 5, Table 125-2 and 45 with reported crude capacities in California Energy Commission, California's Oil Re?neries; Available at it? ncn. rov almanac atroleum data refineri .html. 7 Compare FluxSense Report, Figures 10 and 13 with DEIR. Figures 2-2 (de 81), 215 (pdf 115), 217 (pdf 123). 3.2.1 (pdf142). 3.5-2 (panes). 3.7-1 (pdf 207), 4.3-1 (pdf 266), 4.3.3 (pdf 253). 5.1.1 (pdf 342). etc. 9 FluxSense Report, pdf45. 9 FluxSense Report, 16 a week and a half (September 28 to 7 October), measurements were also conducted inside the Refinery A at the main eastern tank farm") '0 DEIR, p. 1-5, 24 (?The proposed project is designed to better integrate the Wilmington Operations and Carson Operations"). 11 FluxSense Report, 32 ("it has been collaborating with SCAQMD to support this campaign and making it possible to carry out 7 days of onsite measurements?) 2 significant VOC air quality impact that was not identi?ed and mitigated in the FEIR. It also demonstrates that the increase in benzene emissions will result in a signi?cant increase in cancer risk at the maximum exposed individual residential receptor (MEIR). I. The Increase in VOC Emissions from a 6,000 Increase In Crude Throughput Is Significant The DEIR estimated the Project would increase VOC emissions by 49.09 lb/day, which is less than the SCAQMD VOC significance threshold of 55 Ib/ day.12 This estimate is based on a unit-by-unit analysis in which emissions are estimated using various emission calculation procedures.? The FluxSense report demonstrated that these calculation procedures significantly underestimate refinery emissions, including at Carson.? The luxSense Report demonstrated that the Tesoro Carson refinery underestimated its VOC emissions reported to the SCAQMD in its emission inventories by a factor of 6.4. Similarly, the six re?neries studied in the SCAQMD underestimated their VOC emissions by an average factor of 6.2, compared to those reported to the SCAQMD in emission inventories.15 A ratio of 6.2 means that the emission inventories underestimated VOC emissions by a factor of 6.2 compared to measured VOC emissions. This is consistent with results reported elsewhere for other refineries that also estimate their emissions using AP-42 and other similar methods. Johansson et a1 (2014), for example reported for refineries in Texas that, ?Despite some significant variations from year to year and from area to area, there is a clear pattern of measured VOC emissions (alkanes, ethane, and propene) exceeding reported emissions with almost an order of magnitude on average, while no similar pattern exists for SOzand Nozf?? The emission inventories for VOCs are based on calculations similar to those used in the DEIR as VOCs are not monitored, unlike and 50x, for which good agreement between inventories and measurements was found.? The average ratio between measured and reported emissions for NO): was 0.8 and for 50x, 1.5, compared to an average ratio of 6.2 for consistent with other studies. Most major refining processes that emit and 50x are continuously monitored using continuous '1 DEIR, Table 4.2-4, 234. '3 DEIR, Appendix B. FluxSense Report, Table 43, 95. '5 FluxSense Report, Table 43, 95. '6 et al. 20I4, p. 1983 (Exhibit 2). '7 FluxSense Report, Table 43, 95; Johansson et al 2014, p. 1983 (Exhibit 2). 1' FluxSense Report, Table 43, 95. emission monitoring systems (CEMSs), while VOC emissions are calculated using emission factors from AP-42, or estimated from infrequent stack tests, explaining the discrepancy between VOCs, which are grossly underestimated, and and 50x, which are much more accurately estimated in emission inventories. This supports the importance of assuring that permit conditions for VOCs include monitoring to assure they are practically enforceable. The FluxSense results can be used to estimate the increase in VOC emissions due to the Project?s estimated increase in throughput of the Re?nery, approximately 6,000 bbl/ day?. The FluxSense Report demonstrated that from 0.017% to 0.045% or an average of 0.024% of the crude oil throughput at the six studied refineries is emitted as VOCs. At Carson, 0.020% of the crude throughput was emitted as VOCs, compared to an average of 0.024% for the six refineries that were studied. Assuming a 6,000 bbl/day increase in crude throughput (and my prior comments on the DEIR demonstrate that the actual increase could be much higher), the increase in VOC emissions based on FluxSense measurements at Carson would be 367 lb/ clay.20 Assuming the lowest measured percent VOC emitted (Refinery F), the increase in VOC emissions due to the Project would be 312 lb/ day. In sum, the increase in VOC emissions, just due to the asserted 6,000 bbl/ day increase in crude throughput, would exceed the CEQA significance threshold of 55 lb/ day by a factor of nearly six. This is a significant air quality impact that was not identified in the DEIR and which must be mitigated. As discussed elsewhere in these comments, VOC emissions increase due to increases in crude throughput is only one of several components of the Project that would increase the emissions of VOCs. Volatile organic compounds (VOCs) are converted into ozone, otherwise known as smog, in the atmosphere. The South Coast Air Basin is in extreme nonattainment with the federal 1-hour and 8-hour ozone standards and in nonattainment with the state 1-hour and 8-hour ozone standards.21 The federal and state ozone standards were exceeded on 92 and 129 days or 25% to 35% of the time in 2014.22 Infect, the area where the Project will be located has the worst ozone pollution in the entire United States. The Los Angeles - Long Beach area, where the Project is located, has remained at the top of the worst ozone pollution list for 17 out of the 18-years that the American DEIR, p. 2-27. 30 Increase in VOC emissions from a 6,000 bbl/ day increase in crude throughput [(1,086,215 mo of crude throughput at Carson) (6,000 bbl/ day increase in crude throughput due to Project) 257,300 day of crude throughput at lb/ Mmmo/ 30.4 day) 367 lb/day. =1 DEIR, Table 3.2.2. '23 DEIR, p. 3-4. Lung Association (ALA) has been ranking pollution in its annual State of the Air Report.? Seven groups of people are especially vulnerable to the effects of breathing ozone." In the Project area in 2016, these include over 18 million people that were at risk from ozone pollution, including: 4,383,662 under 18; 2,376,130 that are 65 and over; 313,246 with pediatric asthma; 1,099,027 with adult astl'una; 571,985 with COPD25: 8,096 with cardiovascular disease; and . 1,409,515 living in poverty.? The revised Health Risk Assessment? identified 23 sensitive receptors - schools, hospitals, child care facilities and churches - located in close proximity to the Project. The American Lung Association, in its 2017 State of the Air report, summarized recent research on children that would be present in these facilities due to ozone exposure. The impacts include: premature death; developmental harm; reproductive harm; asthma attack; wheezing and coughing; shortness of breath; cardiovascular harm; susceptibility to infections; lung tissue redness and swelling?3 11' Laura Parker, See the Best and Worst Places for Breathable Air in the U13, National Geographic, April 19, 2017; Available at: http:/ news.nalionalgeographic.com/ 2017/ 04/ ozone-pollution-city-rankings- particles-Clean-Air-Act/ . 3* American Lung Association, stag of the Air 2017, 2017); Available at: http:/ 2017/state-of-the-air- 2017.html (Exhibit 3). 35 Chronic Obstructive Pulmonary Disease, also known as COPD, includes and chronic bronchitis. 26 ALA 2017. p. 17 (Exhibit 3). 17 Ashworth Leininger Group, Health Risk Assessment for the Tesoro Los Angeles Refinery Integration and Compliance Project, Carson and Wilmington, California, February 2017, Appendix 8-4 (llevised HRA). 1' ALA 2017. p. 32 (Exhibit 3). I increased admission to hospitals for asthma, with younger children and those from low-income families more likely than others to need hospital admission; . some children with certain genes are more likely to develop asthma as adolescents; and I lower birth weight and decreased lung function in newborns.? The revised HRA failed to identify any of these well known, significant health impacts from elevated ozone levels that would result from the Project. Further, a major new study found evidence that people with lung cancer faced greater risk from ozone and other outdoor air pollutants. The 2016 study tracked the air pollution levels from 1988 to 2011 experienced by more than 350,000 cancer patients in California. The researchers found that the ozone and other air pollutants shortened their survival}!0 In addition, some evidence suggests that other groups, including women, people who suffer from obesity and people with low incomes, may also face higher risk from ozone.31 Numerous studies document the serious public health impacts of ozone.32 Thus, it is critical that the huge increase in ozone precursors that will result from this Project be fully mitigated. II. The Increase in Project VOC Emissions from Correcting the Underestimate in Tank VOC Emissions Is Significant The DEIR estimated that storage tanks are the major source of VOC emissions, amounting to 322.62 lb day. Of this amount, 141.64 lb/ day is from two new tanks at Wilmington and 112.51 lb/ day are from six new tanks at Carson.33 In addition, conversion of two existing fixed roof tanks to internal floating roof tanks and increased utilization of 11 existing tanks combined contribute an additional 68.4 lb/daySM I previously commented that if VOC emissions from these new and existing tanks were as little as 2% greater than estimated in the DEIR, operational VOC emissions from the Project (considering other increases and decreases as reported in DEIR Table 4.2-4) would exceed the SCAQMD daily VOC significance threshold. This would be a significant impact not disclosed in the DEIR. 29 ALA 20l7, p. 34 (Exhibit 3). 3? 5. P. Eckel and others, Air Pollution Affects Lung Cancer Survival, Thorax, v. 71 . 20l6, pp. 89 I-898 (Eckel et al. 20l6. Exhibit 4). ALA 2017, p. 33 (Exhibit 3). 31 See, for example, Exhibils 5 - ?19. 33 DEIR, Appx. 3-3, p. 3-3-45. 1' DEIR, Table 4.2-4 and Appr-S, p. 8-3-45. I identified a large number of errors and omissions in the tank calculations that could increase tank VOC emissions from the Project enough to exceed the VOC significance threshold of 55 lb/ day.35 I supported my conclusion that the method used to estimate tank VOC emissions in the DEIR (TANKS 4.09d) underestimated tank emissions by factors of two to fifteen by citing actual measurements at re?nery tanks using optical remote sensing methods similar to those used in the FluxSense Report.? The FluxSense Report used mobile Optical measurements at the Carson tank farm for eight days between September 28 and October 7, 2015 to estimate tank VOC and BTEX37 emissions. This study con?rmed that the tanks are the major source of VOC emissions at the Carson refinery, comprising 71% of the total measured VOC emissions.38 This is consistent with results reported elsewhere for refinery tank emissions?!9 The FluxSense Report also demonstrated that measured VOC emissions from the Carson Refinery (214 ton/ mo) are 6.5 times higher than reported to the SCAQMD in emission inventories (33 ton/ mo) using the same calculation procedures used in the As the majority of these emissions is from the tanks, by extension, the TANKS model or the Applicant?s use of this model selection of input parameters, such as temperature, vapor pressure, vapor molecular weight) underestimates VOC emissions. 35 Fox, Comments on the Draft Environmental Impact Report (DEIR) for the Tesoro Los Angeles Re?nery integration and Compliance Project, Los Angeles, California, June 10, 2016 (Fox DEIR Comments), Comment VI, pp. 81 - 90. 3" Fox Comments, pp. 83-84 and Table 6. 1" Benzene. toluene, and xylene (BTEX). 3' The VOC emissions from the Carson tank farm, reported as alkane flux (less methane) is 191 kg/ hr (FluxSense Report, Table 38). The total VOC emissions from the Carson Refinery including its tank farm, reported as alkane flux (less methane), is 269 kg/ hr (FluxSense Report, Table Thus, the percent of Carson VOC emissions that originates from the tanks is (191/ 269)100 71%. 3? Johansson 2014, p. 1989 (Exhibit Kihlman, Application of Solar Fl'iR Spectroscopy for Quantifying Gas Emissions, Thesis for the Degree of Licentiate of Engineering, Chalmers University of Technology, Goteborg, Sweden, Paper A, Monitoring of VOC Emissions from Re?neries in Sweden Using the Solar Occultation Flux Method, 81, 90, Table 4 (?0f the emitted gas 26% originates from the process, 31% from the crude-oil tanks, 32% from product tanks, 8% from the water treatment facility and 2% from transport related activities), Available at: webcache. oo rleusercontentcom 5e rch? =cach :motv-i A .58 - contentz uploadsz im- ugloads? pd M. Kihlman, J. Mellqvist, and]. Sameulsson, Monitoring of VOC Emissions from Re?neries and Storage Depots Using the Solar Occultation Flux Method, 2005 (thlman et a1. 2005), Table 40 (Exhibit 20). 4" FluxSense Report, Table 43, Meas. Alk BTEX 214 tons/mo; Rep. Tot VOC .- 33 tons/ mo. Thus, reported VOC emissions, which are 71% from tanks, are underestimated by a factor of 214/ 33 6.5. 7 Assuming the DEIR underestimated tank emission as reported in the FluxSense Report, by factor of 6.5, the net change in Project VOC emissions would increase from 49.09 lb/ day? to 1,422 lb/ clay?, exceeding the CEQA signi?cance threshold of 55 lb/ day by a factor of 25. Thus, the underestimate in tank VOC emissions alone is sufficient to result in a sigrdficant VOC impact. In sum, the FluxSense report confirms my DEIR comment on the significant underestimate in tank VOC emissions. Cancer Health Risks Are Significant The DEIR conducted a health risk assessment focused solely on hazardous air pollutants (HAPs) and concluded that health impacts were not significant.43 In response to comments on the DEIR, the HRA was revised. The Revised HRA also concluded that health risks are not significant. However, the Revised HRA failed to consider the results of the FluxSense report or to acknowledge the very significant health impacts that result from the highly signi?cant increase in VOCs, which are ozone precursors known to cause many very significant health impacts, discussed in Comment I. In particular, the revised HRA failed to acknowledge the synergistic impact of very high ozone levels with very high emissions of hazardous air pollutants. The Revised HRA reported an increase in cancer risk at the maximally exposed individual residential receptor of 3.7 in one million,44 which is less than the cancer significance threshold of 10 in one million. The Revised I-IRA also reported that emissions from the tanks and associated fugitive components are the major source of cancer risk at the MEIR, contributing 78% of the total cancer risk.? Benzene was reported to be responsible for 33.2% of the cancer risk at the The FluxSense Report concluded that benzene was underestimated in emissions reported to the SCAQMD by a factor of 3.2 to 202 at the six refineries where benzene was measured. The Carson Refinery underestimated its benzene emissions by a factor 4' DEIR, Table 4.2-4. 41 Revised increase in VOC emissions due to Project, assuming tank emissions are 6.4 times larger than estimated, based on DEIR Table 4.2-4 8 total project VOC emissions increase at Wilmington and Carson tanks - Regulation compliance - Regulation prior compliance: 401.15 lb/ day - 141.64] (Wilmington tanks) - 112.51] (Carson tanks) - 317.33 - 34.73 1,422 lb/day. 43 DEIR, 44, 55-56 and Appendix B. 4" Revised I-IRA, p. 8-4-34, 36: *5 Revised HRA, Table '13, 48. 45 Revised HRA, Table 12, 46. of 43.47 Assuming a factor of 43 underestimate in benzene, the cancer risk at the MEIR would increase from 3.7 in one million to 55 in one million.? In fact, the cancer risk at all of the "most exposed sensitive receptors?, which are all schools, would exceed the cancer significance threshold by a significant amount, when adjusted as for the Thus, emissions from the Project would result in a highly significant cancer risk impact at numerous nearby sensitive receptors, requiring mitigation. The major source of cancer risk is benzene emissions from tanks and associated fugitive components. As I explained in my comments on the DEIR, emissions from tanks and fugitive components can be mitigated by requiring ?oating roof tanks to be controlled with geodesic domes and by requiring the use of leakless fugitive components.50 These feasible controls were not required in the DEIR. In sum, the Project would result in significant air quality and public health impacts when the results of actual measurements at the Carson Refinery and other nearby refineries are considered. In my opinion, it is a serious violation of CEQA to fail to use the most accurate available information to evaluate air quality and public health impacts and to fail to acknowledge and mitigate the significant air quality and cancer impacts that would results from the Project in the communities surrounding the Refinery, including many schools. The DEIR should be revised to incorporate the new FluxSense measurements and recirculated for public review. Sincerely, Fox, RE. *7 FluxSense Report, Table 43, 95. 43 The cancer risk at the MEIR, assuming benzene is underestimated by factor of 43: 55x10-?. 49 Revised HRA, Table 11, 45-46. Selecting the lowest reported cancer risk at local schools, 1.12 ?10? at Wye Tech National Institute of Tech, the revised cancer risk is: - 17310". 5? Fox DEIR Comments, pp. 82, 89.