ft EPA
United States
Environmental Protection
Agency
National Center for
Environmental Assessment
Research Triangle Park, NC 27711
EPA/600/R-95/134
August 1995
Proceedings of the
Conference on MTBE and
Other Oxygenates:
A Research Update
-------
EPA/600/R-95/134
August 1995
PROCEEDINGS OF THE CONFERENCE ON
MTBE AND OTHER OXYGENATES:
A RESEARCH UPDATE
July 26-28, 1993
Fairvievv Park Marriott
Falls Church, Virginia
Sponsored by
U.S. Environmental Protection Agency
American Petroleum Institute
Oxygenated Fuels Association
Prepared by
National Center for Environmental Assessment (MD-52)
U.S. Environmental Protection Agency
Research Triangle Park, NC 27711
-------
DISCLAIMER
These proceedings have been reviewed by the U.S. Environmental Protection Agency
and approved for publication. The information presented herein represents the opinions of
the authors at the time they provided their material and does not necessarily represent the
views or policies of the sponsors (U.S. Environmental Protection Agency, American
Petroleum Institute, and Oxygenated Fuels Association).
n
-------
ACKNOWLEDGMENTS
The sponsors are extremely grateful to numerous individuals and organizations who
were instrumental in conducting and presenting the research that was the subject of this
conference and to those who contributed to the organization and conduct of the conference
and to the preparation of these proceedings.
We especially thank Colleen Rose Schwoerke of Research and Evaluation Associates,
Inc. (Chapel Hill, NC), who provided exceptional assistance in the organization of the
conference as well as in the development of the proceedings, and Marianne Barrier, •
John Barton, Suzy Bornemann, Shelia Elliott, Sandra Eltz, and Sheila Lassiter of ManTech
Environmental Technology, Inc. (Research Triangle Park, NC), who provided invaluable
graphics, editorial, and word processing support.
We also are indebted to the organizing committee for creating a conference with
stimulating chairpersons, speakers, and discussants.
in
-------
CONFERENCE ORGANIZING COMMITTEE
Dr. J. Michael Davis
Environmental Criteria and Assessment
Office (MD-52)
U.S. Environmental Protection Agency
Research Triangle Park, NC 27711
Dr. Robert T. Drew
Health and Environmental Science
Department
American Petroleum Institute
1220 L Street, NW
Washington, DC 20005
Mr. Stanley Durkee
Office of Science, Planning,
and Regulatory Evaluation (H-8105)
U.S. Environmental Protection Agency
401 M Street, SW
Washington, DC 20460
Dr. Timothy Gerrity
Health Effects Research Laboratory
(MD-58)
U.S. Environmental Protection Agency
Research Triangle Park, NC 27711
Dr. Judith A. Graham
Environmental Criteria and Assessment
Office (MD-52)
U.S. Environmental Protection Agency
Research Triangle Park, NC 27711
Dr. Kenneth Knapp
Atmospheric Research and Exposure
Assessment Laboratory (MD-46)
U.S. Environmental Protection Agency
Research Triangle Park, NC 27711
Mr. John Kneiss
Oxygenated Fuel Association
1330 Connecticut Avenue, NW
Suite 300
Washington, DC 20036
Dr. Paul J. Lioy
Environmental and Occupational Health
Sciences Institute
681 Freelingheysen Road
Piscataway, NJ 08855-1179-
Ms. Meredith Miller
Office of Mobile Sources (6406J)
U.S. Environmental Protection Agency
401 M Street, SW
Washington, DC 20460
Mr. Richard Paul
American Automobile Manufacturers
Association
7430 2nd Avenue, Suite 300
Detroit, MI 48202
Dr. Randy Roth
ARCO
P.O. Box 2679-TA
Los Angeles, CA 90051
Dr. Jane Warren
Health Effects Institute
141 Portland Street
Cambridge, MA 02139
IV
-------
TABLE OF CONTENTS
Parti Page
ORGANIZATION AND DESCRIPTION OF THE PROCEEDINGS .............. 1
SESSION ONE: INTRODUCTION/WELCOME 5
Chair: Judith A. Graham
SESSION TWO: EMISSIONS AND AIR QUALITY UNDER WINTER
CONDITIONS 21
Chair: Steven Cadle; Discussant: Robert Sawyer
SESSION THREE: HUMAN EXPOSURES .................... r 24
Chair: Charles Powers; Discussant: Steve Colome
SESSION FOUR: ACUTE HEALTH EFFECTS OF MTBE EXPOSURE ......... 27
Chair: Lawrence Reiter; Discussant: Roger O. McClellan
SESSION FIVE: CHRONIC HEALTH EFFECTS OF MTBE EXPOSURE 36
Chair: Robert Drew; Discussant: John Doull
SESSION SIX: NEW FINDINGS FOR OTHER FUEL OXYGENATES ......... 38
Chair: Randy Roth; Discussant: Bernard Goldstein
SESSION SEVEN: CONFERENCE SUMMARY . 43
Discussant: Gareth Green
Part II
Speakers' Abstracts and Presentation Materials
APPENDIX A: SESSION ONE A-l
William J. Piel (ARCO) - Overview of Fuel Oxygenate Development ....... A-2
APPENDIX B: SESSION TWO B-l
Kenneth T. Knapp (U.S. Environmental Protection Agency) -
Cold Temperature MTBE Dynamometer Study ..................... B-2
Chandra B. Prakash (Environmental Canada) - Emissions Under
Cold Conditions .B-22
-------
TABLE OF CONTENTS
Part II (cont'd) Page
David Veazey (University of Alaska at Fairbanks) with Marcus Martin,
Perry Klein, and Richard Benner - Carbon Monoxide Emissions in
Fairbanks, Alaska . . . B-34
Gerry Guay (Alaska DEC) - Air Quality Monitoring - Oxygenated Fuels B-48
Roy B. Zweidinger (U.S. Environmental Protection Agency) - Air Quality
Measurements in Fairbanks, Stamford, and Albany B-49
Larry G. Anderson (University of Colorado, Denver) with Pamela Wolfe
and John A. Lanning - The Effects of Oxygenated Fuels on Carbon Monoxide
and Aldehydes in Denver's Ambient Air B-67
APPENDIX C: SESSION THREE C-l
Ted Johnson (IT Corp) - Service Station Exposures C-2
Jack Hinton (Texaco) - American Petroleum Institute Occupational
Exposures - MTBE C-21
P.J. Lioy (Environmental and Occupational Health Sciences Institute) with
C. Weisel, E. Pellizzari, and J. Raymer - Volatile Organic Compounds from
Fuels Oxygenated with MTBE: Concentration and Microenvironmental
Exposures to MTBE in Automobile Cabins C-56
Alan H. Huber (U.S. Environmental Protection Agency) - Human Exposure
Estimates of Methyl Tertiary Butyl Ether (MTBE) C-87
APPENDIX D: SESSION FOUR D-l
John Middaugh (State of Alaska, DHHS) with Michael Beller - Potential
Illness Due to Exposure to Oxygenated Fuels - Fairbanks, Alaska . D-2
John Middaugh (State of Alaska, DHHS) with Bruce Chandler - Potential
Illness Due to Exposure to Oxygenated Fuels - Anchorage, Alaska D-7
Ronald Moolenaar (Centers for Disease Control and Prevention) - Methyl
Tertiary Butyl Ether (MTBE) in Human Blood After Exposure to Oxygenated
Fuels in Fairbanks, Alaska . D-12
VI
-------
TABLE OF CONTENTS
Part II (cont'd) Page
Mary C. White (Centers for Disease Control and Prevention) -
An Investigation of Exposure to Methyl Tertiary Butyl Ether Among
Motorists and Exposed Workers in Stamford, Connecticut ............. D-42
Nancy Fiedler (EOHSI) with Sandra Mohr and Kathie Kelly-McNeil -
Response of Sensitive Groups to Methyl Tertiary Butyl Ether (MTBE) D-65
Sandra N. Mohr (EOHSI) with Nancy Fiedler and Kathie Kelly-McNeil -
Health Effects Among New Jersey Garage Workers ................. D-85
Mary E. Gordian (Municipality of Anchorage, Alaska) with M.D. Huelsman,
M.L. Bracht, and D.G. Fisher - Using Insurance Claims Data to Investigate
Effects of Oxygenated Fuels on Community Health in Anchorage, Alaska . . . . D-128
Gerhard K. Raabe (Mobil) - American Petroleum Institute Health
Complaint Survey D-144
Richard Clark (Unocal) - Odor Threshold Studies of Oxygenates and
Oxygenate/Gasoline Blends ................................ D-160
Timothy R. Gerrity (U.S. Environmental Protection Agency) with
James Prah, Robert Devlin, George Goldstein, David Otto, David Ashley,
and Timothy Buckley - Acute Responses of Healthy Human Subjects to
MTBE Exposure D-172
William S. Cain (Yale University) with Brian Leaderer, Enrique
Cometta-Muniz, Jenneane F. Gent, Marion Buck, Larry G. Berglund,
Vahid Mohsenin, Edward Monahan, Gary L. Ginsberg, Larry S. Andrews,
and J. Soren Kjaergaard - Human Reactions to One-Hour Exposures to
Methyl Tertiary-Butyl Ether (MTBE) D-211
APPENDIX E: SESSION FIVE .................................. E-l
Larry S. Andrews (ARCO Chemical) - Chronic Health Effects of MTBE E-2
Jeffrey S. Gift (U.S. Environmental Protection Agency) - Derivation of the
Methyl Ten-Butyl Ether (MTBE) Inhalation Reference Concentration (RfC) . . . . E-31
Charlie Hiremath (U.S. Environmental Protection Agency) - The EPA
Cancer Assessment of MTBE E-43
vn
-------
TABLE OF CONTENTS
Part II (cont'd) Page
APPENDIX F: SESSION SIX ........ ..... . . . ...................... F-l
Michael Wells (AMOCO) - Health Effects ofETBE F-2
Russell D. White (Chevron) - Health and Ecosystem Effects of TAME F-3
Robert C. MacPhail (U.S. Environmental Protection Agency) - Health
Effects of Inhaled Ethanol F-22
Vlll
-------
ORGANIZATION AND DESCRIPTION
OF THE PROCEEDINGS
The "Conference on MTBE and Other Oxygenates: A Research Update" was held on
July 26 through 28, 1993, at Falls Church, Virginia under the sponsorship of the U.S.
Environmental Protection Agency (EPA), the American Petroleum Institute (API), and the
Oxygenated Fuels Association (OFA). A primary purpose of the conference was to provide
a forum for the presentation of results from studies that had recently been conducted on
exposure and health effects aspects of methyl tertiary butyl ether (MTBE) and MTBE.-
oxygenated fuels.
The conference comprised seven sessions: (1) Introduction/Welcome, (2) Emissions and
Air Quality under Winter Conditions, (3) Human Exposures, (4) Acute Health _Effects of
MTBE, (5) Chronic Health Effects of MTBE, (6) New Findings for Other Fuel Oxygenates,
and (7) Conference Summary. A chairperson introduced the topic of each session, and after
the speakers presented their studies and entertained questions from the audience, a discussant
summarized the findings for that session. Following the conference, the speakers were asked
to provide abstracts of their presentations and copies of their slides. In addition, the
discussants submitted written summaries for their respective parts of the program, including
the overall summary of the conference. The discussants' summaries were based on the
information available to them at that time and represent solely their opinions about that
information.
These proceedings consist of two parts: Part I contains presentations in the
Introduction/Welcome session and the discussant summaries for each of the remaining
sessions; Part II contains the abstracts and slides of the speakers. These proceedings of the
conference are not a verbatim transcript of the meeting. Speakers and discussants were given
an opportunity within the few weeks following the conference to revise their abstracts and
slides. Also, oral statements have been edited to reduce the number of verbalisms (although
much of the introductory section still has the character of an oral transcript rather than a
statement that was intended for publication). In addition, the discussion at the end of each
session has been summarized, rather than attempting to identify individual speakers'
questions and answers. Otherwise, these proceedings present the information as it was
presented in July 1993 or shortly thereafter.
Because many of the studies presented at the conference had only just been completed
or were still in the stage of data analysis and interpretation, the material in these proceedings
does not reflect analyses that may have been conducted subsequent to the conference.
Reports of some of the studies presented at the conference have since been published. These
and some related papers are listed below.
-------
Recent Oxyfuel-Related Publications
American Petroleum Institute (1993) Gasoline vapor exposure assessment at service stations.
Washington, DC: Health & Environmental Sciences Department; March; API Publication
No. 4553.
American Petroleum Institute (1994) Odor threshold studies performed with gasoline and
gasoline combined with MTBE, ETBE and TAME. Washington, DC: Health &
Environmental Sciences Department; January; API Publication No. 4592.
American Petroleum Institute (1994) A study to characterize air concentrations of MTBE at
representative service stations in the Northeast. Washington, DC: Health & Environmental
Sciences Department; April; API Publication No. 4619.
American Petroleum Institute (1995) Petroleum industry data characterizing occupational
exposures to MTBE. Washington, DC: Health & Environmental Sciences Department;
August; API Publication No. 4622 (in preparation).
American Petroleum Institute (1995) Service station personnel exposures to oxygenated fuel
components. Washington, DC: Health & Environmental Sciences Department; August; API
Publication No. 4625 (in preparation).
American Petroleum Institute (1995) Anecdotal health-related complaint data pertaining to
possible exposures to methyl tertiary butyl ether (MTBE): 1993 and 1994 follow-up surveys
(1984-1994). Washington, DC: Health & Environmental Sciences Department; August; API
Publication No. 4623 (in preparation).
Anderson, H. A.; Hanrahan, L.; Goldring, J.; Delaney, B. (1995) An investigation of health
concerns attributed to reformulated gasoline use in southeastern Wisconsin: final report.
Madison, WI: Wisconsin Department of Health and Social Services, Bureau of Public
Health; May 30.
Belpoggi, F.; Soffritti, M.; Maltoni, C. (1995) Methyl-tertiary-butyl ether (MTBE)—
a gasoline additive—causes testicular and lymphohaematopoeitic cancers in rats. Toxicol. Ind.
Health 11: 119-149.
Bonin, M. A.; Ashley, D. L.; Cardinal!, F. L.; McCraw, J. M.; Wootin, J. V. (1995)
Measurement of methyl tert-butyl ether and tert-butyl alcohol in human blood and urine by
purge-and-trap gas chromatography-mass spectrometry using an isotope-dilution method.
J. Anal. Toxicol. 19 (May/June): in press.
Cain, W. S.; Leaderer, B. P.; Ginsberg, G. L.; Andrews, L. S.; Cometto-Muniz, J. E.;
Gent, J. F.; Buck, M.; Berglund, L. G.; Mohsenin, V.; Monahan, E.; Kjaergaard, S. (1996)
Acute exposure to low-level methyl tertiary-butyl ether (MTBE): human reactions and
pharmacokinetic response. Inhalation Toxicol. 8 (January): in press.
-------
Cirvello, J. D.; Radovsky, A.; Heath, J. E.; Farnell, D. R.; Lindamood, C., III. (1995)
Toxicity and carcinogenicity of t-butyl alcohol in rats and mice following chronic exposure in
drinking water. Toxicol. Ind. Health 11: 151-165.
Daughtrey, W. C.; Bird, M. G. (1995) Genotoxicity and twenty-eight-day subchronic
toxicity studies on tertiary amyl methyl ether. J. Appl. Toxicol. 15: 313-319.
Duffy, L. K. (1994) Oxyfuel in Alaska: use of interleukins to monitor effects on the immune
system. Sci. Total Environ. 151:253-256.
Fiedler, N.; Mohr, S. N.; Kelly-McNeil, K.; Kipen, H. M. (1994) Response of sensitive
groups to MTBE. Inhalation Toxicol. 6: 539-552.
Hartle, R. (1993) Exposure to methy tert-butyl ether and benzene among service station
attendants and operators. Env. Health Pers. Suppl. 101 (Suppl. 6): 23-26.
Lioy, P. J.; Weisel, C. P.; Jo, W-K; Pellizzari, E.; Raymer, J. H. (1994)
Microenvironmental and personal measurements of methyl-tertiary butyl ether (MTBE)
associated with automobile use activities. /. Exp Anal. Env. Epid. 4: 427-441.
Maltoni, C.; Soffritti, M. (1995) Editorial: gasoline as an oncological problem. Toxicol.
Ind. Health 11: 115-117.
Mannino, D. M.; Schreiber, J.; Aldous, K.; Ashley, D.; Moolenaar, R.; Almaguer, D.
(1994) Human exposure to volatile organic compounds: a comparison of organic vapor
monitoring badge levels with blood levels. Int. Arch. Occup. Environ. Health 67: 59-64.
Mohr, S. N.; Fiedler, N.; Weisel, C.; Kelly-McNeil, K. (1994) Health effects of MTBE
among New Jersey garage workers. Inhalation Toxicol. 6: 553-562.
Moolenaar, R. L.; Hefflin, B. J.; Ashley, D. L.; Middaugh, J. P.; Etzel, R. A. (1994)
Methyl tertiary butyl ether in human blood after exposure to oxygenated fuel in Fairbanks,
Alaska. Arch. Environ. Health 49: 402-409.
Nihlen, A.; Walinder, R.; Lof, A.; Johanson, G. (1994) Toxicokinetics and irritative effects
of methyl tertiary-butyl ether in man. Presented at: International Society for Environmental
Epidemiology/International Society for Exposure Analysis joint conference; September;
Research Triangle Park, NC.
Prah, J. D.; Goldstein, G. M.; Devlin, R.; Otto, D.; Ashley, D.; House, D.; Cohen, K. L.;
Gerrity, T. (1994) Sensory, symptomatic, inflammatory, and ocular responses to and the
metabolism of methyl tertiary butyl ether in a controlled human exposure experiment.
Inhalation Toxicol. 6: 521-538.
Rudo, K. M. (1995) Methyl tertiary butyl ether (MTBE)—evaluation of MTBE
carcinogenicity studies. Toxicol. Ind. Health 11: 167-173.
-------
Rudo, K. M.; Pate, W. J.; Smith, C. G. (1995) The public health impact of the oxygenated
fuels program in North Carolina. Toxicologist 15: 140.
Smith, S. L.; Vetrano, K.; Duffy, L. K. (1994) Report to the Oxygenated Fuels Association
on the effect of cold temperatures on odor thresholds of Alaska and Lower-48 gasoline
oxygenated with 15% MTBE and 15% IPA. Institute of Arctic Biology; grant no. 258740.
Tepper, J. S.; Jackson, M. C.; McGee, J. K.; Costa, D. L.; Graham, J. A. (1994)
Estimation of respiratory irritancy from inhaled methyl tertiary butyl ether in mice.
Inhalation ToxicoL 6: 563-569.
U.S. Environmental Protection Agency (1993) Assessment of potential health risks of
gasoline oxygenated with methyl tertiary butyl ether (MTBE). Washington, DC: Office of
Research and Development; November; EPA report no. EPA/600/R-93/206.
U.S. Environmental Protection Agency (1994) Health risk perspectives on fuel oxygenates.
Washington, DC: Office of Research and Development; November; EPA report no.
EPA/600/R-94/217.
White, M. C.; Johnson, C. A.; Ashley, D. L.; Buchta. T. M.; Pelletier, D. J. (1995)
Exposure to methyl tertiary butyl ether from oxygenated gasoline in Stamford, Connecticut.
Arch. Environ. Health 50: 183-189.
-------
Session One
INTRODUCTION/WELCOME
Chair: Judith A. Graham, Office of Research and Development,
U.S. Environmental Protection Agency
Welcome
Peter Preuss
Director of the Office of Science, Planning, and Regulatory Evaluation,
Office of Research and Development, U.S. Environmental Protection Agency
I am very pleased to be here and to welcome all of you to this workshop on behalf of
EPA and the Office of Research and Development (ORD). There are a good number of
people in the audience here today that I haven't seen in quite awhile. And so I welcome all
my friends and colleagues to this meeting.
Basically what we're trying to do with this meeting is very simple. We want to hear
what the new scientific evidence on MTBE and MTBE oxyfuels is about. The reason that
we're doing this is, in large part, because over the past few months a number of questions
have arisen about this topic and we felt it was important both to do research to try to answer
some of the questions and to review the information that was available on this topic. So we
have a very clear purpose that we've set out for ourselves and that is to look at the science,
to try to understand the science, to come to grips with the difficult questions in the science,
and to see what it is that we know and determine what it is that we don't know. One of the
things that we have to do yet at EPA is take that science and finish our analysis of it and
then give it to our program office, the Office of Mobile Sources (OMS) in the Office of Air
and Radiation, so that the Agency can make a decision about MTBE and oxyfuels and
whatever needs to be done. This is a very critical meeting for us, and I hope all of you will
participate and help us to try to clarify what it is that we know, what it is that we don't
know, and where the areas of major uncertainties lie, so that any assessment that we do can
be as clear as possible.
I think it's very important to both understand and acknowledge what has happened over
the past year or so since this issue first was raised. Essentially we've had a group of people,
both in EPA and outside of EPA, with sponsorship from API, OFA, and a number of other
places, and we have had a number of states involved, coming together trying to understand
what it is that people were saying and the kinds of questions that people were asking, and
trying to see what kind of scientific data were needed in order to answer some of these
questions. One of the things that I'd like to acknowledge here is the very impressive
turnaround, the very impressive reaction that all of these groups gave to this, the attention
that they gave to this, so that we could in fact, in a very timely way, come together at a
workshop like this and discuss it. Much of what we'll be discussing at this workshop really
did not exist a year ago.
-------
And so I'd particularly like to make mention of the various EPA laboratories, API, and
OFA, who have not only helped us to fashion and put together some of this work, but also
have helped us to put together this workshop.
Finally, I'd like to just reemphasize this effort of cooperation. I think the EPA, in
particular in the Office of Research and Development, is moving in a somewhat different
direction from where we have been in the past. I think we're going to wind up doing a lot
more work together with other entities outside of the federal government. Whether it's
industry or whether it's state government or whether it's associations or what-have-you, we
are trying to come together hi the areas where we have common interests, where we have
common concerns and where we have common questions; and we are trying to jointly
understand what kind of research needs to be done and trying to jointly come together to see
that the research is actually done. This is one example; there are other examples in the area
of alternative fuels, of tropospheric ozone research—there are lots of areas where EPA is
trying to begin to work in this new fashion with groups outside of the federal government.
And so as we go through this, I hope you keep that in the back of your mind and see how
we can try to develop this kind of an idea and this kind of shared and cooperative research.
So welcome all of you and thank you very much for being here.
-------
Legislative History and Regulatory Overview of Oxygenates and MTBE
Mary Smith
Director of the Field Operations and Support Division
Office of Mobile Sources, U.S. Environmental Protection Agency
After EPA implemented a nationwide oxygenated gasoline program in November 1992,
concerns about possible health issues related to MTBE became evident. I will provide some
background on the oxygenate program and why we are here to discuss health effects and
exposure research on MTBE and other oxygenates. The principal oxygenates used in-the
program are MTBE and ethanol. Future oxygenates—which we will address on the last day
of the conference—include tertiary amyl methyl ether (TAME) and ethyl tertiary butyl ether
(ETBE).
In terms of background on oxygenates and mobile source emissions, carbon monoxide
(CO) is an odorless, colorless gas, which impairs cardiovascular functions and has particular
health effects on infants and unhealthy individuals. Carbon monoxide emissions in motor
vehicles represent a large proportion of all CO emissions, about 75%—more so in those
cities where there's heavy vehicle traffic and lack of industry. The Washington, DC, area is
one of those areas where we have a heavy contribution of motor vehicle CO emissions.
Carbon monoxide nonattainment occurs principally in the winter. It's estimated, based
principally on lab testing at warmer temperatures, that mobile source emissions of CO from
motor vehicles can be reduced by 15 to 20%. Oxygenates can contribute to CO emission
reductions because they allow the vehicle combustion process, which is somewhat inefficient
in the winter, to be more efficient.
Why do we have an oxygenated gasoline program? When Congress passed the Clean
Air Act Amendments in 1990, they had a specific provision in Section 211(m) that required
oxygenated gasoline to be used in various areas of the country. Basically, if an area did not
attain the CO National Ambient Air Quality Standards, based on 1988 and 1989 data, the
area was required to implement an oxygenated gasoline program by November 1 of 1992.
Thirty-nine areas met this criterion. Other areas would necessarily be in the program if
subsequent data showed that they had a CO nonattainment problem, and this year, in 1993,
Salt Lake City meets mat criterion. Prior to the implementation of the program in about
39 areas in November 1992, there were seven existing programs operating in the country,
including Denver, which started the first oxygenated fuels program hi 1988.
Oxygenated gasoline is seen by the states as the major CO reduction strategy in the
Clean Air Act amendments. The specific requirement is that gasoline must contain
oxygenates to achieve.2.7% oxygen by weight hi the gasoline. This is achieved with about
15% by volume MTBE and 7.7% by volume ethanol.
The oxygenate program is not a year-round program because CO is mainly a wintertime
problem. Because of differences in climate and the degree of an area's CO problem, the
oxygenate program in the 39 areas runs anywhere from 4 to 7 mo each year. Thus,
-------
programs start anywhere from September to November, with the majority running a 4-mo
program from November to February.
• - - -••.•-•.•-.
Oxygenated gasoline was not intended just to be sold only in the CO nonattainment
area, which can be small, but must be sold in the entire metropolitan statistical area
surrounding the nonattainment area. The rationale was that cars travel in and out of CO
nonattainment areas, and it is necessary to encompass all those cars that are contributing to
the CO problem in the area. This is a state-enforced program, not a federally enforced
program.
Under the Clean Air Act, a state may ask EPA to waive the program based on certain
factors, but none of these are relevant to the discussion today. The statute does provide that
if an area comes into CO attainment and can demonstrate CO attainment without the program
by means of a 10-year maintenance plan, then the area basically does not have-to use
oxygenated gasoline. There are a few areas applying for such attainment demonstrations
without the oxyfuel program. Cleveland is one of them, for example.
On the third day of the conference, other oxygenates will be discussed that are relevant
to another section of the Clean Air Act that mandates oxygenates in reformulated gasoline.
The reformulated gasoline program starts in 1995 in at least the nine worst ozone
nonattainment areas. Other areas that are in nonattainment with ozone may opt into the -
program. Currently, the opt-in areas are primarily in the northeast where they have a
particular problem with ozone nonattainment and where there's a large region of ozone
transport. So oxygenates are used in more than one program, and I suspect, in 1995, more
of the other oxygenates are going to be used.
We do have a sense of how well the oxyfuels programs performed this past winter.
Thirty-six of the programs started on time, although, as I mentioned before, all of the
programs do not start at the same time because the control period depends on when CO is
elevated. In this past season, 70% of the gasoline had MTBE, 30% had ethanol. The
gasoline in these oxygenated gasoline program areas constitutes about a third of the gasoline
used in the country. California areas hi the program adopted a 2% oxygen requirement, and
their petition with regard to this reduced percentage is pending.
Carbon monoxide data for November 1992 to January 1993 have been evaluated.
We're still analyzing the February data. The non-California areas were evaluated because
these are the areas that had 2.7% oxygen. We did not include data from the older programs
because obviously you don't get a good picture of how well oxygenated gasoline has worked
when there has been a preexisting program. The evaluation showed that there was about a
95% reduction in the number of CO exceedances. California alone had about an 80%
reduction, and even the preexisting programs had about a 50% reduction.
Now for why we're here. As Bill Piel from ARCO Chemical will discuss shortly,
oxygenates have been in commerce for a long time, since the early 1980s and even possibly
late 1979. And so I don't think anybody anticipated any problems with the oxygenates. But
in November of 1992, as Alaska was implementing its program in both Fairbanks and
Anchorage, there were a significant number of health complaints registered. The state asked
8
-------
the Centers for Disease Control (CDC) to investigate the health complaints, which included
headaches, nausea, and coughing. The CDC did a study in the November and December
time frame and issued a report to Alaska in early 1993, which said that there were some
preliminary indications of a correlation between the acute health complaints and the
oxygenated gasoline program. I should remind you that 100% of the gasoline hi Alaska
contains MTBE. The CDC is represented at this meeting and later on will present all their
findings, so I defer to them in terms of a more detailed description. There was a lot of press
coverage on health concerns. As a result of all of this, Governor Hickel suspended the
program. On December 11, we were made aware of that suspension.
Concerns also were raised in other areas, particularly Missoula, MT, where the city
convened a health panel. They decided to continue the program to the end of February, but
obviously they are very anxious to hear what comes out of this workshop and what
conclusions are reached about these health issues. Concerns also were raised later in
Anchorage and New Jersey, and the people in these areas also are anxiously awaiting the
results of this workshop.
Although unexpected, there was obviously an issue here that needed to have some
resolution. So EPA, CDC, and some members of the industry met in January of 1993 to
review what we knew about the acute and chronic effects from MTBE and to map out a
strategy for research that we felt needed to be completed in 6 mo. Time was limited
because, although most of the programs stop for the season at the end of February, they are
nonetheless due to resume as early as September 1993. We therefore felt that we had about
6 mo to look at the issue and see what other data we could collect.to address a lot of the
issues that had arisen, particularly the acute health complaints.
This brings us to the workshop. I would like to reiterate Peter Preuss' compliments
(i.e., I think that people worked very well hi the 6-mo period on an aggressive research
program). People worked hard within EPA, CDC, industry, and other groups to collect
data, and to quickly analyze it so that we could target it for presentation at this workshop, hi
the hope that we could try to come to some sort of resolution and conclusions with regard to
MTBE before the next oxyfuel-season started hi this country. I must compliment everybody's
effort in this regard.
-------
Overview of the Health-Related Issues and Framework for the Research
Judith Graham
Associate Director of the Environmental Criteria and Assessment Office, Office of
Research and Development, U.S. Environmental Protection Agency
I would like to expand on the history of the research program to which Mary Smith
referred. Discussions on MTBE research actually began in earnest in EPA around 1986 or
1987, resulting early in 1988 in a consent decree between EPA and industry according to the
provisions of the Toxic Substances Control Act. The MTBE Task Force, a consortium of a
number of industry groups, was created to conduct a large number of research projects
investigating acute, subchronic, and chronic inhalation effects of MTBE hi animals. Those
studies included reproductive and developmental effects.
In 1991, EPA took the subchronic studies that had been completed by the MTBE Task
Force and developed a reference concentration, or RfC. An RfC is defined as a
concentration (with an uncertainty about it) of a chemical that can be inhaled over a lifetime
by even sensitive people, and is thought not to pose an appreciable deleterious
noncancer—noncancer, I want to emphasize that—health hazard. The 1991 RfC for MTBE
was 0.5 /xg/m3.
In late 1992, the chronic inhalation studies from the MTBE Task Force became
available and the oxyfuel program began. Very soon after the program began, numerous
people started complaining of health symptoms in some areas. Especially noteworthy were
the complaints coming from Fairbanks and Anchorage, AK, and from Missoula, MT. There
were a number of more isolated complaints coming from citizens in New Jersey and some
other areas. There were still other areas of the country using oxyfuels, such as Denver,
where a significant number of complaints were not made.
Due to complaints in Alaska, the state requested that the CDC conduct an evaluation.
The CDC, working with the state, initiated some epidemiology studies in Fairbanks while
MTBE was still being used. As part of these studies and as part of an EPA interest in
obtaining even more information, there was a collaborative effort to measure MTBE air
levels as well as other compounds hi the air, hi some sites hi Fan-banks. At the same tune,
the Environmental and Occupational Health Sciences Institute in New Jersey conducted some
pilot studies of exposure levels inside vehicles, using commuting scenarios. In some of the
scenarios, people stopped and refueled their cars with gasoline, whereas in other scenarios
they didn't.
The Office of Research and Development within EPA also conducted initial studies of
emissions from vehicles down to temperatures as low as 0 °F. In February 1993, ORD
issued a report that evaluated the public health issues and indicated numerous uncertainties
and gaps hi the information available. I think it's important to spend a few moments here on
the conclusions of ORD's February report, so that everybody can have a sense of where we
were when the research started and so that we can think clearly about where we are when the
10
-------
research is completed. It also gives you an idea of the genesis of some of the research
projects that were undertaken.
We had concluded that the use of MTBE oxyfuels can decrease CO emissions from
vehicles at temperatures as low as 0 °F, and perhaps even lower. But we really didn't go
lower than that, leaving the emission effects at lower temperatures undefined. One of the
main concerns was that in the state of Alaska it can get down to —30 °F or even —50 °F.
From limited emission studies, it appeared that MTBE oxyfuels would reduce the total
mass (the net emissions) of 1,3-butadiene, benzene, and formaldehyde. The data showed that
concentrations of formaldehyde increased and concentrations of butadiene and benzene
decreased. But the net overall effect was a decrease. We don't know what this decrease
implies for health. If there's a decrease in total cancer risk—and that's a big if—that
decrease would likely be very small.
At the end of February when we issued this report, we really couldn't draw any
definitive conclusions about the potential for MTBE oxyfuels to cause acute health
symptoms. But the initial reports that came out from CDC and the anecdotal reports that
were received from various areas of the country were suggestive of a problem, although they
were not definitive.
When we evaluated the RfC for MTBE and compared it to very—emphasis on the word
"very"—limited preliminary exposure data, it didn't appear at that time that MTBE was
going to pose a significant risk of chronic noncancer health effects.
As we looked at the results of chronic inhalation toxicity studies in February, we noted
that tumors were observed in some of the animals at high MTBE concentrations. But we
decided that an assessment was needed and that we had to evaluate the data in depth before
reaching conclusions. It also became apparent that although there were numerous studies of
MTBE itself, we couldn't find any controlled studies of MTBE oxyfuels, and that precluded
drawing conclusions about the effects of the mixture. Our bottom line was that research was
needed.
When we talked about research needs, we did think in terms of a major goal:
to conduct research that would provide information by midsummer. We wanted to get
information before the next MTBE oxyfuel season began, and we wanted the data to increase
our understanding of the potential for health risks. We wanted to be able to develop risk
assessments, insofar as possible, with the new data. And we also wanted to identify any
residual uncertainties to determine whether it was advisable to think about more of a long-
term research program.
The primary groups that have funded this research are API, OFA, and various
organizations within EPA. (If I've left anybody out, I really do offer my sincere apologies.)
There was widespread collaboration and interaction among the groups.
I have dealt with research for a long time, and as Peter Preuss mentioned earlier, this
program really is remarkable in that we had states, several federal organizations, several
11
-------
groups within EPA, industry groups, and several academic institutions, who were all working
toward the same goal. And although communication can always be improved, I think it was
a good example of a lot of different minds coming together to work on a common problem.
The risk assessment paradigm developed by the National Academy of Sciences several
years ago describes how research feeds into risk assessment and how risk assessment feeds
into risk management. Briefly, research would consist of laboratory and field studies, not
only of health effects, but some preliminary exposure data as well. It then becomes
important to identify or determine whether there is there a hazard; that is, does the agent
cause an effect?
But then we need to know at what concentration does an agent cause a problem? What
exposures actually cause effects, and what are those concentrations in humans? Often
laboratory studies have to be extrapolated from high to low dose or from animal to human.
Once we know that a certain exposure causes a certain effect, another significant
question follows—does that exposure actually occur? And if it occurs, is it one person that
gets such an exposure or 50 million people that get such an exposure? That's the importance
of exposure assessment, which needs to be supported by field measurements and
characterization of populations.
All of these pieces of information are integrated together and judgments are made about
risk characterization, which is the estimated incidence of the effect in a given population.
It then becomes the task of risk managers and policy makers to decide how to take that
information and develop regulatory options. Regulatory options consider a number of issues
including health issues, social issues, and the different impacts of the regulatory options.
From these, the Agency makes decisions. So these processes, the risk assessment process
and the risk management process, are different.
This brings us to the research that will be presented here at the conference. I'm going
to give you a brief overview of what people are going to present without giving you any
information on the results of their studies.
The Environmental and Occupational Health Sciences Institute (EOHSI) has been
conducting a lot of work on exposure hi commuting scenarios, including monitoring inside
vehicles. The International Technology (IT) Corporation has been performing some
extensive work characterizing MTBE and other pollutant levels at several gas stations. The
EPA/ORD has looked at vehicular emissions at subzero temperatures. Those emission
studies also have been conducted in collaboration with Environment Canada, using their
facilities as well to look at the effects of low temperatures. And both the ORE) and OMS
worked with Alaska and New York, as well as several of the EPA regions, to analyze the
ambient air samples.
For the health studies, we took a combination approach of epidemiology, human
clinical, and animal toxicology studies; each one of these approaches has its own strengths
and its own limitations. In the interest of time, I won't go into them here today. Together,
12
-------
these approaches are synergistic. When we look at the results of all of these approaches as a
whole, it minimizes the inherent weaknesses of each and it maximizes the strengths of each.
In terms of epidemiology, CDC has done a number of studies. As you recall,
originally they were called in by the state of Alaska to look at Fairbanks while MTBE was
being used. They went back after MTBE had been removed and looked at symptom
complaints; they measured blood levels of MTBE and other compounds and conducted some
other associated studies. They later used similar approaches to compare Stamford, CT,
which uses MTBE, to Albany, NY, which does not use MTBE.
The EOHSI, as part of a broader program, also did some epidemiological evaluations.
They had a panel of people who were already known to be chemically sensitive, so they were
able to investigate symptoms in relation to gasoline exposures in that panel. Because
different areas of New Jersey either had or didn't have MTBE at a particular period of time,
they also were able to compare symptom prevalences within New Jersey.
The API surveyed industry to see what kind of symptom complaints were arising with
workers who were more closely and frequently involved with fuels. Also, the Municipality
of Anchorage looked at health insurance records to see if there was any kind of an increase
hi claims that possibly could be associated with MTBE use.
Two human clinical studies were conducted, one at the EPA Health Effects Research
Laboratory (HERL) at Research Triangle Park and the other at Yale University. In both
studies, volunteers were exposed acutely to MTBE and evaluated for self-reported symptoms,
nasal and eye irritation, and neurobehavioral changes. Previously, API had odor threshold
studies performed for pure MTBE, but further studies were conducted to include MTBE-
blended gasolines.
As I mentioned, the MTBE Task Force sponsored a whole series of animal studies.
There were some uncertainties associated with effects on rat kidneys, and so they did further
evaluation of tissue samples and conducted some new studies to help clarify some of the
effects on kidneys.
In the realm of assessment, the EPA/ORD developed a tentative or draft cancer
classification. When EPA develops a cancer assessment, a particular group of
scientists—in this case, scientists from the Office of Health and Environmental
Assessment—develop a draft that goes to an Agency-wide work group for verification. So at
this point, the cancer assessment is still in that early draft stage; it has not gone to the
EPA-wide verification group.
The ORD also revised the RfC for MTBE. If you recall, the first RfC was based on
subchronic studies. We were able to take advantage of the new chronic information and
derive a new RfC. This is an official reevaluation; the EPA work group just recently met
arid verified the new RfC.
13
-------
The ORD also has developed an exposure assessment, taking advantage of the new data
from EOHSI, IT Corporation, and several other groups. Shortly, we will be developing a
new report that evaluates the potential health risks of MTBE oxyfuels.
We need to keep our expectations reasonable. Looking at the nature of the research
program, without knowing what the results are at this point, we're hoping the new as well as
the older data will enable us to determine what are the likely exposures to MTBE and
oxyfuels and whether acute controlled exposures to MTBE can cause health symptoms, affect
neurobehavioral functions, or cause eye or nose irritation. We also hope to be able to
understand whether acute health symptoms are associated with acute ambient exposures to the
whole of MTBE oxyfuels. Hopefully, we're also going to have some estimate of the
likelihood of MTBE causing cancer, something we didn't have for the February report.
In addition, we should have more information on the effects of MTBE on CO tailpipe
emissions at lower temperatures.
To make sure we have some balance here and keep our expectations reasonable, I'd
like to give some counterpoint, that is, some questions that we think will remain. Again,
I have to remind you that we have not seen the data, so these are our opinions of what will
remain to be understood. If indeed health symptoms are found to be associated with acute
exposures, what are the quantitative risks and what are the associated uncertainties? It's
necessary to know what the key risk factors might be. For most pollutants there are
sensitive subpopulations, and it's important to know sensitive subgroups exist so they can be
enumerated and their risks evaluated.
I think we can presume that there will be a lot of uncertainty about the cancer
classification. So, what sorts of follow-up studies will be needed to decrease the
uncertainties? If we do have concerns for chronic noncancer risks—and again, I don't know
whether we will or we won't because we haven't seen the data yet—but if we do have
concern for chronic noncancer risk, then what follow-up studies would be most appropriate
and necessary?
Finally, the comparative risk is a very important question. Even if we knew everything
we ever wanted to know about MTBE, it would still be important to know what are the risks
that might result from exposure to other oxygenates, such as ethanol and ETBE.
Discussion
It was noted that animal testing of MTBE actually began as early as 1969 and that data
from a 28-day study of MTBE in blended gasoline as well as reproductive subchronic studies
and oncogenicity studies of MTBE alone were available prior to the beginning of the
oxyfuels program in 1992.
In response to a question regarding whether EPA/ORD assessments were being done or
considered for other oxygenates, the answer was no, because not enough information is
available to support such assessments at present.
14
-------
Determining the balance between the health benefits of reduced CO and the potential
health risks of oxygenated fuels themselves was suggested as another objective for the
conference or for future research. This point raised the issue of performing a relative risk
comparison of gasoline versus MTBE-oxygenated gasoline. It was noted that orders of
magnitude more information exists for CO health effects than for MTBE or MTBE oxyfuels.
15
-------
Historical Perspective on the Use of Ethers in Fuels1
William J. Piel
Manager of Business Development of Oxygenated Fuels, ARCO Chemical Company and
Head of the Technical Committee for the Oxygenated Fuels Association
ARCO Chemical has been involved in commercial production of fuel oxygenates since
1969, and so we'd like to share some of that experience of how we got to where we are
today. Here's what I'm going to try to cover. I'll discuss how fuel alcohols were developed
as low-cost gasoline extenders for more expensive crude oil in the days when we went
through the oil embargoes. Then I'll address how ethers were developed as lead-octane
replacements. I'll show worldwide MTBE production growth for the past two decades, and
talk about some of the pre-Clean Air Act oxygenated fuel programs that were conducted by
some of the states. I will then tell you a little about MTBE use in European gasolines and
talk about some of the vehicle emission benefits we've learned from reformulated gasoline
studies.
Fuel oxygenate development was started to develop an alternative to gasoline derived
from high-cost petroleum, particularly during the oil embargoes of 1973 and 1979, though it
in fact did start slightly before that. As I mentioned, ARCO Chemical has been involved in
this since 1969, when we were producing tertiary butyl alcohol (TEA) as a gasoline blending
component, both for gasoline as well as for some octane. We callthat GTBA, for
gasoline-grade TEA. Then we saw where the U.S. Government subsidized, and EPA
granted a waiver for, 10% bio-ethanol in gasoline in 1978.
Soon after that, we saw methanol and co-solvent blends commercialized in 1981. The
incentive there was methanol, which was derived from natural gas at roughly $2 per million
BTU. Crude then was equivalent to about $6 per million BTU. So there was an economic
incentive to look for a lower cost gasoline blending component such as methanol. In that
case, GTBA and other alcohols were used as co-solvents to help blend the methanol into
gasoline.
But then soon thereafter, crude collapsed, down to about $15 a barrel, and with that
was a loss of incentive for a lot of the alcohol blending. So, many of the alcohol blends
were discontinued in 1986. But at the same time, MTBE was being expanded and developed
as an octane alternative to lead in gasoline, and that was really promoted when EPA decided
to accelerate the phaseout of lead compounds in gasoline in the mid-1980s. It also was
helped by the growth in premium gasoline in that same time period.
I'll talk a little about what MTBE is. As I mentioned, it was developed as an octane
enhancer; it was a high-octane alternative to lead compounds or aromatics in gasoline.
It was first commercially used in Europe in 1973 where it was first developed in Italy and
then shortly thereafter was developed, produced, and used in Germany for gasoline blending.
Copies of the slides for this presentation may be found in Appendix A.
16
-------
The U.S. production began in 1979 after the EPA provided approval through a waiver
process. Originally it was allowed at seven volume percent.in gasoline. That limit was
raised to 11% in 1981 under the "substantially similar" ruling, and then was raised to 15% in
1988 after going through a major waiver review sought by Sun Oil Company.
The European Economic Community set a standard of 15% MTBE in 1985, before the
United States did. In Europe, MTBE also was used as a replacement for lead and aromatic
compounds used in gasoline. In actuality, MTBE is generally used throughout the world for
octane in gasoline, both in the Far East as well as Europe and here in the United States.
Recent interest in MTBE has been more from an environmental viewpoint, because of
its use to modify gasoline composition to reduce emissions associated with vehicles. The
largest benefit is reducing CO emissions and that's due to the oxygenate effect or improved
combustion processes, as mentioned already by Mary Smith. But it also helps-reduce other
mobile source pollutants such as volatile organic compounds (VOCs), nitrogen oxides (NOX),
paniculate matter < 10 /im, and toxics such as benzene. Because the oxygenates do not
include sulfur, generally when you use the oxygenate, it helps dilute the gasoline's sulfur
content from crude; so generally it also contributes to reduction in sulfur oxides (SOX) as
well as all the other things I just mentioned.
I'll talk a little bit about the production of MTBE and other ethers. All the ethers I'm
going to address were actually considered as early as the 1930s, when Shell Oil Company
was looking at them. The production technology that's in place today was essentially
developed in the late 1950s and patented around 1960, and it's essentially the same reaction
process and reactor configuration that have been used since then.
The reaction is fairly simple; it's made by combining isobutylene (from a number of
sources) and methanol. The isobutylene has a very active tertiary carbon for adding hydroxyl
groups such as methanol. The reaction is highly selective, with very minimal by-products.
The few that are produced are preliminary products such as di-isobutylene, tri-isobutylene
and TBA, which is produced from the small amount of water that gets in the feed stocks and
combines with the isobutylene to make TBA. All these contaminants have generally been
found in gasoline for a long time.
At present, the three main sources of isobutylene, which is usually the commercially
limiting factor for producing MTBE, are a by-product of the refinery fluid catalytic cracking
process for making gasoline. Isobutylene also is produced by isobutane dehydrogenation
(where the hydrogen is removed to make the isobutylene) and dehydration of TBA, which is
made from the isobutane oxidation. Recently there have been some commercial processes
for making isobutylene called skeletal isomerization, where normal chain olefins are
reconfigured or reshaped to make the preferred tertiary carbon. We can probably expect to
see a few of the latter production processes in the future.
Ethyl tertiary butyl ether is an alternative to MTBE. It's made by combining bio-
ethanol and isobutylene. It has been in a semi-commercial state for the past two years, and
has been made both in Europe and in the United States hi existing MTBE processes. It's
fairly simple to switch over from MTBE to ETBE; the reaction kinetics are similar as well as
17
-------
the process operation. So a few refiners and others have switched from MTBE to ETBE for
trial production purposes.
Tertiary amyl methyl ether is a commercial ether that has been produced in refineries
for a number of years. In Europe it has been in production since the mid-1980s. It's made
by combining refinery isoamylene with methanol. Isoamylene is another active olefin that
has tertiary carbon for combining alcohols or hydroxyl groups. Isoamylene is a very volatile
product that's made by the fluid catalyst cracking process as well and normally goes directly
into gasoline blending.
These latter two ethers, ETBE and TAME, generally have much lower volatility than
MTBE. That's one of their major advantages and why they're being considered and
developed today. Methyl tertiary butyl ether generally has a vapor pressure of about 8 psi at
100 °F, whereas ETBE is about around 4 psi, and TAME is around 2 psi. They're
considerably less volatile than MTBE.
I show a simple diagram of the process flow unit (see Appendix A). For refinery
MTBE production, the isobutylene is usually found in the mixture with other butylenes and
butanes coming from the fluid catalyst cracker, but the isobutylene selectively reacts with
methanol using an acid type catalyst in very mild reactor conditions. So the mixed four-
carbon chain compounds (C4s) from the fluid cat cracker are fed to the reactor along with
methanol. In the reaction, the methanol almost completely reacts with isobutylene with a
high conversion, 96 to 97% or higher. The unreactive C4s will then vaporize off the
separation tower where the higher boiling MTBE is then removed off the bottom of the
fractionating tower. In the refinery, this would go directly to gasoline blending.
A lot of MTBE capacity has been developed based on butane plants for commercial
sales, generally to refiners. In the first step, the isobutane—and this could be directly from
liquified petroleum gas (LPG) or from normal butane, again from LPGs—is dehydrogenated
to make isobutylene, which is combined with methanol to make MTBE in a very
conventional MTBE unit. These would be large-scale units, generally on the order of
500,000 metric tons per year or larger. Today, roughly about two-thirds of the MTBE
produced in the world is derived from butane, versus that made from the fluid catalyst
cracker, C4s, that I mentioned earlier.
The MTBE is then usually shipped in large commercial movements, such as barges,
and moved into industry storage for the oxyfuel program. The production is all year around,
but the demand is very seasonal. During the wintertime, the MTBE is removed from storage
and shipped to the refinery where it will be stored or blended into gasoline to make the
oxygenated gasoline.
The MTBE production increased quickly from the mid-1980s. In 1980, there were
only four process units. (I report the capacity in two different ways: for refiners, that's
19,000 barrels a day; for other people, it's in millions of gallons per year, in this case about
290 million gallons per year.) In 1985, the number of plants started to grow and when we
get to 1990, we're up to 78 plants throughout the world for roughly 230,000 barrels of
capacity or about 3.5 billion gallons per year production capacity of MTBE.
18
-------
There were 117 plants by the end of 1992 for a total capacity of roughly 360,000
barrels per day or almost 6 billion gallons per year of production capacity throughout the
world. We've made estimates of how many plants and how much production capacity will
be hi place. Right now it looks like about 186 plants will be hi place by 1995 for about
600,000 barrels per day of capacity or a little over 9 billion gallons per year of MTBE
throughout the world. This doesn't include any TAME plants or ETBE plants that might be
in place by then as well. In 1995, roughly two-thirds of that MTBE will be needed in the
United States, assuming follow through with the Clean Air Act requirements.
Our oxyfuel experience really started back hi Colorado, where they introduced the first
oxygenated fuels program for reducing CO. They started off with a large demonstration fleet
test that they conducted there in late 1987 and then instituted their first oxyfuel program of
1.5 weight percent oxygen. Just for convention, the 2.0 weight percent oxygen is about 11%
MTBE and 2.7% oxygen is about 15 volume percent MTBE, which right now4s required for
most oxyfuel areas. Colorado increased their oxygen number to 2.0% in the 1988-1989
winter and to 2.6% in the 1989-1990 winter. Also that same whiter season, three other areas
introduced an oxyfuel program: one in Arizona and two in Nevada. So there's actually a
number of years of experience before the Clean Air Act started oxyfuel programs for
controlling CO emissions.
Since that time, a number of these areas have moved up their oxygenate level, and for
this past whiter we're up to 2.7% in all these areas, except for Tucson, which was not a CO
nonattainment area before introducing their oxyfuel program. That roughly describes our
oxyfuel experience in the United States now. As I mentioned earlier, MTBE was actually
around much longer, where it was used mostly in premium gasoline in concentrations on the
order of 5 to 11%, although with the use of the Sun Oil Company waiver in 1988. they used
it as high as 15% in their premium gasolines.
In Europe, it has also been widely used, actually before the United States. A similar
oxygenated gasoline in Europe, introduced by Neste Oy Company, is called "city gasoline."
It was introduced hi Finland, where the company is actually headquartered. This program
started hi May 1991; it requires two weight percent oxygen using 11% MTBE. The gasoline
now represents about 80% of the gasoline sold hi Finland, where it's promoted as an
emission reduction gasoline.
Methyl tertiary butyl ether in Europe is mostly used in premium gasolines as an
alternative to using lead and aromatics for octane. Lead is now being phased out in Europe.
Much of the gasoline now has reduced lead content or no lead. Also, in Europe only about
15% of the cars have catalyst and vapor recovery systems, so they're not quite up to where
the United States is on emission reduction technology on their cars. As a result, their car
emissions of unburned gasoline are generally five to ten tunes greater than for the U.S. cars
that we have today. So the absolute emissions are higher. But when using MTBE and
getting the element benefits, and the combustion improvements, the absolute reduction
benefits are much greater there because they have much higher emissions from their cars on
a grams per mile basis.
19
-------
Some of the emission reduction benefits of reformulated gasoline using 15% MTBE and
blending that to the refinery are based on EPA's "complex model", which is the most recent
version that just came put about two weeks ago. This model takes all the emissions data and
tries to determine the relative impact of all the different fuel parameters, working toward the
reformulated gasoline program goal of 15% reduction of VOCs and toxics, starting January
1995. That requirement then jumps to a 25% reduction of VOCs and toxics in the year
2000. The estimates show CO down 17%, VOCs generally reduced by 6%, NOX down by
3%, SOX (due to dilution of the sulfur) down about 15%, and total toxics down about 18%.
Of the individual toxic pollutant contributions in milligrams per mile, benzene is the largest,
and you generally see almost a 20% reduction of benzene with MTBE. Of all the toxics
mentioned here, the only one that really goes up is formaldehyde. To some degree that's
offset when you look at the VOC reduction, because much of the linburned hydrocarbon
coming out of the tailpipe generally will eventually go through a formaldehyde stage and be
oxidized in the atmosphere, and there's less VOCs coming out of the tailpipe. -So when you
look at the total inventory, there's somewhat of an offsetting effect from just the VOC
reduction.
So where does this leave us now? We went Jhrpugh the history of using oxygenates
and looked at the emission reduction benefitsfor the: Clean Air^ Act and reformulated
gasoline. Today, MTBE and bio-ethanol are the main oxygenate options. Future options
include TAME and ETBE, which are much less volatile ethers; thanMTBE.Volatile organic
compound reductions will be more important as well as lower volatility, so these ethers will
take on more importance.
Methyl tertiary butyl ether is one of the most widely produced petrochemicals in the
world. It ranks with petrochemicals such as toluene, benzene, and others. It has now been
in commercial use for over 20 years, so there's been a lot of experience with it, and it has
been used just about everywhere (i.e., all the major gasoline markets in the world). Also,
the oxyfuel programs were demonstrated in the United States for 3 years prior to the Clean
Air Act being developed.
Based on all the emission testing that's available, the use of MTBE helps reduce major
air toxic pollutants and generally contributes to reductions in the six criteria pollutants. As a
replacement for lead, MTBE helps reduce CO, helps reduce ozone by reducing VOC and
NOX, as well as reduces the NOX directly, along with SOX and particulates. So generally, the
use of MTBE is viewed as helpful and beneficial in reducing the mobile source pollutants.
Discussion
In response to one question, the speaker clarified that the model showing emission
reductions was the EPA complex model, not the Mobile 5 model. One person asked whether
there would be an overall reduction of pollutants in .view of the reduction in fuel efficiency
and consequently greater consumption of gasoline. The response was that EPA measures
emissions in terms of grams of pollutants per mile driven, not per gallon consumed. Thus,
any increase in emissions associated with a reduction in fuel efficiency would be captured by
that measure.
20
-------
Session Two
EMISSIONS AND AIR QUALITY UNDER
WINTER CONDITIONS2
Chair: Steven Cadle, General Motors Corporation
Discussant's Summary
Robert Sawyer, University of California, Berkeley
It is estimated that by the mid-1990s, approximately 6 billion gallons of MTBE per year
will be used in the United States. This estimate is significant and requires that scientists and
the public understand the consequences of using MTBE in fuels.
During Session Two, three speakers provided experimental data on the effects of
MTBE use on emissions under cold conditions. Ken Knapp of the EPA/ORD reported the
emission results, using standardized Federal testing procedures at temperatures from -20 to
75 °F. These data showed that CO emissions increase as temperature decreases and that CO
is reduced when MTBE is added to fuels. However, the effect of MTBE addition on CO
emissions is not consistent, which suggests vehicle-to-vehicle differences and the need for
more data to establish a statistically significant trend for the effect of MTBE on CO emission
at low temperatures. Chandra Prakash of Environment Canada provided data on two
vehicles tested at temperatures from -20 to 70 F°; the data showed no significant MTBE
effects on CO emissions. Finally, David Veazy of the University of Alaska-Fairbanks
reported on vehicle emissions data observed in Fairbanks, Alaska, under actual use
conditions (rather than FTP conditions). He noted that cold weather cold starts there are
often followed by approximately 15 to 30 min of cold idle to allow the vehicle to warm up.
These operational conditions are quite different from the FTP. The CO emissions during this
Fairbanks cold start sequence dominate. He concluded that adding MTBE had little effect on
cold-start emissions but reduced on-road CO emissions. Increased use of electric engine
heaters would be a more effective means of reducing CO emissions. These data are not
sufficient to establish that adding MTBE reduces CO emissions under cold-weather, cold-start
conditions.
The next three speakers reported on ambient air quality measurements. Gerry Guay of
the Alaska Department of Environmental Conservation noted that extreme winter conditions
in Fairbanks—very cold temperatures, very little sunshine, and stable, very low
inversions—cause relatively low CO emissions (estimated at 50 tons per day) to produce
ambient CO levels that exceed air quality standards. He outlined measurement methods and
difficulties. Roy Zweidinger of EPA/ORD reported on air quality in outdoor and indoor
environments and on fuel composition measurements in Fairbanks, Stamford, and Albany.
Copies of the slides for the presentations in this session may be found in Appendix B.
21
-------
High concentrations of MTBE and other hydrocarbons were found in motor vehicle repair
service bays. Measurements taken from a Fairbanks home having an attached garage also
showed high levels of gasoline vapors, including MTBE. Larry Anderson, of the University
of Colorado-Denver, analyzed the effect of the Colorado fuel oxygenates program on air
quality over the past 6 years. He attempted to determine the relation between CO levels,
which decreased strongly during the 6-year period, and oxygenate use. He concluded that
adding oxygenate had no statistically significant effect on CO emissions.
Both emissions and ambient measurement data raise concerns related to wintertime CO
emissions: "Is MTBE doing any good?"
Discussion
In response to a question, the source of the 6 billion gallons per year figure for MTBE
use in the United States was attributed to an estimate that worldwide use would be about
9 million gallons of MTBE and that two-thirds of this amount would be used in the United
States. What fraction of the MTBE use would be independent of the wintertime oxygenate
program would depend on several factors. If the data were based on the difference between
year-round-reformulated gasoline with 2.0% oxygen and winter gasoline with 2.7% oxygen,
then 10 to 20% MTBE would be used strictly for the wintertime oxygenate program. If all
of the oxygenate (hi addition to that which would be used for octane improvement) were
there for wintertime CO, then approximately 40% of the oxygenate would be used in the CO
program.
Another questioner asked if there were no regulatory requirements for oxygenated
fuels, what fraction of the MTBE would be used? It was noted that MTBE is added to fuels
for a number of reasons, including reducing vapor pressure requirements, which has pushed
displaced butanes into MTBE production.
One commentator noted the difficulty of determining the effect of adding oxygenate on
Denver's CO levels because of the number of Denver's CO-related control measures.
He noted that early analysis of the effect of the wintertime oxygenate program on CO
emissions during the first year showed decreases of 15% hi New Jersey, 11% in
Connecticut, 15% in Washington, 17% in Oregon, and 10% in California. He suggested that
Larry Anderson apply his analyses to these new data. In response, Larry Anderson said that
his analysis would work well in other areas, but that these same areas, as hi the case of
Denver, also were experiencing CO reduction from other programs.
It was noted that the relatively high vapor pressures of the cold weather fuels suggested
a need to measure cold weather evaporative emissions (as they might be important for indoor
emissions). The response was that these measurements, although difficult, would be made,
and that the expectation is to see the increased vapor pressure increasing low molecular
weight hydrocarbons more than increasing MTBE emissions.
An inconsistency was pointed out between vehicle emissions measurements that showed
little effect of MTBE on formaldehyde emissions and data for Denver that showed significant
increases in ambient aldehydes. One respondent noted that secondary aldehydes (formed hi
22
-------
atmospheric reactions) could dominate atmospheric measurements, whereas another person
indicated that ambient increases were from direct emissions.
A member of the audience questioned the statistical significance of the Denver data.
There was no statistically significant effect of MTBE on CO at the 95% confidence level;
whereas a 10% change at some locations may have been missed, this would not have
occurred at the downtown Denver site. Another person questioned whether meteorological
changes were adequately addressed. The response was that using data from six winters
should account for meteorological changes.
The potential beneficial effects of diluting gasoline with MTBE and thereby reducing
gasoline toxicity were pointed out. However, it also was stated that such a dilution effect
would accompany any nontoxic additive and was not peculiar to MTBE. Another beneficial
effect of MTBE, other than CO reduction, is as an octane improver compared 4o other toxic
compounds, such as benzene.
23
-------
Session Three
HUMAN EXPOSURES3
Chair: Charles Powers, Health Effects Institute
Discussant's Summary
Steve Colome, Integrated Environmental Services
Given the extremely rapid turnaround of the research presented in this session on
human exposures, there was not enough time for the authors to write papers; therefore, this
meeting is the first presentation of findings. As a result, this summary was exfemporaneous.
The quality of this research and the speed with which it has been refined and presented
is admirable. The interval between .initial complaints associated with using MTBE last
whiter and the decision to address the issue was extremely short. Sponsors and contractors
quickly assembled the funding and equipment, developed the field protocols, and conducted
measurements and analyses. A few months later, there was evidence of fairly clear data
results—this is indeed impressive, if not unprecedented. The funders and the researchers
who were involved in this project are to be congratulated for their wisdom, skill, and effort
in completing such an extraordinary set of studies.
The picture presented by these investigators is relatively coherent. With this new work,
there is substantially more information available now than there was before measurements
were taken. The exposure values that have been generated by this project may even have
sufficient certainty at this stage to frame the exposure element of an initial assessment of
acute health effects.
Despite the clear picture presented by this research, there are some areas of concern
that need to be addressed. There was a serious lack of information that was available prior
to this meeting on human exposures to, and health effects that may result from, MTBE. The
experience of this meeting should reemphasize the value of collecting at least a modest
amount of exposure data for MTBE and the Other alternative fuels that are under
consideration. The following is a summary of the presentations.
Ted Johnson discussed a sampling program designed to gain insight on MTBE
exposures that the public would experience at service stations during refueling operations.
The measurements that he presented were made during late winter and early spring hi the
northeastern United States. Results from the survey indicate the concentration ranges that
occur during vehicle refueling. The concentrations are less than 10 ppm hi all cases and less
than a couple of parts per million in almost all circumstances. The data also provide some
Copies of the slides for the presentations in this session may be found in Appendix C.
24
-------
evidence that stage-two vapor recovery further reduces exposures, as compared to
uncontrolled releases of tank vapors. Based on the median values presented by Ted Johnson,
it appears from this preliminary data that vapor recovery may provide almost an order of
magnitude of exposure reduction.
Jack Hinton reviewed over 2,000 occupational exposure measurements that have been
conducted on workers in the fuels industry. The vast majority of work-shift exposures were
less than 10 ppm, and in a majority of the labor categories, the maximum short-term
exposures were around 100 ppm. Isolated concentrations up to 1,000 ppm were observed in
barge-related tasks in the transportation sector. These were the highest concentrations
observed, and they were for nonroutine activities. One important point is that the
measurement sets reported by Dr. Hinton were not collected for the purpose of developing
exposure distributions. These measurements represent an assembly of exposure values and a
useful compilation of the data.
Paul Lioy discussed commuting exposures in the same area of the country that Ted
Johnson's work was conducted. Most of these concentrations—note the change in
measurement units—were less than about 30 ppb, with averages generally below around
10 ppb. As is often found in exposure assessments, there was a wide distribution of
concentrations with some evidence of higher outliers. The highest concentrations were
measured inside a vehicle cabin, presumably from exposure to the car's own exhaust.
Alan Huber tried to put these data together and make some sense of the compilation of
exposures. He considered activities and time spent in those activities using
microenvironmental concentrations that were, in most cases, less than l/100th of a ppm,
according to some of the data collected in Alaska. He considered an approach to modeling
MTBE for long-term exposure assessment. This would have a primary focus on cancer risk
assessment. However, the more difficult challenge, and the focus of most of the health
research presented at this meeting, is to understand and model short-term exposures to
MTBE. That should be the direction for future exposure modeling.
Discussion
It was pointed out that EPA now has a method for MTBE analysis that has no
interference from hydrocarbons and that the method is specific for MTBE. Another
commentator noted that even though ambient measurements were initiated in rapid response
to the Alaska situation, intentional efforts were made to get measurements over a range of
different types of exposures to support the exposure modeling.
It also was noted that Dr. Huber used the upper limit for his microenvironmental
exposure assessment, and, thus, the estimates were conservative. However, because
Dr. Huber's exposure ranges were developed in part from Alaska data, one should be aware
that Alaska, unlike most other areas in the United States, has exceedingly high levels of CO
on the neighborhood scale, not just at street-side. Thus, if CO and MTBE are correlated,
then MTBE estimates in Alaska neighborhoods would be much higher. However, it would
probably be inappropriate to extrapolate Fairbanks' background levels to other cities, because
during the winter there are strong inversions and Fairbanks is in a "bowl" in which
25
-------
concentrations build up. In other cities, weather systems come through clearing CO out of
the air. So it would be expected that background levels in other places would be
considerably lower than the background in Fairbanks.
A commentator noted that CO is emitted from the tailpipe, whereas MTBE is an
evaporative emission either from refueling or evaporation from the car. Indeed, the data are
not adequate to support a one-to-one correspondence between CO and MTBE. Another
person added that in the average car about 10 to 40 g/mile of CO is emitted and only up to
about 10 mg/mile of MTBE is released through the tailpipe.
It was observed that eventually, as one attempts to relate the exposure analyses in
microenvironments to other acute or chronic effects information, it will be important to sort
out which people it is that the law is asking us to protect in this particular case, which will
finally turn out to be a risk management question.
26
-------
Session Four
ACUTE HEALTH EFFECTS OF MTBE EXPOSURE4
Chair: Lawrence Reiter, Office of Research and Development,
U.S. Environmental Protection Agency
Discussant's Summary
Roger O. McClellan, Chemical Industries Institute of Toxicology
This session included ten papers that addressed potential acute health risks from MTBE
exposure. The topics included measuring blood levels of MTBE and, in some "cases, its
metabolite (TEA) under both field and controlled exposure conditions. Other topics included
assessing various health indicators in individuals exposed to MTBE under environmental
conditions or in controlled exposure studies. These papers provide substantial information
for assessing the potential human acute health risks of MTBE exposure.
Figure 1 is a paradigm linking sources of MTBE with exposure, dose, and response
data. It is necessary to understand the quantitative relationship between these data.
The data presented in Session Four, and in other sessions at this conference, considered
a range of MTBE sources resulting in exposure of people in their occupational, consumer,
and environmental settings. Occupational exposures include those occurring during the
manufacturing, blending, transportation, and distribution of MTBE or MTBE-containing
fuels, as well as exposures of service station, garage, and taxicab personnel. Consumer
exposures can occur during refueling or during vehicle use; environmental exposures can
occur anywhere in the locale where MTBE is being used as a fuel additive. These exposures
occur principally when oxygenated fuel use is mandated, although MTBE is routinely used as
an octane enhancer for premium fuels.
Exposure
Summary data on MTBE exposures from various microenvironments are shown in
Table 1. These data are presented as background for the discussion that will follow. The
exposure values are typically well below 1 ppm (3.6 mg/m3), the exception being refueling
values. Other values of special interest are those for automotive shops (0.1 to 0.5 ppm or
0.36 to 1.8 mg/m3).
Several sets of field data on acute health effects of MTBE exposure are of special
value. These are the data from Fairbanks, AK, and Stamford, CT, in which
Copies of the slides for the presentations in this session may be found in Appendix D.
27
-------
to
00
SOURCES
Occupational
Manufacturing
Blending
Transport
Distribution
Servicing Vehicles
Consumer
Fueling
Servicing
Environmental
EXPOSURE
Air MTBE
MTBE on skin
DOSE
Blood MTBE
Blood TBA
RESPONSE
Symptoms
Behavioral
measurements
Physiological
measurements
Figure 1. Paradigm for evaluating potential human health risks of exposure to methyl tertiary butyl ether (MTBE) and other
oxygenates.
-------
TABLE 1. Summary Data on Microenvironmental Exposure to
Methyl Tertiary Butyl Ether
Source
Microenvironment
Concentrations
EPA Estimate
ppm (mg/m3)
Gas refueling
Commute/in vehicle
Automotive shop
Public garage
Residential garage
Residence
Office
School/public buildings
Outdoors
Pump
Other
High
Low
High
Low
High
Low
1-10 (36)
0.1-1(3.6)
0.005-0.1 (0.36)
0.1-0.5(1.8)
0.1-0.5(1.8)
0.1-1 (3.6)
0.001-0.005 (0.018)"
0.005-0.01 (0.036)
0.001-0.005 (0.018)
0.001-0.01 (0.036)
0.001-0.01 (0.036)
0.01-0.1 (0.36)
0.001-0.01 (0.036)
measurements were made of both air and blood concentrations of MTBE. In Fairbanks,
during MTBE fuel use, measurements of MTBE in the garage workplace were made along
with measurements of MTBE in the blood of 18 garage workers. The median value for the
garage was 370 /xg/m3 (range of 40 to 2,930 /xg/m3), which decreased to 130 /xg/m3 (ranging
from nondetectable to 510 /ig/m3) following the cessation of the distribution and sale of
MTBE fuel. Similar MTBE air data were developed for Stamford along with data on MTBE
in blood. The blood and air measurements were not collected simultaneously. The air
concentration of MTBE was measured in four garages: 10 samples yielded values from 5 to
1,549/xg/m3.
In addition to the above data on occupational and environmental exposures, exposure
data were available from two controlled human MTBE exposure studies. One study was
conducted at Yale University and involved a 1-h exposure to MTBE at 1.7 ppm (6.0 mg/m3)
MTBE. Blood measurements of MTBE were made on some subjects. A second study was
done at the EPA/University of North Carolina-Chapel Hill facility and involved a
1-h exposure to 1.4 ppm (5.0 mg/m3) of MTBE.
Dose (Blood Methyl Tertiary Butyl Ether and Tertiary Butyl Alcohol)
A major strength of the present data set is the availability of information on dose, in
this case, blood MTBE and blood TBA, which complements the exposure data and provides
29
-------
a linkage between exposure and response data (Figure 2). Blood TEA levels are available
for Stamford and for several individuals exposed under controlled exposure conditions.
In Fairbanks, the median blood MTBE levels in workers (mechanics* service station
attendants, and garbage collectors) were slightly above 1 mg/L (preshift) and were further
elevated (postshift) to approximately 2 /xg/L. Individual values ranged up to slightly more
than 30 /xg/L. As a group, the values for gasoline pump attendants were highest, followed
closely by mechanics, with the lowest MTBE blood concentrations being observed in taxicab
drivers. Blood MTBE measurements also were made on commuters, who had MTBE blood
concentrations from 0.18 /xg/L (precommute) to 0.83 /xg/L (postcommute) when MTBE fuel
was being sold in Fairbanks. A good correlation was observed between workplace exposure
concentration of MTBE and blood MTBE levels; a 10-fold increase in workplace air
concentration (from 30 to 300 /xg/m3) resulted in approximately a 10-fold increase in blood
MTBE over the preworkshift value (from 0.5 /xg/L to 5 /xg/L). These data may be
normalized as follows: 1,000 /xg/m3 (0.28 ppm) MTBE exposure yields a 17-/xg/L increase
in MTBE (or restated, 0.10 ppm [360 /xg/m3] MTBE exposure yields a 6.0-/xg/L increase in
MTBE). These values have been calculated by assuming an equilibratory relationship
between air MTBE and.blood^MTBJE^py^a^range of exposure concentrations from
environmental levels to the high concentrations measured in occupational settings (and those
used in the controlled chamber studies). Blood TEA values were not reported for Fairbanks.
In Stamford, Connecticut, blood MTBE and TEA measurements were obtained from
gas pump attendants, car repairmen, and commuters. The MTBE measurements ranged from
just over 30 /xg/L (car repairmen) to less than the detection limit of 0.05 /xg/L. The median
value was highest for three gas pump attendants (about 15 /xg/L), intermediate for 21 car
repairmen (about 1.5 /xg/L), and lowest for 14 commuters (about 0.1 /xg/L). The median
blood TEA values were similar in ranking: 75 /xg/L, 15 /xg/L, and 2 /xg/L, respectively.
The MTBE blood values for garage workers in Stamford were generally similar to
those found in Fairbanks, whereas the commuter blood MTBE levels appeared to be slightly
lower. Because the air and blood concentrations of MTBE were measured on samples
collected at different times, it is not possible to directly correlate them. However, using the
normalized relationship developed earlier from the Fairbanks data, the normalized
relationship between air-to-blood MTBE developed earlier from the Fairbanks data appear to
fit with the Stamford data. Stamford air data yielded estimates of MTBE in blood that are
consistent with the measured values.
The relationship between blood MTBE and blood TBA is of interest. The extent to
which the TBA blood levels exceeded those of MTBE is in accord with the MTBE clearing
rapidly from blood and being distributed to tissues, such as fat, where it is slowly released
and metabolized to TBA, which has a long residence time in blood and tissue before being
metabolized and/or excreted.
In controlled human exposure study at Yale University, four subjects (two males and
two females) provided blood samples before, during, and up to 1.5 h after a 1-h exposure to
1.7 ppm MTBE (approximately 6.0 mg/m3). The peak blood MTBE concentration was
17 /xg/L at the end of the 1-h exposure and decreased to 7 /xg/L 1 h after exposure was
30
-------
EXPOSURE DOSE RESPONSE
Air MTBE Blood MTBE Blood TBA
10,000
3,000
1,000
m
.E
? 300
LU
CQ
i-
2 100
30
10
3
r Y4 100
EPA4
- F2 30
-
S»
F2M
-
— -
- F
_
10
_j
? 3
in
CD
H
1
0.30
0.10
S2
0.03
100
F- 55
-
-
| 30
S1M
! S 10
i
si
F2M
—
-
F
s£
"o
2
S- ^"
>M
3 B 3
3-
ca
H 1
0.30
- S2 saM o.10
Q
0.03
~S1 Y i
j 4 I Negative
EPA '
S,M
"S S2
1
~ T F 1
_ o } Suggestive
EPA4 ss bJ
-
SjSgM
I
-
_
_
F - Fairbanks
S - Stamford
EPA - Environmental Protection Agency
Y - Yale
1 - Gas pumpers
2 - Garage workers
3 - Commuters
4 - Controlled study subjects
M - Median
Figure 2. Synthesis of Exposure-Dose-Response Paradigm for methyl tertiary butyl ether (MTBE).
-------
terminated. Data were not reported on blood TEA levels, but would be of interest if they
could be obtained. .
In the Yale study, it appeared that the blood MTBE levels were still increasing at the
end of the 1-h exposure period. Using the air-to-blood relationship normalized from the
Fairbanks data, the MTBE blood level might be approximately 100 pg/L for exposure to
1.7 ppm (or 6 mg/m3) of MTBE, contrasted to the 17 /zg/L of MTBE that was measured.
These differences may relate, in part, to the differences in exposure duration (1 h for the
controlled exposure and longer for the worker exposures) and equilibrium time for MTBE
blood levels of exposure concentrations exceeding 1 h. An additional factor may be the
ventilation (exercise) levels of the controlled exposure subjects versus the exposed workers.
It is possible also that the Fairbanks study systematically underestimated the air
concentrations of MTBE to which the workers were exposed, thus overestimating the blood
MTBE to air MTBE relationship. Or there may be unidentified factors that are influencing
the relationship.
In EPA's controlled human exposure study, two individuals who were exposed for
1 h to 1.4 ppm of MTBE had blood concentrations of 15 and 27 jtg/L, respectively. These
concentrations decreased with exposures of 15 min, reaching 1 to 2 /xg/L at
4 h postexposure. Blood TEA levels rose steadily during the 1-h exposure and maintained a
steady concentration of 7 to 10 jig/L up to 7 h postexposure. Using the values from the •
Fairbanks study, it is estimated that the 1.4 ppm (5.0 mg/m3) exposure to MTBE would have
yielded blood MTBE concentrations of about 85 jig/L, as compared to the 15 and 27 /zg/L
MTBE measured, respectively. With the Yale study, the apparent discrepancy may be from
the exposure duration, exercise levels, or other unidentified factors. The apparent half-time
of blood TBA associated with the 1-h exposure to MTBE suggests that if the duration of the
MTBE exposure had increased, the blood TBA levels would have increased further from both
MTBE exposure concentration and duration.
Health Responses
The range of health effect responses evaluated for individuals who were exposed to
MTBE is impressive. The field studies conducted in Fairbanks and Stamford used
questionnaires to ascertain health complaints, which included headaches, eye irritation,
burning sensations hi the nose or throat, cough, nausea or vomiting, dizziness, spaciness or
disorientation, as well as other health complaints not thought to be associated with MTBE
exposure. In Fan-banks, the questionnaires were administered both during the period when
MTBE-oxygenated fuels were used (Phase I) and after then- use had ceased (Phase II).
Between Phase I and Phase II, the complaints decreased.
Of special note are the analyses conducted using postshift blood levels of MTBE as
exposure measures. Even with the small study population of 18 individuals, there was a high
correlation between blood levels of MTBE and the health complaints (headaches, eye
irritation, burning of nose or throat, cough, nausea, dizziness, and spaciness) attributed to
MTBE exposure. There was no similar association for health complaints, such as muscle
aches and fatigue, not relatable to MTBE exposure. The authors emphasized the limitations
of the study related to possible bias and the limitations of the sample size.
32
-------
The Stamford study was similar to that conducted in Fairbanks although it was
conducted later and in a community whereas health officials were not aware of any health
complaints or public concerns related to MTBE use. The questionnaires asked about the
prevalence of 15 symptoms over the previous month that were unrelated to the cold or flu.
The data from occupationally exposed individuals were analyzed by grouping the population
into two groups based on blood MTBE levels (over 3.8 fig/L [top quartile] and under
3.8 ng/L). This analysis suggested an association between blood MTBE and key symptoms,
such as was seen in Fairbanks. The toxicokinetic data now available suggest that it may be
useful to perform a similar analysis using TEA, because it may be a better surrogate internal
dose measure for MTBE exposures.
The investigators who conducted the Stamford, study offered recommendations for
conducting future epidemiological studies with MTBE-exposed populations. These included:
• The location and timing of a study of the general population should be chosen to
maximize potential exposure to MTBE and to minimize bias in reporting.
• Occupational studies should focus on persons with exposure to gasoline.
• Studies should focus on measuring symptoms that are present at the time MTBE
exposure is measured.
• Comparison groups should have comparable exposures to gasoline fumes, but not to
MTBE exposures.
• The contribution of exposure routes (other than inhalation) to the body should be
examined.
• Follow-up studies should be planned for more definitive evaluation.
The two controlled human exposure studies represent substantial undertakings. The
Yale University study involved 43 persons, ages 18 to 34 years, and the EPA study involved
37 persons, ages 18 to 35 years, with about equal numbers of males and females in each
study. The Yale study involved three exposures: a control exposure to air, a 1-h exposure
to 1.7 ppm MTBE, and a 1-h exposure to a 17-component mixture of volatile organic
compounds. The EPA study involved a control exposure to air and a 1-h exposure to
1.4 ppm MTBE.
Both studies used a range of behavioral and physiological indicators that might be
altered in response to exposure to low levels of organic solvents. For the EPA study, these
included:
• Indicators of symptomatic responses, including headaches, nasal irritation, throat
irritation, cough, eye irritation, odor quality, and dizziness. These were measured
before and during exposure using questionnaire responses.
• Indicators of neurobehavioral responses measured before and at the end of exposure.
• Indicators of upper airway inflammation measured before, immediately after, and
18 h after exposure. This included nasal lavage to assess cytological and
biochemical indicators of response.
• Indicators of ocular response, before and after exposure. These included hyperemia,
noninvasive tear film breakup, and impression cytology.
The assessments conducted in the Yale University study were quite similar. When
considering the results of both the EPA and Yale studies, it is important to keep in mind that
33
-------
the data-gathering phase of both studies has only very recently been concluded, that a
tremendous amount of data was generated, and that only preliminary analyses had been
completed at the time of the MTBE conference. Based on their preliminary analyses, the
investigators concluded that there were no symptomatic responses or findings suggestive of
ocular or nasal inflammation, cell damage, or irritation in normal, healthy, young people
exposed for 1 h to 1.4 or 1.7 ppm MTBE.
Synthesis
The information on acute health..effectsof...MTBE in humans can be assessed within the
framework of an exposure-dose-response paradigm. The controlled exposure studies
involving 80 individuals exposed to either 1.4 ppm (5.0 mg/m3) or 1.7 ppm (6.0 mg/m3) for
1 h with rigorous evaluation of health responses yielded no symptomatic responses, nor any
findings suggestive of ocular or nasal inflammation, cell damage, or irritation. -These
negative findings, with exposure concentrations at the level of the highest values seen in
occupational settings and several orders of magnitude above typical environmental levels, are
reassuring, especially regarding exposures of up to 1 h. The results of the controlled studies
are even more reassuring when the blood MTBE levels in the controlled exposure studies are
compared to those observed in the field studies. The peak MTBE blood levels observed in
the controlled exposure studies are typically greater than those observed in the field studies.
However, it is important to note that the blood MTBE levels were still increasing at the end
of the 1-h exposures and, if the exposures had been continued, even higher blood MTBE
levels would have been realized.
The results of the field epidemiological studies in Fairbanks and Stamford appear to
contradict the results of the controlled exposure studies. The studies, launched on short
notice, were basically exploratory. The conduct and analysis of the studies were
handicapped by the potential for bias, confounding, and small sample size. The suggestive
effects at exposure concentrations typically lower than those of the controlled exposure
studies are difficult to reconcile with the clearly negative results of the controlled exposure
studies. One possibility might be confounding, which is related to the low odor threshold for
MTBE. It is possible, particularly in the Fairbanks study, that the suggestive effects
observed were linked to the odor of the new MTBE fuel. This may have been compounded
by the publicity and controversy over the mandated oxygenates program and the 140/gallon
increase hi prices. However, the data analysis was strengthened in both field studies in
which small populations were stratified by blood MTBE level.
An additional factor that needs to be considered is the disposition of the TBA
metabolite of MTBE. As may be seen in Figure 2, the blood TBA levels observed in the
field studies were greater than the blood MTBE levels by factors of about 5 (TBA to MTBE)
for the highest exposed individuals, about 10 for the intermediate exposures, and about
20 for the lowest exposures. The highest TBA levels observed in Stamford (over 50 fig/L)
exceeded those observed in the controlled exposure study (5 to 10 /zg/L). This reflects both
concentration and duration of MTBE exposure, with MTBE clearing rapidly from blood,
and TBA clearing relatively slowly. These data suggest that TBA may be a useful internal
dose surrogate for MTBE, perhaps better than measurements of blood MTBE. It would be
34
-------
very useful to reanalyze the epidemiological data, stratifying the subjects by blood TEA
levels, as a marker for cumulative MTBE exposure.
To strengthen the present impressive data set on the potential acute health effects of
MTBE exposure, additional targeted research studies would be advantageous. This snould
include additional pharmacokinetic studies to allow more adequate modeling of blood MTBE
and TEA levels for various exposure scenarios. This will aid in linking the results of
controlled human exposure studies to data from field epidemiological studies. The results of
modeling of MTBE pharmacokinetics can then be used in designing additional controlled
exposure studies directed at validating the pharmacokinetic models through variations in
exposure concentration and duration and the level of exercise (ventilation) of the subjects.
The controlled exposure studies can use a more targeted set of response indicators than the
earlier studies to assess the potential for responses from longer duration exposures.
The present database clearly indicates a lack of response to the short-term exposures
typically encountered by the general population. However, additional epidemiological studies
focusing on occupationally exposed populations with measurements of both blood MTBE and
TBA as indicators of internal dose would be useful in ascertaining any potential effects of
longer term exposures to MTBE and in differentiating responses to MTBE from those arising
from other fuel components found in the same environment. A key element of any future
studies is to use an exposure-dose-response paradigm orientation with blood levels of MTBE
and TBA serving to document the internal (or effective) dose for various exposure scenarios
and to provide a mechanistic link to any possible MTBE-exposure related responses.
Further, it should be noted that the experience gained with MTBE provides a "blueprint" for
investigating exposure-dose-response relationships for other oxygenates, such as ETBE and
TAME.
35
-------
Session Five
CHRONIC HEALTH EFFECTS OF MTBE EXPOSURE5
Chair: Robert Drew, American Petroleum Institute
Discussant's Summary
John Doull, University of Kansas
Session Five focused on the chronic health effects of MTBE exposure and included
presentations by Larry Andrews of ARCO Chemical on the chronic inhalation studies, Jeff
Gift of EPA on RfCs, and Charlie Hiremath of EPA on cancer assessment for MTBE.
Dr. Andrews discussed the results of the cooperative oncogenicity study in rats and
mice, and he presented preliminary data from a follow-on, 28-day study in male rats. The
major finding was an increase in kidney tumors in the male rats, and there was considerable
discussion within the presentation and in the subsequent discussion as to whether this MTBE
effect met all of the EPA-recommended criteria for an alpha-2/z-globulin mechanism. Methyl
tertiary butyl ether clearly meets most of the criteria, and some of the other agents that are
thought to act through the alpha-2/*-globulin mechanism do not meet all.of the EPA
recommended criteria. Although there also was a dose-related increase in interstitial
adenomas of the rat testis, the incidence range of these tumors was within the range of
historical controls. Methyl tertiary butyl ether produced an increased incidence of
hepatocellular tumors in mice, and there also was a dose-related increase in endometrial cell
hyperplasia, but these effects occurred only at dosage levels that exceeded the maximum
tolerated dose for MTBE. In his discussion of how the EPA might view this data, Dr.
Hiremath stressed the need for a weight-of-evidence approach involving all tumor types in
deciding whether a linearized multistage analysis and potency estimation should be made for
MTBE. Dr. Doull urged the agency to use the actual incidence data from the Bushy Run
study to calculate potency rather than the slope provided by a model that has multiple
uncertainties and is hypothetical at best.
In describing the methodology used to establish the RfC for MTBE, Dr. Gift discussed
neurologic/behavioral, reproductive, kidney nephropathy, and other noncancer endpoints that
were considered by the EPA RfC Workgroup. The workgroup determined that the
no-observed-adverse-effect level (NOAEL) was 403 ppm for increased liver and kidney
weights and increased severity of spontaneous renal lesions in female rats and for prostration
in female rats, along with swollen periocular tissue in both male and female rats. The
lowest-observed-adverse-effect level for these effects was 3,023 ppm. After adjusting this
value for the exposure duration, the NOAEL was divided by safety factors of 10 (for
sensitive subpopulation), 3 (for interspecies extrapolation), and 3 (for database deficiencies)
Copies of the slides for the presentations in this session may be found in Appendix E.
36
-------
to obtain an RfC for MTBE of 1 ppm. This value can be compared to a threshold limit
value (TLV) of 40 ppm that has been recommended by the American Conference of
Governmental Industrial Hygienists (ACGIH) and 100 ppm that has been adopted by the
"WEEL" committee of the American Industrial Realm Association, recognizing that these
occupational levels are for an 8-h/5-day week, whereas the RfC is for a continuous life-time
exposure.
During the discussion in response to a question about MTBE-induced endocrine effects,
it was suggested that such effects may be from the tumors seen in the MTBE rats and mice,
and it was pointed out that TEA (a metabolite of MTBE) has produced thyroid tumors in an
National Toxicology Program study.
One of the most stimulating aspects of the discussion for this session was the dialogue
between the public health ..officials who were asking for advice on what to tell-their
constituents and the scientists both within and outside the Agency who were challenged to
assess the public health significance of their studies. The EPA scientists indicated that their
evaluation of MTBE carcinogenicity is still in progress, but that it appears there is limited
evidence in animals for this effect, and if MTBE is a possible human carcinogen, it is likely
to have a very low potency.
In his opening remarks for this session, Dr. Drew pointed out the importance of this
type of cooperative research effort to both industry and the Agency and his hope that this
cooperative effort will ..serve as ..a. modelfor sirnilar ventures between industry, the regulatory
bodies, and academia.
37
-------
Session Six
NEW FINDINGS FOR OTHER FUEL OXYGENATES6
Chair: Randy Roth, ARCO
Discussant's Summary
Bernard Goldstein, Environmental and Occupational Health Sciences Institute
Session Six began with a reminder from Judith Graham of EPA that MTBE is only one
of a large number of oxygenates that have been proposed for meeting the mandates of the
Clean Air Act. Next, Randy Roth of ARCO Chemical provided information about U.S.
oxygenated fuel production capacity. As of January 1, 1993, there are about 170,000 barrels
per day of MTBE produced in the United States. For methanol, 120,000 barrels per day of
MTBE are produced; and for fuel-grade ethanol, the number of barrels produced is
approximately 90,000. There is far less TEA, ETBE, and TAME produced, the latter being
produced in quantities of about 5,000 barrels per day. Among the physical characteristics,
vapor pressure is critical: MTBE is about 7.8 to 8, ETBE is 4, TAME is 215 to 2, and
ethanol is 2.3. The octane ranges from about 110 to 115.
Ethyl Tertiary Butyl Ether
Michael Wells, a toxicologist with Amoco, reviewed the existing data on ETBE. After
briefly outlining acute toxicity studies that show ETBE's low toxicity, Dr. Wells presented
the results of a subchronic study in Fischer rats sponsored by Amoco.
Ten rats of each sex per group were exposed to ETBE for 6 h/day for 5 days a week
for a total of 4 weeks. Target exposure concentrations were 0 (for control), 500, 2,000, and
4,000 ppm ETBE. The parameters were mortality, body weights, and physical and clinical
observations. There was a neurotoxicity assessment with a functional observation battery.
Necropsies were performed on all the animals, and histopathology was performed on the
controls as well as on the high exposure group.
None of the rats died during the study, and there were no treatment-related effects
observed on body weights. In terms of clinical observations, animals exposed to 500 and
2,000 ppm generally showed no different signs than the control animals. However, during
the exposure, the 4,000-ppm ETBE group did display treatment-related effects, which were
primarily central nervous system (CNS) suppression, including general sedation, reduced
motor activity, and mild-to-moderate ataxia. Within 15 mm following termination of the
exposure, the high exposure group appeared to return to normal, suggesting that ETBE is
eliminated very rapidly.
Copies of the slides for the presentations in this session may be found in Appendix F.
38
-------
Among the many endpoints of the functional observation battery performed for the
neurotoxicity assessment, only high limb splay showed a significant difference from control.
This occurred only in the 4,000-ppm group and was notable on Day 20. Also observed in
the 4,000-ppm group, but this time on Day 5, was a significant decrease in body temperature
hi male animals. Other observations included an increase in white blood cell counts in
female rats at 2,000 and 4,000 ppm, and some increase in organ weights at the higher
exposure levels, which were not correlated with any specific histopathology.
Tertiary Amyl Methyl Ether
Health and ecological effect data on TAME were presented by Russell White, a
toxicologist at Chevron who also chairs the API Air Fuels Health Effects Work Group.
Dr. White pointed out the similarities and differences between TAME and JMTBE. The same
fluid catalytic cracking unit in a refinery that produces the equivalent of 2,000-barrels per
day of MTBE could be used to produce about 2,500 to 3,000 barrels per day of TAME.
The lower vapor pressure of TAME has advantages over MTBE in terms of the drive to
lower the vapor pressure of gasoline to meet requirements related to the ozone standard.
The toxicological information available includes a pre-manufacturing test battery
performed by Exxon Chemical, a 28-day inhalation toxicity study with neurotoxicity
assessment by Amoco (identical to the study done on ETBE), and from API, some
ecotoxicology studies, dermal sensitization studies, and odor and taste threshold studies.
Review of the single-dose acute study, the median lethal dose (LD50) study, revealed
that the signs and symptoms that are observed—ataxia hyperactivity, dyspnea, and the LD50
values—are comparable for both ETBE and MTBE. There is not a significant difference
between males and females. There was no indication of dermal irritation or sensitization by
TAME. Both an Ames test and a micronucleus study were negative.
A repeated-dose oral study was performed in which the animals were dosed 7 days a
week for 29 consecutive days with doses up to Ig/kg. Food consumption and body weights
were reduced in the high-dose males only. There were no effects in the female animals.
In the males, though, two animals out of the 10 died, apparently in relation to exposure to
the compound. The hematology and clinical chemistry were fairly unremarkable. There
were no treatment-related histopathological findings in any of the tissues.
The repeated-dose study of TAME by inhalation used an identical test design to that
reported earlier by Dr. Wells for ETBE, including identical measurement endpoints. There
was mortality in the high-dose group: about 28% of the test animals at 4,000 ppm died.
The clinical signs, including sedation, ataxia, hypoactivity, and hypothermia, were very
similar to those observed.with ETBE. Central nervous system effects that were observed
immediately after exposure were transient, with the animals essentially back to normal the
next morning. There was a decrease in body weight in high-dose males, whereas there was
no effect with ETBE. Again, no histopathological findings were observed in the high-dose
tissues. A small but significant increase in cholesterol levels was observed at the highest
dose in both males and females.
39
-------
The API has initiated some preliminary ecotoxicology studies in common test strains
evaluating the effects of TAME on aquatic organisms. Using Daphnea magnio, a fresh
water invertebrate, the no-observed-effect level (NOEL) for the test was 83 mg/L, with the
actual interpolated EC50 as 100 mg/L, which can be considered essentially nontoxic.
Preliminary results in the rainbow trout suggest a NOEL over 300 mg/L, indicating that
TAME does not appear to be a particular problem in the fresh water vertebrate species.
Biodegradation studies suggest that under anaerobic conditions, little biodegradation occurs,
which is probably related to TAME being both an ether and a tertiary structure.
Based on a review of the physical characteristics and reactivity of TAME, Dr. White
suggested that it was unlikely that human exposure levels would be higher than those
observed for MTBE. He further suggested that the current evaluation of MTBE would be a
useful base for formulating hypotheses to study TAME's potential toxicity.
Specific studies that were proposed included some 90-day repeat-dose studies focusing
on the rat kidney and mouse liver. Pharmacokinetics and metabolism studies would
particularly be of interest in understanding the extent to which the oxygenates persist in blood
and how they are metabolized. A third suggested approach would be a combined
reproductive/developmental study. Finally, based on other presentations at the meeting,
Dr. White suggested that human studies would be lowest on his list, but would not be
ignored. The smell of TAME is very similar to MTBE and some people find it even more
objectionable. Although TAME has a lower vapor pressure, more of it must be used in
gasoline to reach the same oxygen content. In addition, Dr. White remarked that it is
unlikely TAME will be the only oxygenate in the gasoline. A blend of MTBE and TAME,
or even with ETBE mixed in, would be more likely. Dr. White did not recommend a 2-year
animal study of TAME.
Health Effects of Inhaled Ethanol
The health effects of inhaled ethanol were reviewed by Robert MacPhail, Chief of the
Neurobehavioral Toxicology Group at EPA's HERL, and research professor at the University
of North Carolina. There is a rich database on ethanol, although there is relatively much
less information on inhalation than on ingestion. The targets of concern for ethanol are the
CNS for acute and chronic effects, the developing fetus, and the liver.
In his review of the literature, Dr. MacPhail found 26 papers related to inhalation
toxicology of ethanol in animals and humans. In humans, there are some old reports that
include numbness and drowsiness following exposure to 3,000 to 9,000 ppm ethanol, and
sensory irritation at 10,000 ppm ethanol, with 20,000 ppm ethanol considered to h>e
intolerable by human volunteers. The acute CNS effects of ethanol begin with CNS
stimulation, including excitement and hyperactivity, at low concentrations. At higher
concentrations, depression is seen, accompanied by muscle relaxation, incoordination, and
narcosis, which could lead to death by respiratory failure.
There are a variety of studies on hyperactivity and hypoactivity, reflex deficits, and
performance impairments in animals exposed to ethanol. Recovery is ordinarily rapid and
complete. Concentrations generally range from about 10,000 to 40,000 ppm, and durations
40
-------
are generally several minutes to a few hours. There is evidence from studies of a variety of
alcohols that the lipid-water partition coefficient is an important determinant of acute toxicity,
further suggesting that it is a membrane effect of ethanol that is responsible for toxicity, at
least in the nervous system.
Tolerance to a number of the acute CNS effects has been reported for ethanol, as well
as cross-tolerance with other alcohols and with anesthetic agents. Most of the so-called
chronic animal studies in the literature are relatively short term and have the goal of
developing an animal model of ethanol dependency.
In humans, chronic ethanol use can produce severe central and peripheral nervous
system effects, including the Wernicke-Korsakoff syndrome. Among the many other effects
reported in humans is an increased susceptibility to a number of infections. The cancers that
are associated with chronic ethanol exposure are usually limited to the oral cavity, to the
pharynx, larynx, and esophagus.
Dr. MacPhail reviewed a long and useful compilation of individual studies showing the
effects of chronic ethanol administration in a variety of organ systems. There is evidence
that the acetaldehyde metabolite of ethanol may be important, particularly in chronic CNS
toxicity.
There are good animal models of the fetal alcohol syndrome, which is characterized by
growth retardation, by a small brain at birth, hypotonia, generally decreased body tone,
cranial facial dysmorphology, and mental retardation. Initially, the fetal alcohol syndrome
was considered to be a high-dose phenomenon, but as more was learned about the syndrome,
it has become apparent that effects on the fetus can be seen in moderate-to-mild alcohol
consumers during pregnancy.
Dr. MacPhail reviewed the work of B.K. Nelson, of the National Institute for
Occupational Safety and Health, who evaluated occupational exposures to ethanol inhaled by
rats for 7 h daily. No prominent developmental effects have been observed. There are data
indicating some neurochemical changes in the offspring of rats exposed prenatally to inhaled
ethanol, but the pattern of effects is that they are generally small and inconsistent.
Discussion
It was noted that the amount of information on MTBE is of an order of magnitude
greater than that of ETBE or TAME, and the information on CO is of many orders greater
than that of MTBE. The increase in oxygenate use this past winter, which resulted in health
complaints leading to this meeting, was required by Congress to meet the CO standard.
However, in almost all areas of the country that exceeded the CO standard, the only adverse
effect of CO that can be considered to be scientifically demonstrable would be a shortening
of time to angina attacks among the population of individuals with arteriosclerotic heart
disease (i.e., angina syndrome) who have been shown to be at risk from this reversible
effect. There is a comparative risk issue in determining whether it is appropriate to cause
acute symptoms in one population, if this is in fact occurring, to protect another population.
It is possible that no angina attacks were avoided in Alaska by MTBE, in view of the
41
-------
relatively young population and the 8- to 12-h time to equilibration of CO levels with human
hemoglobin. It is unlikely that many people with angina would be outdoors exerting
themselves for 8 h during an Alaskan winter.
The almost complete lack of mixture studies also was pointed out, particularly the
mixture of an oxygenate in gasoline to which humans would be exposed. Additionally,
studies should address what the addition of an oxygenate to gasoline might do to dermal
absorption of benzene. There was a call for more basic studies of mechanism of action as a
necessary approach to introducing oxygenates to gasoline. Commentators from the floor
provided additional information about the potential distribution of TAME, supported the call
for pharmacokinetic studies and mixture studies, and emphasized that the database on human
exposure was confined to healthy individuals rather than to those who were potentially
susceptible.
42
-------
Session Seven
CONFERENCE SUMMARY
Discussant's Summary
Gareth Green, Harvard School of Public Health
The risk assessment paradigm for analyzing problems in this very complex
interdisciplinary field is well reflected in the organization of this conference. Science defined
by this conference and other research work will provide a rational basis for managing risks
posed by MTBE and other fuel oxygenates in the future. The purpose and goals of the
workshop were to answer the following questions:
What are the population exposures to MTBE?
What are the health effects of acute exposures?
Are reported symptoms due to ambient MTBE exposures?
Is MTBE a cause of cancer?
Does MTBE decrease CO at very low temperatures?
What are the quantitative risks?
What are the key risk factors (cold temperature, age, etc.)?
Are follow-up studies needed?
What are the risks of alternative oxygenates?
In her charge to the conferees, Dr. Graham focused on the relationship among
epidemiological studies, clinical studies, and animal toxicology in the health effect analysis.
There was much discussion on effects of acute exposures to MTBE from epidemiological
studies and from clinical test exposures. Also discussed was the biological activity of MTBE
in vitro. There was little information on structure—activity relations, a useful focus for
exploring and analyzing greater or lesser risks of alternative oxygenates. There were
extensive data on human exposure, particularly as to concentration and, to a lesser degree,
on tune of exposure; there were few data on the potential number of people exposed. Most
of the focus was on recent experimental findings.
Adding MTBE to gasoline reduced CO concentration in proportion to the percentage
concentration of MTBE, indicating a diluent effect rather than a chemical interaction.
However, there is a variation in emissions, depending on engine design, vehicle age, engine
size, operating conditions, and temperature effects, as shown by the cold-soak effect on
increased CO emissions.
Several pieces of data showed rather striking progress hi reducing CO hi several
communities over the last decade, primarily related to improvement in engine design and
combustion technology. This decline in the community CO levels suggests that other
approaches to CO control may be effective and may dimmish the need for oxygenate
additives; conversely, use of MTBE correlated with the reduction of CO exceedances in
some communities, notably in Denver. Methyl tertiary butyl ether was noted to increase
43
-------
carbonyls, aldehydes, and formaldehyde, which might account for some of the irritant
symptomatology that has been reported. The Alaskan studies also pointed to alternate
strategies for CO control—for example, the potential for engine preheating to reduce the
cold-soak effect on CO emissions. Another possibility for CO control is to eliminate the
10% of vehicles responsible for 70% of pollutant emissions.
The exposure parameters from indoor/outdoor to occupational settings, in auto shops,
and garages were laid out quite clearly (see Table 2). There appears to be a higher risk for
significant MTBE exposure in closed spaces, compared with ambient levels. The data show
that the pollutant levels of MTBE are worse in closed environments, such as residential
garages, work places, service stations, and microenvironments in quasi-occupational settings.
A higher proportion of time may be spent indoors in the winter months in Alaska than in
cities in the lower 48 states. This time factor for exposure needs to be factored in to
quantify chronic exposure values. -
TABLE 2.. Estimated Exposures to Methyl Tertiary Butyl Ether
Indoor 0.001 - 0.01
Outdoor 0.001 -0.1
Auto shop 0.1 - 0.5
Garage 0.1-1.0
Chambers 1.4-1.7
Gas fill-up 0.1 - 10.0
(Stage IH)
Occupational 1.0-1000.0
ppm x hours
Residence 20-41
Commute 52
Gas fill-up 13-26
Public garage 30
Outdoors 10
Auto shop 0.5
The studies using experimental chambers show where the human experimental exposure
model fits into this scale. It is very difficult to quantify the range of the gas fill-up exposure,
due to high variation. For example, instances of gasoline spillage at the pump may increase
the risk of dermal, as well as respiratory tract, absorption. There has been much progress in
controlling vapor release at the pump, as shown by the data on vapor concentration reduction
44
-------
at pumps employing Stage II vapor recovery. However, there are still huge variations in
MTBE exposure concentrations that occur in specialized occupational settings. Despite the
proposed reduction of the ACGIH TLV, there still are opportunities for massive exposures
over short periods of time, which should be examined more carefully to understand the
potential toxicity of this or related substances.
Introduction of the time factor into the exposure equation almost reverses the rank order
of total exposure to MTBE in comparative settings, raising many questions about the relative
significance of peak exposures versus chronic low-level exposures. We need to know how
well humans metabolize andexcrete MTBE..at low levels of exposure and whether there may
be an exceedance of a threshold in a short-term, high-concentration exposure.
The epidemiologic studies and the clinical tests do not document a risk for serious
illness with acute short-term exposure. The availability of a blood biomarker that correlates
with exposure concentration allows the epidemiologic studies to minimize the exposure/dose
variable. Future studies of this and related agents ought to include biomarker measures
where available. Two of the epidemiological studies reported a significant correlation of
some key symptoms at the high end of exposure to MTBE, one study in the cold climates of
Alaska, and the other in the milder temperatures of Connecticut. In contrast was the study
that failed to find differences hi symptoms in comparative populations in neighboring New
Jersey. A third study showed no effect of MTBE exposure on health insurance claims, as
may be seen hi ah- pollutant episodes. However, this indicator may be neither highly
sensitive nor highly specific.
The clinical experimental exposures showed no symptomatic effects of 1 h of MTBE
exposure at 1.4 and 1.7 ppm, yielding 16 to 20 jig MTBE/L of blood. The strengths of the
clinical studies were in using controlled exposures, using a dose biomarker, and using both
subjective and objective measures of effect (a significant contrast to some of the
epidemiological studies). In addition, the findings were reproduced in two different exposure
studies with consistent results.
Despite all their strengths, both types of human studies have limitations. There were
design flaws in the epidemiological studies, difficulty with sensitivity and specificity of the
symptomatic indicators of effect, lack of simultaneous exposure measurements, the bias of
self-selection, and psychological and economic confounders. Limitations hi the clinical
studies include a lack of a dose range, a relative lack of an age range in comparison with the
community epidemiological studies, and lack of simultaneous co-exposures. The
epidemiological studies took place in complex work environments with people working hard
and exposed to other stresses and agents with similar symptomatic effects. The clinical
studies comprised a single brief exposure of 1 h to the agent hi question.
Chronic health effects of MTBE are unknown for humans. In animals, the pathologic
potency of MTBE is relatively low, with increased mortality and nephro-pathology only at
very high ranges (3,OCX) to 8,000 ppm) hi rats and mice. Testicular, liver, and kidney
tumors, as well as the other pathologic findings, seemed to reflect exacerbations of endemic
disease rather than the production of new types of cancer.
45
-------
In a variety of in vitro tests, MTBE is nongenotoxic and nonclastogenic. The RfC for
continuous human exposure is 3 /xg/m3. The limit value for the workplace exposure is
currently being modified. Human carcinogenicity at this point is indeterminant, although an
evaluation is in process.
Other fuel oxygenates (e.g., TAME) are under development. Their lower vapor
pressure values may bring the benefit of lower exposures to tank evaporations and
evaporations in the course of pumping gasoline.
Conclusions
1. The association of reported symptomatology with MTBE use in the community was
documented in some epidemiological studies but not in others reported at this meeting.
A precise exposure-response relation sufficient to invoke causality was not demonstrated.
2. Methyl tertiary butyl ether does not cause symptoms or objective changes at the dose and
durations tested in the human experimental studies. The challenging task facing the
assessment of risk from MTBE exposure is how to resolve the differences among the
several epidemiological studies and between the clinical and the epidemiological findings.
Although not conclusive, the data bear sufficient validity to accept both sets of findings at
this point and consider how to come to terms with those conflicts.
3. Co-factors may influence the symptomatic responses in the field studies, In Alaska, the
economic cost of increased fuel prices, the negative attitude toward government
intervention, and general anxiety in the community, suggest that significant co-factors
influenced the response in some of those field studies.
4. Methods are available to measure exposure, dose, and response in systematic studies.
Any future studies should be required to incorporate those methodologies for assessment
of these exposures.
5. Chronic health effects have not been demonstrated by the data. From a synthesis of the
animal data and the kinds of findings and lack of findings in the acute exposures, one can
reasonably conclude that chronic effects are unlikely at ambient exposures.
6. Future initiatives, to assess health risks of alternate oxygenates or other additives, are to
pursue preventive strategies by using the understanding and methodology that now exist.
Unanswered Questions
1. What accounts for the differences in the conclusions between the clinical tests and the
epidemiological studies?
2. What is the human dose response?
3. What are the net health benefits of the oxygenates?
4. Are there susceptibility factors such as age, working conditions, work stress, etc., to
explain the difference in the epidemiological studies and the clinical studies?
5. What are the risks of other oxygenates?
6. What are the mechanisms of the noncarcinogenic, biological, and health effects?
7. Are the emissions and ambient data adequate?
8. What is the influence of psychological factors on health effects in humans?
9. What is the effect of mixed exposures on the response to MTBE?
46
-------
10. Finally, what is the comparative risk of MTBE and CO at the measured levels of
exposure?
Discussion
A member of the audience from the State of Colorado stated that, Colorado believes that
oxygenated gasolines are a very effective method of reducing CO pollution. This last winter,
the Colorado Department of Health believes that CO tailpipe emissions were reduced by
approximately 25%, which translates into a reduction of approximately 5 days this last winter
that the City of Denver did not violate the CO standard. It is believed that this one program
gave an incremental benefit such that 50% of CO exceedance days were prevented. It also is
believed that CO is a health threat, even at 9 ppm, for some groups of people, and that there
is a health benefit even going below 9 ppm. Although Colorado advocates the use of a
variety of clean gasolines, MTBE is just one of those gasolines. It is felt that giving the
consumer a choice of fuel diversification played a large role in the major public acceptance
of the Colorado program. Gasoline is composed of hundreds of different components, 90%
of which have not undergone the same scrutiny as MTBE; and it is suggested that not only
MTBE be subject to this battery of testing, but there are many other components in gasoline
that need to be subjected to the same level of scrutiny.
Another speaker noted that the big issue in Fairbanks is the cold temperature effects
and that none of the experimental health studies were done at the temperatures of Fairbanks.
Cold start emissions are a large source of CO; during the first 1 to 2 min of the operation,
CO can be as much as 8 to 10%. This indicates that a lot of uncombusted fuel is getting
through. The question is how much MTBE gets through in those first 1 to 2 min in
Fairbanks when people start their cars. A calculation assuming an air to fuel ratio of about
four to one, which is reasonable for the first 1 or 2 min of operation during a cold start, and
assuming that none of the MTBE was combusted, indicates there would be 5,000 ppm
coming out of the tailpipe. This raises the question of what would be the effects on
somebody at a gas station breathing 5,000 ppm for maybe even two or three breaths at
—40 °F, especially if they are breathing 8% CO at the same time. Also, what are the
toxicity and cancer effects of just fuel alone compared to that of MTBE?
In response to the last question, a speaker noted that unleaded gasoline has been tested
in similar studies in rats and mice, and the rat kidney tumor and the mouse liver tumor are
issues in exposures to unleaded gasoline also. Furthermore, it was noted that the Chemical
Industries Institute of Toxicology has done a lot of work on those findings and has come up
with a lot of information on the cancer mechanisms. In addition, there has been an extensive
10-year sequence of animal studies and several large epidemiology studies of both
distribution workers and refinery workers associated with the manufacture of gasoline. The
carcinogenic risk was considered essentially minimal, with the exception of aromatics and,
specifically, benzene. This speaker also asserted that it was premature to draw a conclusion
that MTBE is associated with symptoms at high doses based on the epidemiologic findings,
for although there is some suggestive evidence of effects in a few studies, the MTBE
biomarker could be equally correlated with total hydrocarbon exposures associated with
sitting next to vehicles or refueling vehicles.
47
-------
It was noted that the negative findings of the New Jersey study should be viewed in
light of the difficulty of assuring an effective partition of automobiles by north and south due
to the north/south traffic in New Jersey. On the other hand, air and personal monitor
samples were collected and a subset of those people who had the highest exposure were
compared to the people who had the lowest exposure, with no statistical difference in
symptoms.
One commentator encouraged the EPA, perhaps in conjunction with the National
Institute for Occupational Saftey and Health, to establish a round-robin testing program for
MTBE and other oxygenated fuels, because of the analytical problems associated with MTBE
or other oxygenated fuels in complex hydrocarbon environments. It also was recommended
that concurrent monitoring be available for additional aromatics, such as benzene, when these
studies take place.
A speaker noted that comparative risk assessments are needed to develop intelligent,
useful, and good public policies. The information provided at this meeting was helpful, but
there are many important areas that were not addressed and therefore cannot be used to draw
conclusions about the public policies related to use of MTBE. For example, little was
presented on the potential increase hi emissions of aldehydes from the use of MTBE, which
could be related to some of the symptoms seen in Alaska. The possibility that aldehydes are
increased might provide the kind of exposure that produces a secondary effect from the use
of MTBE as opposed to an effect of MTBE itself. Also, there were very few data on the
potential levels of exposure hi the microenvironment of the motor vehicle itself. Previously,
studies on CO in Anchorage, where individuals drove through traffic for short periods of
time, showed excursions of CO levels 50 times higher within the vehicle than in ambient air
at an intersection. So, very high levels of CO could be transiently experienced.
In response to a question about the regulatory use of the RfC and its relationship to the
recommended TLV, an EPA representative responded that EPA believes that exposure of
people for a lifetime to MTBE at the level of the RfC will not have a significant risk of
noncancer adverse effects. This form of a dose-response assessment needs to be interpreted
as part of a risk characterization, which hi turn is used to create policy. In the risk
assessment of MTBE, the RfC is used to characterize the risks of chronic noncancer health
effects, but is interpreted in light of an exposure assessment. The RfC is very different from
the TLV because TLVs are intended to protect workers specifically, whereas RfCs are
intended to protect the entire population. The primary difference is the TLVs deal with a
40-h work week, whereas RfCs deal with a continuous 24-h/day, 7-day/week community
exposure. The RfC is also intended to protect sensitive subpopulations and adds another
uncertainty factor to account for those populations.
Another speaker noted a need for vehicle test protocols appropriate to local conditions—
not just with respect to temperature, but also in terms of the way the vehicles are actually
used and driven. Also, from a public health standpoint, no one has yet addressed the effects
on the very young, the very old or, hi some way, the very sick, hi addition to very cold
temperatures and some of the other climatic differences that occur in Alaska. Such
susceptibility factors remain an unknown at this tune.
48
-------
A commentator noted that EPA's Total Exposure Assessment Methodology (TEAM)
studies routinely found that aldehyde levels—formaldehyde and acetaldehyde—are higher
indoors than outdoors, as is benzene. This raises the question of why carboxyhemoglobin
levels are higher in some residents of Fairbanks and the possibility there is an indoor source
of CO, such as kerosene heaters or attached garages. Another speaker added that 40 ppm
for the CO TLV is a recommended value that is still at the committee level; it is not an
adopted value.
Some discussion was devoted to Dr. Green's conclusion about a correlation of symptom
reports with the high exposure groups in occupational settings. A CDC scientist addressed
this issue by noting that "therer was a."small"association7 between being in the upper quartile
[of blood MTBE levels]..... and being more likely to report symptoms."
It also was noted that, starting in 1995, reformulated gasoline will be required in
O3 nonattainment areas, and there is a congressional mandate to have an oxygen content of
2.0% in that gasoline. So the issue of oxygenates will not go away.
7A subsequent published report of the CDC/Alaska study by Moolenaar et al. (1994) states that there was "a
relationship between the highest quartile of blood MTBE concentrations and key health complaints on the day of
testing, but this finding was not statistically significant." A similar analysis of data from workers in the
CDC/Stamford, CT, study indicated a statistically significant association.
49
-------
APPENDIX A
SESSION ONE: SPEAKER ABSTRACTS AND PRESENTATIONS
A-1
-------
AUTHOR(S): William Piei (ARCO Chemicals)
TFTLE: HISTORICAL PERSPECTIVE ON USE OF ETHERS IN FUELS
NO ABSTRACT SUBMITTED
A-2
-------
OVERVIEW OF
FUEL OXYGENATE DEVELOPMENT
BY
WILLIAM J. PIEL
ARCO CHEMICAL COMPANY
PRESENTED AT
WORKSHOP ON MTBE AND OTHER OXYGENATES
AN UPDATE OF HEALTH RISK RESEARCH (
FALLS CHURCH, VA.
JULY 26-28, 1993
-------
>
•tfc
OVERVIEW OF FUEL OXYGENATE DEVELOPMENT
ALCOHOLS AS LOW COST GASOLINE EXTENDERS
ETHERS AS LEAD-OCTANE REPLACEMENT
WORLD-WIDE MTBE PRODUCTION GROWTH
PRE-CAA OXYGENATED FUELS PROGRAMS
MTBE IN EUROPEAN GASOLINES
VEHICLE EMISSION BENEFITS
ARCO CHEMICAL COMPANY, JULY 1993
-------
FUEL OXYGENATE DEVELOPMENT
ORIGINALLY DEVELOPED AS ALTERNATIVE TO S30/BBL CRUDE
T-BUTYL ALCOHOL COMMERCIALIZED IN 1969
SOLD AS GASOLINE COMPONENT (GTBA)
> US GOVERNMENT SUBSIDY FOR 10% BIO-ETHANOL IN 1978
a\
METHANOL / COSOLVENT BLENDS COMMERCIALIZED IN 1981
MANY ALCOHOL BLENDS DISCONTINUED IN 1986 WITH $15 CRUDE
MTBE EXPANDED WITH U.S. LEAD PHASEDOWN IN 1985-87
ARCO CHEMICAL COMPANY, JULY 1993
-------
WHAT IS MTBE?
METHYL TERTIARY BUTYL ETHER - OCTANE ENHANCER
HI-OCTANE ALTERNATIVE TO LEAD AND AROMATICS
MTBE FIRST COMMERCIALLY USED IN EUROPE (1973)
US PRODUCTION IN 1979 AFTER EPA WAIVER APPROVAL
> ORIGINALLY ALLOWED AT 7 VOL % IN GASOLINE
LIMIT RAISED TO 11% IN 1981, THEN 15% IN 1988
EUROPE SET EEC STANDARD OF 15% IN 1985
RECENT INTEREST HAS BEEN ENVIRONMENTAL
LARGEST BENEFIT IS REDUCING "CO" EMISSIONS
HELPS REDUCE MOBILE SOURCE VOCs, NOX, TOXICS £ PM-10
DILUTES GASOLINE SULFUR CONTENT FOR LOWER SOX
ARCO CHEMICAL COMPANY, JULY 1993
-------
MTBE AND OTHER ETHER PRODUCTION
MADE BY COMBINING ISOBUTYLENE AND METHANOL
REACTION IS HIGHLY SELECTIVE, MINIMAL BY-PRODUCTS
DI-ISOBUTYLENE, TRI-ISOBUTYLENE & T-BUTYL ALCOHOL
THREE MAIN SOURCES OF ISOBUTYLENE
BY-PRODUCT OF REFINERY FLUID CATALYTIC CRACKERS
5 ISOBUTANE DEHYDROGENATION
DEHYDRATION OF T-BUYTL ALCOHOL FROM ISOBUTANE OXIDATION
ETBE : ETHER ALTERNATIVE TO MTBE
COMBINE BIO-ETHANOL AND ISOBUTYLENE
TAME: COMMERCIAL ETHER PRODUCED IN REFINERIES
i
COMBINE REFINERY ISOAMYLENE WITH METHANOL
ARCO CHEMICAL COMPANY, JULY 1993
-------
r REFINERY MTBE PRODUCTION
ISOBUTYLENE SELECTIVELY REACTED FROM C4 HYDROCARBON MIXTURE
UNREACTANTS TO
MIXED C4's FROM
>
GO
FLUID CATALYTIC CRACKER
METHANOL
REACTION
&
SEPARATION
GASOLINE PROCESSING
OR GASOLINE BLENDING
MTBE TO
GASOLINE BLENDING
ARCO CHEMICAL COMPANY, JULY 1993
-------
BUTANE BASED MTBE PRODUCTION
FOR COMMERCIAL SALES
ISOBUTANE
>
-------
WORLD MTBE EXPANDED GREATLY SINCE MID 1980'S
NUMBER OF
PLANTS
YEAR END CAPACITY
M BBLS/DAY MM GAL/YR
1980
19
290
1985
19
49
750
1990
78
232
3500
1992
117
367
5600
1995
ESTIMATED
186
600
9200
ARCO CHEMICAL COMPANY, JULY 1993
-------
YEARS OF EXPERIENCE WITH OXYGENATED FUEL PROGRAMS
WINTER-TIME MINIMUM OXYGEN STANDARDS - WT% OXYGEN *
WINTER
SEASON
87/88
88/89
89/90
90/91
91/92
92/93
UULUI-IMUU
1.5
2.0
2.6
2.6
2.6
2.7
ARIZONA
PHOENIX TUCSON
2.3
2.3 1.8
2.7 1.8
2.7 1.8
NEVAD
LAS VEGAS
2.5
2.6
2,7
2.7
(A
REN
2.0
2.0
2.7
2.7
2.7% OXYGEN EQUAL TO 15 VOL%MTBE
ARCO CHEMICAL COMPANY, JULY 1993
-------
MTBE USE IN EUROPEAN GASOLINES
"CITY" GASOLINE INTRODUCED IN FINLAND
STARTED MAY 1991 BY NESTE OY
2 WT % OXYGEN USING 11 % MTBE
REPRESENTS 80% OF GASOLINE IN FINLAND
£
ro
MTBE IN MOST PREMIUM GASOLINES THROUGHOUT EUROPE
OCTANE ALTERNATIVE TO LEAD & AROMATICS
ONLY 15% OF CARS HAVE CATALYST & VAPOR RECOVERY
CAR EMISSIONS 5 TO 10 TIMES GREASER THAN U.S. CARS
ABSOLUTE EMISSIONS REDUCTIONS EVEN GREATER WITH MTBE
ARCO CHEMICAL COMPANY, JULY 1993
-------
MTBE'S NET IMPACT ON VEHICLE EMISSIONS
FTP EMISSIONS
BASE
. 57
6.7
7.7
4.2
3.1
+ MTBE
47
5.7
9.2
3.7
2.8
% CHANGE
(17)
(6)
(3)
(15)
(18)
CO
VOCs
NOx
SOx
TOXICS (mg/mile)
BENZENE
BUTADIENE
FORMALDEHYDE
ACETALDEHYDE
POM (PM-10)
* ERA'S COMPLEX PREDICTIVE MODEL FOR REFINERY BLENDING OF 15% MTBE
ARCO CHEMICAL COMPANY, JULY 1993
-------
SUMMARY
MTBE AND BIO-ETHANOL ARE TODAY'S MAIN OXYGEN OPTIONS
FUTURE OPTIONS LIKELY INCLUDE TAME & ETBE
ONE OF THE MOST WIDELY PRODUCED PETROCHEMICALS
MTBE NOW IN COMMERCIAL USE FOR 20 YEARS
OXY FUEL PROGRAMS DEMONSTRATED FOR 3 YEARS PRIOR TO CAA
USE OF MTBE HELPS REDUCE MAJOR AIR POLLUTANTS
CONTRIBUTES TO REDUCTIONS IN SIX CRITERIA POLLUTANTS
i
LEAD, CO, OZONE, NOx, SOx & PARTICULATES
ARCO CHEMICAL COMPANY, JULY 1993
-------
APPENDIX B
SESSION TWO: SPEAKER ABSTRACTS AND PRESENTATIONS
B-1
-------
AUTHOR(S): Kenneth T. Knapp
TTTLE: COLD TEMPERATURE MTBE DYNAMOMETER STUDY
ABSTRACT
Several research questions have emerged while implementing the wintertime oxygenated
fuel program in Alaska. These questions are: (1) does the cold temperature increase CO
emissions; (2) does MTBE decrease CO emissions; and (3) is MTBE safe? This paper addresses
the first two questions.
Little or no data existed on the effects of temperature and MTBE on CO emissions at very
low, or subzero, temperatures. The Mobile Sources Emissions Research Branch (MSERB) of the
Atmospheric Research and Exposure Assessment Laboratory (AREAL) had studied effects at
temperatures down to 40°F. To assist in evaluating these problems, MSERB ran Federal Test
Procedures (FTP) on a 1984 Buick Century V-6 carburetored car with two fuels: (1) a base fuel
and (2) a base fuel with 15% MTBE added (the vapor pressure and octane of the MTBE fuel were
adjusted to that of the base fuel) at temperatures of 75, 40, 20, and 0°F. This vehicle had been
tested with other MTBE fuels at temperatures down to 40°F. The results of these tests indicated
that there was a major increase in CO emissions when temperatures decreased: MTBE reduced
the increases of CO emissions. However, at 0°F, the CO emissions with MTBE fuel were still 3.5
times higher than that at 75°F for either fuels. Fuels with lower levels of MTBE gave smaller
decreases in CO emissions. Thus, the decrease in CO emissions appears to be MTBE-
concentration-dependent.
The next phase of the study was to develop a test program for testing a fleet of vehicles
at low temperatures. A protocol was developed including a fleet of test vehicles that is similar
to the vehicle distribution in Fairbanks, Alaska. The fleet consisted of 10 vehicles: five cars and
five light duty trucks (including a four-by-four truck). The vehicles are currently being tested at
75,20, 0, and -20°F, with the FTP in the MSERB cold cell dynamometer and in the Environment
Canada cold cell. Half of the vehicles are fueled with carburetors and half with fuel injection.
Two of the vehicles are identical, one pair is fueled by carburetors, and the other pair by fuel
injection. One set is being tested at MSERB and the other set is being tested at Environment
Canada. Two fuels are being tested, one base fuel having specifications of those used in
Fairbanks, and the second with about 15% MTBE added. The MTBE content turned out to be
only 12.5%. The vapor pressure and octane were adjusted for the MTBE fuel to match that of
the base fuel. MSERB provided Environment Canada with the two fuels. Emissions analyses
included regulated emissions, CO, NOX, HC, and CO2, both real time and integrated, and
speciated organics including over 250 HC, aldehydes, ketones, and MTBE. Both real-time
monitors and FTIR are being used. Temperatures of the ceil and vehicle oil, coolant, and catalyst
are also being continuously measured during the FTP.
Results to date indicate there is a strong dependence of CO emissions on temperature,
with the CO emissions increasing when temperatures decrease. With some vehicles, there are
advantages in using MTBE while others show no differences. The data show that the change
in CO emissions are dependent on several other factors including fuel composition, full vapor
pressure, and vehicle type. These results of multi-dependency agree with those found in the
Auto/Oil Air Quality improvement Research Program sponsored by the auto and oil industries.
B-2
-------
o
cc
111
cc
UJ
UJ
o
cc
UJ
m
O
o
ID
cc
LU
LU
S
O
<
o
LU
CD
LU
CC
I
CC
LU
0.
LU
O
O
I
I
a
B-3
-------
TEST CONDITIONS
00
STUDY TEST VEHICLES
Phase I Buick Century
FUELS TEMPERATURES
°F
Exxon 87 Base 75,40,20,0
Exxon 87 MTBE 75,40,20,0
Sun Base 75,40
Sun MTBE 75,40
TEST CYCLES
FTP, I/M240
FTP, I/M240
FTP
FTP
Phase II
Fleet
Alaska II Base 75,20,0,-20
Alaska II MTBE 75,20,0,-20
FTP, I/M240
FTP, I/M240
-------
Table I
CD
cn
Proposed Low Temperature MTBE Test Fleet
Year Make
Model Engine
Type Size
Fueline Svs
Drive Dyno Set
M HP
84 Buick Century V-6 3.0L Carb 2V FWD 3280 11.9
86 Chey Monte C V-8 5.0L Carb 4V RWD 3750 9.6
87 Chev Caprice V-8 5.0L Carb 4V RWD 4000 9.7
92 Chev Corsica V-6 3.1L MPFI FWD 3250 6.5
92 Chev Corsica V-6 3.1L MPFI
90-2 Toyota Corolla 1-4 1.6L MPFI
91 Dodge Pick-up V-6 3.9L TBI
87 Chev* Blazer V-6 2.8L TBI
FWD 3250 6.5
FWD
4x4
2750 8.3
RWD 4750 16.5
3875 12.0
84 Chev* S-10 LDT V-6 4.3L Carb 2V
RWD 3500 11.7
84 Mazd* pick-up 1-4 2.2L Carb 2V RWD 3000 11.4
-------
60
00
d)
50
Sun Base Sun 9.5% MTBE EX87-Base EX87-14.4% MTBE
- 20
- 10
0
40 20
Test Temperature, Deg F
1984 Buick Century operated on fuels with varying levels of MTBE
0
-------
FUEL SPECIFICATIONS
FUELS
Ex87 Base Ex87 M Sun B Sun M Alaska II B
03
OCTANE
RVP
T90
COMP vol 9
MTBE
Benzer
Aroma
Olefini
Paraffi
?
e
tics
ns
87
13.3
343
l-.l
1.0
31.1
12.5
53.1
87
10.8
367
14.4
0.9
30.1
10.1
41.4
87.8
9.0
346
0
1.4
38.0
9.5
51.4
88.8
8.9
294
9.5
1.5
23.8
11.3
54.6
87.0
14.5
321
0
1.4
25.6
15.4
57.1
Alaska II M
88.6
15.5
298
12.1
0.9
19.7
9.6
59.8
-------
MEASUREMENTS IN MTBE STUDY
TEMPERATURES
Vehicle Oil
Vehicle Coolant
Dynamometer Cell
Catalyst, before and after
CHEMICAL ANALYSES
250 components (benzene, 1,3-butadiene, etc.)
Carbonyls -12+ (formaldehyde, etc.)
Oxygenates -10+ (MTBE, mcthanol, etc.)
-------
ep
CD
CO, g/mi GO, g/mi
Base Fuel 12.1% MTBE Fuel
H — A —
75 70 65 60 55. 50 45 40 35 30 25 20 15 10 5 0
Test Temperature, Deg F
1984 Buick Century
-------
40
CD
*_J.
o
30
0)
*
oc
.1 20
CO
.S2
LLI
CO, g/mi CO, g/mi
Base Fuel 12.1% MTBE Fuel
- 10
0
75 70 65 60 55 50 45 40 35 30 25 20 15 10 5 0
Test Temperature, Deg F
1992 Chevrolet Corsica
-------
50
40
Q)
c
30
uu
20
10
CO, g/mi CO, g/mi
Base Fuel 12.1% MTBE Fuel
—a— — A—
D
D>
50
40
30
20
10
80 75 70 65 60 55 50 45 40 35 30 25 20 15 10 5 0 (5)
Test Temperature, Deg F
1992 Chevrolet Corsica
o
-------
CO
a_*
IS3
75 70 65 60 55 50 45 40 35 30 25 2015 10 5
Test Temperature, Deg F
1992 Chevrolet Corsica
- 500
- 400
Bag 1 Bag 1
Base Fuel 12.1% MTBE Fuel
— —A—
600
- 300
- 200
- 100
0
-------
50
CD
I
CO
40
Bag 2 Bag 2 Bag 3 Bag 3
Base Fuel 12.1% MTBE Fuel Base Fuel 12.1% MTBE Fuel
B ---£,--. 0
0
75 70 65 60 55 50 45 40 35 30 25 20 15 10 5 0
Test Temperature, Deg F
1992 Chevrolet Corsica
-------
Buick Bag 1, 0 Degrees F
O
8
CO
T3
C
o
SL
Time (505 Sees.)
B-14
-------
Buick Bag 2, 0 Degrees F
3000
O
O
2500-
2000-
1500-
1000-
500-
0
-500
IJU
Base
MTBE
Time (872 Sees.)
B-15
-------
Corsica Bag 2, 0 Degrees F
900
O
Time (872 Sees.)
B-16
-------
600
03
500
w
E 400
03
D)
CC
c 300
.9
'v>
V)
LU
O 200
100
Bag1
Base Fuel
B
Bag 2
12.1% MTBE Fuel
*--
Bag1
12.1% MTBE Fuel
---A---
Bag 3
Base Fuel
Bag 2
Base Fuel
Bag 3
12.1% MTBE Fuel
600
500
400
300
200
100
•MUJMUMKUMLUMUMJMti
75 70 65 60 55 50 45 40 35 30 25 20 15 10 5 0
Test Temperature, Deg F
1991 Dodge Truck
Q
-------
CD
—k
oo
50
40
0)
c
g
'to
to
111
30
20
10
0
CO, g/mi CO, g/mi
Base Fuel 12.1% MTBE Fuel
—a—• —A---
i i
I J
80 75 70 65 • 60 55 50 45 40 35 30 25
Test Temperature, Deg F
Dodge/Buick/Corsica/Monte Carlo
A
A
50
40
30
20
10
20 15 10 5 0 (5)
0
-------
CD
30
25
20
0)
75
DC
§15
V)
to
111
10
0
MTBE, mg/mi MTBE, mg/mi
Base Fuel 12.1% MTBE Fuel
B • —A —
I I I
A
D
80 75 70 65 60 55
1992 Chevrolet Corsica
50 45 40 35 30 25
Test Temperature, Deg F
A
J I i
30
25
20
15
10
20 15 10 5 0 (5)
0
-------
oo
t\>
o
20
I15
o
X
cc
c
o
'w
I 10
LU
0)
0)
2
(0
E
o 5
LL.
.0.-
Corsica Corsica Dodge Dodge
Base Fuel 12.1% MTBE Fuel Base Fuel 12.1% MTBE Fuel
- B -
---A--
*
<-"*O
o
J_
80 75 70 65 60 55 50 45 40 35 30 25
Test Temperature, Deg F
Dodge and Corsica Formaldehyde Emissions
•*
o
A...
20
15
10
20 '15 10 5 0 (5)
0
-------
CD
rb
CONCLUSIONS
1. CO emissions increase with a decrease in operating temperature.
2. The CO increase with temperature is less with MTBE in the fuel
under certain conditions.
3. The emissions data spread increases with a decrease in
temperature.
4. The data suggest additional variables affect the CO emissions.
-------
AUTHOR(S): Chandra B. Prakash
TITLE: EMISSIONS UNDER COLD CONDITIONS
(1) Test fuels: Baseline gasoline
15% MTBE (oxygenated gasoline)
(2) Test vehicles: 1992 Chevrolet Corsica
1987 Chevrolet Caprice
1991 Toyota Corolla
Fuels and vehicles were provided by EPA-RTP, except for the Toyota Corolla, which came
from the Environmental Canada fleet.
(3) Test temperature range: -20°F to 75°F
FUEL EFFECT
The emissions results were analyzed using the student t-test to determine 95% confidence
intervals. The results indicate that there was no statistically significant difference in regulated
emissions or carbon dioxide (CO2) emissions between fuels over the ambient temperature range
of 75°F to -20°F.
The difference in formaldehyde and total carbonyl (aldehyde and ketone) emissions for
the two fuels were also found to be statistically insignificant. However, vehicle-to-vehicle
differences were noticeable. The Corsica and Caprice gave higher formaldehyde and carbonyl
emissions compared to baseline gasoline, while the trend was reversed with the Corolla. This
finding is based on two runs at each fuel-temperature combination over the 20°F to -20°F
temperature range and a single run at 75°F for either fuel.
It is generally recognized that the use of oxygenated gasolines compared to base
gasoline reduces CO emissions, while it increases NOX and formaldehyde emissions. However,
the present data on two test vehicles as well as the recent emissions test results at Environment
Canada on eight new technology vehicles (3-way catalyst with a closed loop system) using
oxygenated and baseline gasolines suggest that the fuel effect on these vehicles is rather small
and is, in most cases, insignificant.
Therefore, the wintertime oxygenated fuels program should be expected to offer
diminishing CO benefits (reductions) as older vehicles are gradually replaced with new
technology vehicles.
The MTBE in the vehicle exhaust, while using MTBE containing fuel, was present only in
bag 1 and ranged between 0.3 to 0.9 ppm.
B-22
-------
TEMPERATURE EFFECT
Regulated emissions increase with decreasing temperature. The increase in HC and CO
emissions can be attributed to a reduction in combustion and catalyst efficiency. Immediately
following start-up, particularly at lower ambient temperatures, the engine operates in a fuel-rich
mode so as to sustain combustion. Excess fuel is required to ensure that sufficient fuel remains
vaporized within the cylinder during intake and compressions strokes. These conditions result
in higher HC and CO emissions, in addition, the engine out HC and CO are not converted
efficiently in the catalyst. The catalyst efficiency remains poor immediately following engine start-
up until the catalyst reaches its operating temperature ("light-off" point).
Nitrogen oxide emissions are a thermal phenomena that would not be expected to be
affected by a reduction in ambient operating temperatures. Rather, the increased NOX emissions
trend can be explained by the overall reduction in catalyst conversion efficiency. Due to the low
catalyst temperature and low A/F ratio (rich) following engine start-up, the NOX reduction
capability of the catalyst is extremely limited. Of interest is the apparent plateau in NOX emission
from the Corsica model at temperatures of 20°F or less. This may be the result of the engine
electronic control system.
Carbon dioxide emissions are vehicle specific and can be attributed to the increased fuel
consumption due to reduced combustion efficiency and increased engine/vehicle friction until the
vehicle has warmed up.
The formaldehyde and carbonyl emissions followed the general trend of HC emissions
and increased with decreasing ambient temperature. Typically, formaldehyde emissions at -20°F
were about twice as compared to emissions at 75°F.
GENERAL OBSERVATIONS
The Caprice (equipped with a carburetored fueling system) showed an unusual behavior
by giving higher CO and HC, while lower NOX emissions with MTBE fuel compared to baseline
gasoline. Although the differences in exhaust emissions between the two fuels were not
statistically significant, it should be mentioned that this car also gave very high evaporative
emissions during tests at EPA. It is likely that the abnormal behavior of the Caprice gave a much
larger variation {n exhaust emissions and the unusual trend of CO emissions.
The data supports the general belief that vehicles and their control systems are designed
for optimum emissions performance at certification conditions, and, in most cases, the emissions
are much higher during off-cycle and low ambient temperature conditions. Finally, the data also
indicate that there is considerable variation in regulated emissions from one vehicle to another.
B-23
-------
CD
rb
16-
1?
.e
.8
o
o
-,*,
-f-
Onsoline
-20 -10 0 10 20 30 40 50 GO 70 00
Ambient Temperature, deg F
500
'-20 -10 0 10 20 30 40 50 60 70 80
Ambient Temperature, deg F
1.4
c* 1
c
|O.R
o OR
0.2
-20 -10 0 10 20 30 40 50 60 70 80
Ambient Temperature, deg F
»-
-20 -10 0 10 20 30 40 50 60 70 80
Ambient Temperature, deg F
Composite emissions as a function of temperature for Corsica using FTP drive cycle
-------
tv
GO
0> 50
w 40
0
•jf 30
OJ
O 20
10
n
[f-20F]|
• 1B43
- 19.78
( T-OF |
• 13.73
i?m
[T«20F |
|- 7 16 726
\W#]
3.41 "J°
WIBE Base»ine MTBE Baseline MfBE Bnseline MTBE BasclinB
Fuel
25
6
•£
W)
111 3
O
P 2
t
0
r=-20F 1
i
-1.53
•1.59
IT-OF |
0.70
O.B9
JT.20F|
0.41
047
[WSF]
» V
Fuel
Fuel effect on composite emissions for Corsica, showing 95% confidence intervals
-------
•20 -10 0 10 20 30 40 50 GO 70 80
Ambient Temperature, deg F
-20 -10 0 10 20 30 40 50 60 70 80
Ambient Temperature, deg F
CD
&
-20 -10 0 10 20 30 40 50 60 70 00
Ambfent Temperature, deg F
"•20 -10 0 10 20 30 40 50 60 70 80
Ambient Temperature. d«j F
Composite emissions as a function of temperature for Caprice using FTP drive cycle
-------
1W
140-
120
t>> too-
to
1 8°~
t
UJ GO-
CD
40
20
n-
(ES
4095
?T>.42
TeOFJ
25.54
10.78
£S]
1707 Ltd
1
T.75FJ
201 3.7G
MTBE Baseline MTBE Baseline MIBE Basoline MfBE B.TSplino
Fuel
&•
4-
t».
•Jssions, g
CJ
r
z
1
o-
[T.-20F |
I 184
2.26
IT-OF
1.05
2.01
|T.20F|
.1.59
.1.75
[T.75F|
1.12
088
MtBE Baseftne MTBE Baseline MTBE Baseline MTBE Baseline
Fuel
DO
tooo-
E 000
O)
Emissions,
§
g 400
O
200-
[T.-20
-617
D
632
T=OF]
502
572
^
r f#M
u WO 1
1
[T~75F[
" -
MTBE Baseline MTBE Baseline MTBE Baseline MTBE Baseline
Fuel
20
0>
M 15
.Q
to
tu to-
5
o-
|U*J
1
334
2.54
|T-OF
I 1.63
132
lajj
|-'^1 Li 09
|T-75F|
:
0.80 000
MTBE Baseline MTBE Baseline MTBE Baseline MTBE Baseline
Fuel
Fuel effect on composite emissions for Caprice, showing 95% confidence intervals
-------
OJ
rb
CO
0.45
-20 -10 0 10 20 30 40 50 60 70 80
Ambient Temperature, deg F
„. | ( j__ _^ | j 1 ( |
"•20 -10 0 10 20 30 40 50 60 70 80
Ambient Temperature, deg F
0.35
.5
LU
6
0.25
0
-10 0 10 20 30 40 50 60 70 80
Ambient Temperature, deg F
-20 -10 0 10 20 30 40 50 60 70 80
Ambient Temperature, deg F
Composite emissions as a function of temperature for Toyota using FTP drive cycle
-------
00
35-
30-
1 ° 8"
o
'05 0.6-
in
e
LU
604-
02
n-
j T = 20F j
- 0.30
- 0.43
|T=OF]
027
-029
l^f]
• 023
032
| T = 75F |
0_33
020
Baseline MTBE Baseline MTBE Baseline MTBE Baseline
Fuel
Q)
.2
LU
J
2.5-
2
1.5-
1
05-
j T= 20F |
0.83
• 1.06
E3
MTBE Baseline MT
1
o.ss
- 0.58
E=J
• 0.32
038
|T = 75F|
013 °_16
BE Baseline MTBE Baseline MTBE Baseline
Fuel
Fuel effect on composite emissions for Toyota, showing 95% confidence intervals
-------
03
8
16-
<*
12
.
.8
K
UJ
C)
O
0:
Oasolinn
-20 -10 0 10 20 30 40 50 GO 70 00
Ambient Temperature, deg F
500
'20 -10 0 10 20 30 40 50 GO 70 80
Ambient Temperature, deg F
niiii:i>ii•ivi
-20 -10 0 10 20 30 40 50 60 70 80
Ambient Temperature, deg F
1.4
•
.5
E
i/f
c
Q „ „
.12 UH
(-
HI n r
o OC
x
i—
0.4
.«•?
,
— r* —
/ \
/ \
. ./ ..x . .
' \
/ ! x
r \ „
^^^^J"^^*^fc****-«. ^ ^^*
-j 1 . , . . .
15% Ml DP
-«»-
'Basofrie
"20 -10 0 10 20 30 40 50 60 70 80
Ambient Temperature, dog F
Composite emissions as a function of temperature for Corsica using IM240 drive cycle
-------
50
CO
GJ
40
•S
t?
E-
c
-.2
o
o
Flnr-pli no
-20 -10 0 10 20 30 40 50 60 70 80
Ambient Temperature, deg F
680
GOO
9
E
a
ui
t\»
O
O
3GO.
440""— «•-;.•;;
-10 0 10 20 30 40 50 60 70 80
Ambient Temperature, deg F
-20 -10 0 10 20 30 40 50 60 70 80
Ambient Temperature, deg F
3.5
I,2-5
w
.Q 0
I LI
15% Ml BE
tki-ajtino
-10 0 10 20 30 40 50 60 70 80
Ambient Temperature, deg F
Composite emissions as a function of temperature for Caprice using IM240 drive cycle
-------
tp
6
ro
52 -M
\
O)
t 23
w .
.9 2«r
v>
v>
'£ _,
£ 2'
o>
"O 5O
>, 20
a)
X) in
o 10-
Caprice
Formaldehyde Emissions vs Temperature
*-'\
.-.<...... V
>*~ \
\
\
N
I V
"^^^-^
•
•0 -10 0 10 20 30 40 50 GO 70 8
Ambient Temperature, deg F
15% MTBE
Bar.pline
0
"°
V)
£
X) 40
Formaldehj
0 8
Caprice
95% Confidence Intervals - Formaldehyde
| T « -20F j
*» *«
JT -OF |
MfBE BaseDne M
•24.73
| T = 20F ]
22.50
-18.77
T-75F
BE Baseline MTBE Baseline MTBE Baseline
Fuel
Corsica
Formaldehyde Emissions vs Temperature
-20 -10 0 10 20 30 40 50 GO 70 80
Ambient Temperature, deg F
25-
5?
Emissions, rr
-------
Caprice
Carbonyl Emissions vs Temperaturn
20 -10 0 10 20 30 40 50 60 70 00
Ambient Temperature, deg F
100
I »
S 60
O
w
W 40
"* 20
O
0
Caprice
95% Confidence Intervals • Carbonyl
|T--20FJ
-
|T -OF |
>
|T-2or
T»75F
-
MTBE Baseline MTBE Baseline Ml' BE Baseline Ml BE Risplino
Fuel
0) \A
g
?„
\f
i'
O
"> «rv
(/> 10
111
">• &
C &
8
fTJ
4.
Corsica
Carbonyl Emissions vs Temperature
*.>„
^v
V
V
X
N
*x
V
^x
:::.^....
"v.
N. ^x
1
>0 -10 0 10 20 30 40 SO 60 70 8
Ambient Temperature, deg F
15% MTBE
Brepline
0
35
0)
E
c
o 20-
V)
V)
6 ir»
§ 10
o ^
Corsica
95% Confidence Intervals - Carbonyl
|T--20F|
; -
|T.OF|
-
T.20FJ
[T-75F]
-
MfBE aisolinn MTBE Baseline M1BE Baseline MIBE Baseline
Fuel
Temperature and fuel effect on composite carbonyl emissions for both Caprice and Corsica
-------
AUTHOR(S): Dave Veazey, Marcus Martin, Perry Klein, & Dr. Richard Benner
TITLE: CARBON MONOXIDE EMISSIONS IN FAIRBANKS, ALASKA
Vehicles at Fairbanks' wintertime temperatures were tested for exhaust emissions both
for cold start and on-road operation. For cold start emissions, engine pre-heating reduced CO
emissions by as- much as 70%, while oxygenated fuel had no detectable effect. On-road
emissions of CO were reduced by an average 28% from using oxygenated fuel. From the results,
a CO inventory was revised for Fairbanks, which estimated a total burden of 49.2 tons per day
(a 54% decrease from a previous inventory).
Cold start CO emissions appeared to be independent of temperature from +20°F to -40°F.
The amount of cold soak time was the primary factor responsible for increased emissions. Due
to the inherent chaotic behavior of such a transient mode, cold start emissions contained
variability that would not allow serious statistical analyses. One hundred and twenty of the 166
cold start tests conducted were instead combined into a normalized data set such that trends
in emissions behavior could be examined. The calculation method used to convert from %CO
to grams of CO was found to over-estimate emissions by 10-20%. Fortunately, the results are
based on the magnitude of the relative difference between tests on each particular vehicle, and
the validity of the results should not be affected. Not all cold start tests comparing pre-heating
and fuel type were conducted at identical temperatures. Clear fuel tests were conducted from
-40°F to +30°F, while oxygenated fuel tests were conducted from -18°F to +30°F. AS a result,
if the oxygenated fuels tests had significantly decreased emissions, it would have been difficult
to determine any effect from oxygenated fuels. Because there appeared to be no change in CO
emissions from either oxygenated fuels or ambient temperatures, it might be surmised that using
oxygenated fuels is not effective in reducing cold start emissions.
Four vehicles were tested for on-road emissions. As discussed previously, the calculation
from %CO to grams of CO introduces error that is not taken into account. This creates even
more error for on-road emissions as the volumetric exhaust flow may change considerably during
operation. It is the magnitude of the difference between clear and oxy fuels that is of interest.
Run-to-run repeatability is crucial to believing these results. The relative standard deviation
ranged from 5-40%. In all cases except one, CO emissions were reduced with oxygenated fuel.
Only one vehicle, the high emitter, gave results that indicated a statistical difference between
dear and oxygenated fuel. Although not all of the results were statistically significant, they were
used to estimate the reduction in CO emissions from oxygenated fuel use. Carbon monoxide
was reduced by 22% for highway driving and 32% for city driving.
A remote sensor was used to measure 462 vehicles over 2 days, ft was found that 14.5
tons, over 70% of the CO from on-road emissions, came from the dirtiest 10% of the Fairbanks
vehicle fleet. The total CO from on-road emissions was 20.3 tons/day.
Based on these results, and the fact that the prior CO inventory was done without benefit
of any vehicle measurements, a revised CO inventory estimated only the CO emissions that
directly affected attainment of the CO air quality limit. Our results conclude that 49.2 tons CO/day
are emitted into Fairbanks. This is a 54% decrease from the prior estimate of 105 tons/day. This
represents an upper limit to emissions in Fairbanks. A worst case scenario was used for cold
start emissions (with 40% of the vehicles being pre-heated) in which 475 grams/vehicle was
emitted by half the registered vehicles each day. All of the fuel purchased each day was
attributed to on-road emissions (none to cold starts), providing a high estimate for on-road
emissions. Although the numbers used for this inventory are speculative, it was felt that a CO
B-34
-------
inventory based on real numbers with some uncertainty would benefit policy-makers who were
at the time consulting a CO inventory which was done with no measurements at all. Although
this study was not conducted following federal guidelines and procedures, the assumptions
made and the conclusions drawn are reasonable, and further study should be initiated to
investigate our findings.
From our revised inventory, it was estimated that ambient CO would be reduced by 11 %
from an oxygenated fuels program. If drivers pre-heated their vehicles all the time, the reduction
would be estimated at 8.2 tons, or 17%. By fixing the dirtiest 10% of the vehicles on the road
a benefit of 27%, or 13.2 tons could be realized (this estimate changes gross polluters into
moderate polluters). If all vehicles were pre-heated and the dirtiest 10% of the vehicles were
repaired, there would be a 43% reduction in ambient CO. In the last 4 years, the highest
exceedance was 11.9 ppm. A 25% reduction in ambient CO would be required to reach
compliance. Neither an oxygenated fuels program or increasing the use of engine pre-heaters
will provide this reduction. Only fixing the dirtiest 10% of the vehicles or combining gross polluter
repairs with increased pre-heater use would accomplish this according to our results.
B-35
-------
FAIRBANKS CO EMISSIONS STUDY
1- COLD STARTS
2- ON-ROAD EMISSIONS
3- CO INVENTORY & CONTROL STRATEGIES
B-36
-------
FAIRBANKS 1990 CARBON MONOXIDE
INVENTORY
SIERRA RESEARCH, 1992
VEHICLE EMISSIONS 96 TONS
WOODBURNQNG 9 TONS
OTHER 13 TONS
TOTAL 105 TONS
*91% OF CO EMISSIONS ARE FROM VEHICLES
B-37
-------
co
£
COLD START EMISSIONS CO% VS, TIME
170 MIN COLD SOAK -40 F (MI-005)
o
o
2
1
0
0
100 200
300 400
elapsed time (sec)
500 600
700
%CO
-------
NORMALIZED CO EMISSIONS
AS A FUNCTION OF TEMPERATURE AND SOAK TIME
(COMBINED DATA SET)
NORMALIZED
CO EMISSIONS
AMBIENT
TEMPERATURS-io
( F) -20
-30
-40
200
400
50C
800
COLD SOAK TIME (MIN)
B-39
-------
03
CO EMISSIONS VS COLD SOAKTIME
NORMALIZED, COMBINED DATA SET
NORMALIZED CO EMISSIONS
-i .
i
OQ-
.y
OQ
.0
07 .
. /
Ofi-
.O
Oc.
.O
0.4-
00_
.0
00-
,fL
0.1-
Q_
SA A fTl
D A IB
an A
* A A A
A r-i A
•••[_] a _, ,-,
CA AA LJ a
Dr— 1 ., i—i . __ D
AirTVi^A1^. AO
rii^ tLj A
^7 CDA^A A A
A ' AAO ^
i^ J n
0 100 200 300 400 500 600 700 80
COLD SOAK TIME (MIN)
a REG
OXY
-------
07
CO EMISSIONS VS COLD SOAK TIME
PLUGGED IN ALL TEMPS
NORMALIZED CO EMISSIONS
4 j
1
OQ-
.y
Op.
.0
0-7.
,(
Oft-
.0
0^-
.0
OA-
.*>
00.
,
-------
<
O
100-
90-
LJ? 80-
I 70-
60-
50-
40-
30-
CO
z
o
55
CO
8 201
101
0
HIGHWAY
CLEAR OXY
CITY
CLEAR OXY
/ l^ift^yiiS&Haerf /
z
z
/HONDA
/TOYOTA
BRONCO
ESCORT
B-42
-------
MASS CO EMISSIONS (TONS/DAY)
ON-ROAD OPERATION
Q
CO
? 9^
6-
3-
O
o
CO
CO
10 20 30 40 50 60 70 80 90 100
PERCENTILE OF VEHICLE FLEET
B-43
-------
CO EMISSIONS DISTRIBUTION
FAIRBANKS VEHICLE FLEET
< 3000
20 30 40 50 60 70 80
PERCENTILE OF VEHICLE FLEET
90 100
clearfuel oxyfuel
-------
COLD START EMISSIONS INVENTORY
COLD START WORST CASE SCENARIO = 475 g/vehicle
48,140 VEHICLES = 25.2 tons/day _
COLD START CO EMISSIONS AT VARYING PLUG-IN RATES
0% PLUG IN
20% PLUG IN
40% PLUG IN
60% PLUG IN
80% PLUG IN
100% PLUG IN
COLD START CO
EMISSIONS (tons/day)
25.2
22.4
19.7.
17
14.2
11.5
The above data represents an upper limit to CO emissions from cold starts in the
FNSB.
B-45
-------
FAIRBANKS CO EMISSIONS ESTIMATES
CO LEVEL REQUIRED
FOR COMPLIANCE
1982 1992
B-46
-------
SUMMARY
COLD STARTS:
• COLD START EMISSIONS ARE DEPENDENT ON SOAK TIME, NOT
TEMPERATURE FROM +20 °F TO -4O °F.
CO EMISSIONS FROM COLD STARTS ARE NOT REDUCED WITH
OXYGENATED FUEL. _
ENGINE PRE-HEATING CAN REDUCE COLD START EMISSIONS BY
70%. ' , .
ON-ROAD EMISSIONS:
• ON-ROAD EMISSIONS OF CO ARE REDUCED BY 28% WITH
OXYGENATED FUEL.
CO INVENTORY:
• A REVISED CO BUDGET BSHWCATES THERE IS $g LESS CO THAN
PREVIOUSLY ESTIMATED.
B-47
-------
AUTHOR(S): Gerry Quay
TITLE: AIR QUALITY MONITORING'- OXYGENATED FUELS
The introduction of oxygenated fuels into the gasoline supply in Fairbanks, Alaska, during
October 1992 resulted in numerous health-related complaints. In response, Alaska's Department
of Environmental .Conservation's (ADEC) Ambient Analysis Group initiated a fast-track program
to procure, install, and operate a monitoring network capable of quantifying ambient
concentrations of hydrocarbon emissions, fuel additives, and aldehydes.
While experience has demonstrated that it is not uncommon to experience an influx of
public apprehension and concern over the introduction of oxygenated fuels (gasoline with an
additional 2.7% oxygen by weight - MTBE) into CO non-attainment areas, the number, type, and
apparent severity of complaints took the Departments of Environmental Conservation and Health
and Social Services by surprise. As the number of complaints increased, a request for
assistance was made by the State Epidemiologist to the National Centers for Disease Control
and Prevention (CDC) to help evaluate the severity and cause of the reported health effects.
Simultaneously, ADEC was in the process of identifying a source of hydrocarbon and aldehyde
monitoring instrumentation that could be made available to Alaska within 2 weeks. The Oregon
Graduate Institute (OGI) was selected to provide hydrocarbon sampling systems, and the U.S.
Environmental Protection Agency's (EPA) Region 10 and Desert Research Institute (DRI) supplied
the aldehyde monitors. Although the original target date of Thanksgiving was missed by a day
or two, the response of all agencies was exceptional.
The original monitoring project was designed to evaluate ambient exposures in Fairbanks,
Alaska. Monitoring sites were selected to survey locations in industrial, urban, residential, traffic
corridors, and background areas of the city. As coordination between the CDC and ADEC
improved, the state realized that an emphasis on ambient air monitoring would not support
CDC's objectives. In a last minute change, the monitoring project was redesigned to focus on
worker exposure. The majority of the monitoring effort was conducted in automotive garages and
fleet vehicles that routinely travel the Fairbanks area.
Monitoring was conducted in three phases. The first phase ran December 1 -15 and was
established to identify the level of hydrocarbon and aldehyde emissions in Fairbanks from the
use of oxygenated fuels. Air sampling for volatile organic hydrocarbons (VOC) was conducted
using OGI volatile organic hydrocarbon samplers. Monitoring and analytical protocols followed
Method 14 of EPA's Compendium of Methods for the Determination of Toxic Organics in Ambient
Air. Sampling for aldehydes was performed using Method 11 and portable aldehyde samplers.
After the decision was made to remove MTBE from the fuel supply, a second phase of monitoring
was conducted between December 17-22 (the phase-out period). Because no baseline data
existed on emission characteristics from fuel in Fairbanks, a third phase of monitoring was
conducted from February 17 through March 3. Phase 1 -3 monitoring dates and locations are
attached.
B-48
-------
ALJTHOR(S): Roy B. Zweidinger
TITLE: AIR QUALITY MEASUREMENTS IN FAIRBANKS, STAMFORD, AND ALBANY
The introduction of oxygenated fuels (15% MTBE) in Fairbanks, Alaska, in 1992 resulted
in many health complaints from consumers. The Governor of Alaska suspended sale of these
fuels on December 11, 1992. Air samples were collected prior to the suspension (phase 1),
during the phase-out period (phase 2), and after about 2 months after suspension, at which time
the MTBE fuels were expected to be at nominal levels (phase 3). For comparison, air samples
were also collected from Stamford, Connecticut, which also sold 15% MTBE oxygenated gasoline
(but had no consumer health complaints) and Albany, New York, where MTBE was only present
in gasoline at nominal levels to enhance octane (Table I). Fuel samples collected from gas
stations in Fairbanks during phase 2 and 3 indicated the average wt % MTBE in unleaded regular
gasoline decreased from 8.46% to 1.00%, while the average for premium gasoline decreased
from 14.66% to 5.56% (Table II).
A limited number of samples were collected from several types of microenvironments in
all three cities. The samples were 8-hour averages with start times varying between 7:30 am and
1:00 pm. Volatile organic compounds (VOCs), including MTBE, were collected in evacuated 6-L
stainless steel canisters that used either positive pressure pump-type samplers or controlled
vacuum-bleed samplers. Aldehydes were collected using 2,4-dinitrophenylhydrazine (DNPH),
coated C-18 (Alaska phase 1 only), or silica gel cartridges. The VOCs were analyzed by gas
chromatograph (GC) with flame ionization detection (FID), while the aldehydes were analyzed by
high-performance liquid chrofnatography (HPLC). These data are not suitable for short-term
exposure assessment. While the data provide approximate ranges of MTBE and other VOC
concentrations in the air, they do not accurately quantify air concentrations in cities over a
particular period. It should also be noted that ambient temperatures and other meteorological
conditions varied greatly among the cities.
For comparison purposes, the sampling sites were divided into seven groups (Table III):
auto traffic (ambient, near roadways); residential (ambient, not adjacent to a major roadway or
intersection); gas station (ambient, near pump island); background (ambient, outside city limits);
garage service area (indoor, cars, gasoline, and other solvents present); parking garage
(ambient, Stamford only); and indoor (inside offices and homes).
Figures 1 -4 are logarithm plots of the average concentration of MTBE, benzene, total non-
methane organic carbon (NMOC), and 1,1,1 -trichloroethane (methylchloroform) observed at sites
in the three cities. Figure 5 is a normal plot for formaldehyde. (Note: Concentrations are in
parts per billion carbon (ppbC), for all compounds except formaldehyde, which are in ppbV. (To
convert MTBE ppbC to ppbV, divide by 4 and not 5 to correct for FID detector response
calibrated to propane.) The highest average concentrations of MTBE (679 ppbC), benzene
(1181 ppbC), total NMOC (80.5 ppmC), and formaldehyde (30.8 ppbV) were found in garage
service bays. Concentrations of 1,1,1 -trichloroethane exceeded 14,000 ppbC (Fairbanks, phase
3). Aside from the service bays, MTBE concentrations were next highest at gas stations (149
ppbC Fairbanks, phase 2; decreasing to 22.6 ppbC, phase 3). The Stamford gas station MTBE
concentrations were the lowest (15 ppbC) but were likely the result of sampler location (Albany
average = 96.1 ppbC). Gas station sampling in Stamford was about 15 feet away from the
pumps, while sampling in the other cities was at the pump island. Benzene levels were higher
in Fairbanks (average roadside 49 ppbC, phase 2; 79 ppbC, phase 3) than in the other cities
(Stamford, 5.6 ppbC; Albany, 2.6 ppbC) and increased slightly with the reduction of MTBE in
gasoline.
B-49
-------
The indoor and outdoor (residential) MTBE concentrations averaged about 28 ppbC in
Fairbanks, with residential outdoor levels falling to 4.1 ppbC in phase 3, Similar concentrations
were also noted indoors, but the overall indoor average concentration was high because of one
home (Hamilton Acres) where concentrations of 80 ppbC were observed (figures 6 and 7). This
home had an attached garage and also had elevated levels of benzene (260 ppbC) and other
compounds associated with gasoline (figures 8 and 10). Formaldehyde concentrations were
higher indoors (10-28 ppbV) than outdoors (2-20 ppbV), which is generally the case, and
appeared typical of those seen in indoor air studies (Figure 9).
Figures 10-12 are simulated chromatograms for several indoor versus ambient samples
that were collected concurrently in Fairbanks. Figure 10 shows the home with an attached
garage mentioned above and indicates the presence of gasoline vapors indoors. Figure 11
shows a situation where most compounds had similar concentrations inside and outside and
lacked an indoor source. Methylchloroform (ME-CHLOROFORM) and hexane are hard to explain
and are likely contaminants. Toluene, however, may have an indoor source.-Figure 12 is a
location near a major intersection; concentrations are slightly higher outside than inside. A
compound found indoors in high concentrations and having a retention time similar to pentane
was identified as 2-propanoI. Figure 13 shows an attempt to indicate the increased evaporative
emissions at a Fairbanks gas station compared to roadside ambient measurements.
Concentrations of VOCs observed at the 2nd and Cushman sites were normalized to the
observed concentration of benzene, and these factors were used to predict concentrations at the
gas station. The observed concentrations of light hydrocarbons are much higher than predicted
concentrations; this indicates increasing evaporative emissions from refueling.
Disclaimer: The information in this document has been funded wholly or in part by the United
States Environmental Protection Agency. It has been subjected to Agency review and approved
for publication. Mention of trade names or commercial products does not constitute
endorsement or recommendation for use.
B-50
-------
MTBE
1000
ro
en
o
.a
Q.
Q.
cc
LJJ
O
z
o
o
LJU
o
<
DC
LJU
100
10
0.1
I
AUTO
RESID
GAS STATION
INDOOR
* * *
BAY I PARK.
BKG
GARAGE
ALASKA1
SAMPLING LOCATIONS
ALASKA2 ^ALASKA3 ES3 STAMFORD KX) ALBANY
t-
I
-------
BENZENE
1000
qo
tn
ro
(J
J3
Q.
a,
Z
DC
h-
2
LU
U
Z
o
(J
LU
(D
<
DC
LU
100
10
0.1
AUTO
RESID
p
/
/
/
/
1
GAS
STATION I SERVK
INDOOR
* *
£
o
o
o
O
o
o
Q
1
7"
/
/
/
/
/
/
'$,
._
\
\
^
* * * *
\
\
\
\
\
\
\
\
\
\
\
\
\
\
\
\
\
\
\
\
\
\
\
\
\
*
BAY | PARK. GARAGE
BKG
ALASKA 1
SAMPLING LOCATIONS
ALASKA2 £23 ALASKA3 R^ STAMFORD EZ3 ALBANY
-------
TOTAL NMOC
oa
U
JD
Q.
Q.
g
<
oc
H
z
LJJ
o
z
o
o
LJJ
CD
<
OC
LJU
100,000
10,000
1000
100
10
0.1
/N
XN
X>
XN
rtf
/N
/N
XS
/N
/N
/N
x^
I
r^>
^>
^
^>]
^>
^>
^>
V
?>
%
1
I
AUTO
/vC
^i
en
^
/\
1
» »
s\
I
^
^
i
* » #
ALASKA1
GAS STATION ] SERVICE BAY IPARK/GARAGE
RESID INDOOR BKG
SAMPLING LOCATIONS
ALASKA2 ^ALASKA3 ^STAMFORD K3 ALBANY
-------
1,1,1 -TRICHLOROETHANE
00
U 10,000 -
Q.
Q.
z
g
<
DC
LLI
U
Z
o
U
LLI
O
<
oc
LLI
1000 -
100 -
10
1 -
0.1
—
_
_ .
«•
1
" l-ia
AUTO
-------
FORMALDEHYDE
.a
Q.
Q.
«.
z
o
CD
61
01
O
Z
o
u
UJ
CD
<
cc
UJ
34
30
26
22
18
14
10
6
2
KXXXXXH
r *
1
0
/
/
/
^
\
\
\
\
\
\
\
\
\
\
\
\
X
X
AUTO
GAS STATION
RESID IND
SERVI
OOR
BKG
PARK. GARAGE
ALASKA1
SAMPLING LOCATIONS
ALASKA2 ^JALASKA3 ^STAMFORD (X3 ALBANY
-------
u
JO
CL
CL
oc
H
03 UJ
81 o
o
u
UJ
CO
h-
0
INDOOR/OUTDOOR MTBE
FAIRBANKS - DEC. 18-20, 1992
HUNTER SCHOOL I HAMILTON ACRES
2ND AND CUSHMAN
-------
CD
u
J3
Q.
Q.
g
<
en
h-
2
LJJ
CJ
2
O
CJ
LJJ
CO
h-
INDOOR/OUTDOOR MTBE
FAIRBANKS - FEB. 17 TO MAR. 2, 1993
2/24/93 2/26/93 3/2/93
2/17/93A
* 2/22/93 2/25/93
HAMILTON ACRES
2/25/93
2/25/93 3/1/93 2/23/93
HUNTER SHCOOL 2ND AND CUSHMAN
-------
w
u
jQ
CL
CL
g
<
DC
h-
z
UJ
O
O
Ul
z
LU
N
Z
UJ
co
340
INDOOR/OUTDOOR BENZENE
FAIRBANKS - FEB. 17 TO MAR. 2, 1993
OUTDOOR
INDOOR
2/25/93
/93
2/17/9312/23/93
2/2i/93 2/25/93
2/24/93 I 2/26/93 I 3/2/93
2/25/93 3/1/93
HAMILTON ACRES
HUNTER SCHOOL
2/2:
2ND AND CUSHMAN
-------
*
s
-Q
Q.
Q.
O
oc
UJ
u
O
u
UJ
Q
>-
UJ
Q
cc
O
INDOOR/OUTDOOR FORMALDEHYDE
FAIRBANKS - FEB. 17 TO MAR. 2, 1993
2/17/93 2/23/93
2/22/93 2/25/93 2/25/93 3/1/93
HAMILTON ACRES HUNTER SCHOOL
2/24/9312/26/93 3/2/93 12/23/93^
2/18/93 2/25/93
2ND AND CUSHMAN
-------
Hamilton Acres February 22,1993
Ambient vs Indoor
-------
w
O)
HUNTER SCHOOL MARCH 2, 1993
INDOOR vs AMBIENT
yu -
80-
70 -
U
•g. 60-
Q.
? 50 -
O
1
< 40-
DC
1
Z 30-
LLJ
U
Z 20-
o
0 K
10-
•
>
n
1,
n-PENTANE
*»
^n.^.
^r~i — n — i — i — i i i
ME-CHLOROFORM
n-HEXANE
MTBE
4
rir A
TTTT1 T 1
n
I
BENZENE
T r
*
i
T T~
1
4.327 15.672
L
•I INDOOR
^ AMBIENT
TOLUENE
r
|]j
T" * i " "
m&p-XYLENE
M^C
1 T" ~
|
|
J J 11 h Jk
* TTT T TTTIT
21.697 27.511
RETENTION TIME minutes
-------
OJ
6>
o
J3
Q.
Q.
Z
g
<
cc
LLI
O
z
o
o
200
190 -
180-
170-
160-
150-
140-
130-
120-
110-
100-
90-
80-
70-
60 -
50-
40-
30-
20 -
10-
0
2nd & Cushmari Streets February 26,1993
Indoor vs Outdoor Location
800 ppbC 2-PROPANOL
n-PENTANE
INDOOR
AMBIENT
TOLUENE
BENZENE
MTBE
4.138
T
16.927
u
(L
m&p-XYLENE
1
lili.
22.61
29.05b
RETENTION TIME minutes
-------
CORNERSTONE CHEVRON MARCH 2, 1993
700
X)
Q.
Q.
600 -
500-
O 400 -
03
cn
QC
O
z
O
O
300-
200 -
100 -
*!
OBSERVED vs PREDICTED VEHICULAR
n-BUTANE (1400)
OBSERVED
PREDICTED VEHICULAR
ISOPENTANE
n-PENTANE
4.246
TOLUENE
BENZENE
MTBE
T-V^T
m&p-XYLENE
T^
J
15.464 21.528
RETENTION TIME minutes
27.762
-------
TABLE I
SAMPLING CONDITIONS
CITY
PHASE 1
FAIRBANKS PHASE 2
PHASE 3
STAMFORD
ALBANY
DATES
DEC. 01-12, 1992
DEC. 17-22, 1992
FEB. 02 -MAR. 5, 1993
APRIL 14-15, 1993
MAY O5-15, 1993
MAY 26-27, 1993
AMBIENT TEMP. °C
-30 to +1.7
-33 to -12
-19 to -2
+6 to +20
+5 to +21
B-64
-------
TABLE II
WT.% MIBE IN FAIRBANKS, ALASKA FUEL SAMPLES
MINIMUM
MAXIMUM
AVERAGE
FUELS TESTED
COLLECTED DEC. 1992
UNLEAD REGULAR
1.86
15.41
8.46
6
PREMIUM
14.16
14.93
14.66
6
COLLECTED FEB. 1993
UNLEAD REGULAR
0.00
8.36
1.00
35
PREMIUM
0.83
8.77
5.56
34
B-65
-------
TABLE III
MTBE SAMPLING SITES
TYPE \LOCATI ON
AUTO TRAFFIC
RESIDENTIAL
GAS STATION
BACKGROUND '
GARAGE SERVICE
AREAS
PARKING GARAGE
INDOOR
VEHICLE
INTERIORS
FAIRBANKS
2ND /CUSHMAN* ;
UNIV . /AIRPORT
HAMILTON ACRES' •
HUNTER SCHOOL*
CONNERSTONE
CHEVRON
OLD NENANA HWYb
AURORA MOTORS;
FAIRBKS MOTORS;
SEEKINS FORD;
TUNDRA TOURS;
GABE'S AUTO;
RAE'S GARAGE
NONE
'HAMILTON ACRES;
HUNTER 'SCHOOL;
OLD POST OFFICE
SEVERAL CARS
AND TRUCKS
STAMFORD
LIBRARY
STARK SCHOOL
PLAYGROUND
MOBILE;
DPW PUMPS
SHIPPAN POINT
DPW;
SUBURBAN
CADILLAC;
DEMOTT'S
GARAGE
GOVT. CENTER
PUBLIC HEALTH
DEPT.; ST.
JOHNS RECTORY
NONE
ALBANY
NEAR PARKING
LOT AT NYDOH
(UNIV. PLACE)
CMT SITE IN
SCHENECTADY
MOBILE;
GETTY GAS
NONE
NONE
NONE
NONE
NONE
•CONCURRENT SAMPLING INDOOR AND OUTDOOR SOME DAYS.
fc!2 HOUR SAMPLING
B-66
-------
AITTHOR(S): Larry G. Anderson, Pamela Wolfe, and John A. Lanning
TITLE: THE EFFECTS OF OXYGENATED FUELS ON CARBON MONOXIDE AND
ALDEHYDES IN DENVER'S AMBIENT AIR
INTRODUCTION
We began studying the effects of oxygenated fuels in Denver in December 1987, a few
weeks before oxygenated fuels were used along the Colorado Front Range. Our initial work
attempted to determine the effects of oxygenated fuel use on concentrations of formaldehyde
and acetaldehyde in Denver's atmosphere. This work has continued with a 4-hour average
monitoring of formaldehyde and acetaldehyde, 24 hours a day, seven days a week, from roughly
October through April of the first few years, and year-round through the past 3 years (see T13)
and up to the present.
After the first year of the program, we learned that the state agency was evaluating the
effectiveness of oxygenated fuel use for reducing atmospheric CO concentrations. Emissions
test data and emissions and air quality modeling were used to calculate the expected reduction
of atmospheric CO concentration. We began looking for and are still researching ways of
analyzing the ambient concentration data to assess the effectiveness of oxygenated fuel use.
CARBON MONOXIDE EFFECTS
The techniques used in this work will not be discussed because detailed presentation
appears in Wolfe et al. (1993). In our analyses, we use a structural time series approach, which
requires a long data series. We have completed analyses using these techniques for five air
monitoring stations along Colorado's front range. These include the CAMP site in downtown
Denver, Carriage west of downtown in a residential area, National Jewish Hospital (NJH) east of
downtown in a residential/commercial area, and two sites outside Denver, one in Boulder and
the other in Colorado Springs. The data analyzed began in 1981 or as late as mid-1983
(depending on the site) and extends through January or February 1993. These data sets include
six complete oxygenated fuel seasons in Denver.
The analyses to be discussed here are for either monthly average CO concentration data,
which is most appropriate for assessing the normal or average behavior, and the monthly
maximum 8-hour average CO, which include only the extreme values, representing only the
extreme 8-hour period of the month. The analysis technique is intended to deal with lognormally
distributed data with serial correlations and to extract information about the seasonal, long-term
trend and oxygenated fuel effects on CO ambient concentrations. The model results for the
monthly averaged CO data and the monthly maximum 8-hour average CO data for the five sites
are graphically shown in the attached figures (T4 - T8).
Figure T9 summarizes the analyses results of the monthly averaged CO data. A
polynomial trend is allowed in the model; the highest order term that is significant is listed for
each data set. For all sites except CAMP there are significant downward trends that are either
linear or quadratic. At CAMP the cubic term is significant, which suggests that there has been
a significant slowing in the downward trend in CO. The oxygenated fuel parameter in the model
is not significant at the a = 0.05 level for any of the five sets. Although the oxygenated fuels
parameters are not significant, they do provide a valuable indication of the change observed.
For CAMP the oxygenated fuel parameter suggests that there was essentially no change in the
B-67
-------
monthly average CO concentration, while at Carriage, National Jewish Hospital, and Boulder
there are indications of a 4-7% decrease in CO attributable to oxygenated fuels. The analysis
for Colorado Springs requires further work to interpret, and will not be discussed. A summary
of the results of the analysis for the monthly maximum 8-hour average CO is shown in Figure
T10, a 2% reduction for CAMP and a 10-11% reduction for Carriage, National Jewish and
Boulder.
We have continued our analysis to try to determine the reasons for the differences
between CAMP and the other air monitoring sites. The CAMP site is in downtown Denver, and
subject to considerable commuter traffic. The other monitors are not in areas that are at the
beginning or end of many peoples journey. The morning CO is largely affected by emissions
from warm vehicles entering the downtown area, while the evening CO is affected by cold
vehicles starting and beginning their commute home. We investigated the trend in the 1-hour
average CO for the most common hour of the wintertime morning and evening CO peaks,
8-9 a.m. and 6-7 p.m., respectively. Figure T11 shows the results of the analyses of the monthly
average of all hours, the morning and evening hour. It is clear that during the winter, the evening
CO peak is much larger than the morning peak, while during the summer, the morning peak is
larger. From our analysis of this data (see T12), the oxygenated fuels effect on the morning CO
data was a reduction of CO by about 2% when the area was most affected by the emissions from
relatively warm vehicles. The effect of oxygenated fuel on the evening CO data was to increase
CO by about 6.5% when the area was most affected by emissions from relatively cold vehicles.
We believe that the cold start emissions have not been adequately dealt with in the models used
to assess the effectiveness of oxygenated fuels.
In downtown Denver, where CO is highest, the effect of using oxygenated fuels on CO
concentrations is a 0-2% reduction; for Carriage, National Jewish Hospital, and Boulder the
decrease is 5-11%. All these effects are much smaller than the about 20% reductions in CO
concentrations being predicted by the modeling for the more recent years. Cold start CO
emissions may explain the reasons CO reductions are smaller in downtown Denver than at the
other sites.
ALDEHYDE EFFECTS
A summary of the techniques used and the data collected has been presented by
Anderson et a!. (1993) and will not be repeated here. We have found very strong correlations
between formaldehyde and CO measured at the same site in downtown Denver, particularly
during the winter (see T14). This suggests that both formaldehyde and CO have the same
source. In Denver 80-90% of the CO emissions are from motor vehicles, which suggests that
motor vehicles are a major source of formaldehyde. Since we do not have CO data from the
same site as our aldehyde data for the entire time period, we have not looked at trends in the
formaldehyde to CO ratio. Figure T15 shows the observed trend in formaldehyde, acetaldehyde
and the correlation coefficient between formaldehyde and acetaldehyde over the 6 years of these
measurements. Figure T16 shows the trend in the formaldehyde-to-acetaldehyde (F/A) ratio for
this period. The formaldehyde concentration and the F/A ratio were significantly greater for each
of the last four winters, when compared to the first two winters. The formaldehyde concentration
during the most recent winter (1992-93) was significantly lower than during the previous three
winters, but the F/A ratio was not significantly different from that for the previous winter.
Emissions data have shown an increase in the F/A emissions ratio when MTBE blended
fuels are used (Hoekman, 1992). Those data show that the F/A emissions ratio for three-way
catalyst equipped vehicles increases from 2.3-2.4 for the reference fuel to 2.6-3.1 when 11% by
B-68
-------
volume MTBE blended fuels are used. These F/A emissions ratios and their increase are quite
similar to the F/A concentration ratios and their increase observed in Denver during the winter.
The required oxygen content of the fuel increased from 1.5-2.7% by mass during the six winters
of this study. These emissions data also suggest that the F/A emissions ratio should decrease
as noncatalyst equipped vehicles are replaced by newer three-way catalyst equipped vehicles.
These data suggest that motor vehicles are a major source of formaldehyde in Denver
during the winter, and that there has been a significant increase in the formaldehyde
concentration and the formaldehyde-to-acetaldehyde ratio. Fleet turnover effects should lead to
decreases in both quantities. Increased vehicle miles traveled should increase formaldehyde, but
not affect the F/A ratio. But the use of MTBE blended fuels is expected to increase both the
formaldehyde concentration and the F/A ratio. This is what we have observed over the 6 years
of oxygenated fuel use in Denver.
REFERENCES
Anderson, LG. et al., 1993. Effects of using oxygenated fuels on the concentrations of
aldehydes in Denver. Paper 93-TP-50.04, 86th Annual Meeting of the Air & Waste
Management Association, Denver, Colorado.
Hoekman, S.K. 1992. Speciated measurements and calculated reactivities of vehicle exhaust
emissions from conventional and reformulated gasolines. Environ. Sci. Technol. 26: t206-
1216.
Wolfe, P., L G. Anderson, and J. A. Lanning. 1993. A structural time series assessment of the
effectiveness of the oxygenated fuel program in reducing carbon monoxide
concentrations in downtown Denver. Paper 93-T-41 B.03, 86th Annual Meeting of the Air
& Waste Management Association, Denver, Colorado.
B-69
-------
The Effects of Oxygenated Fuels on
Carbon Monoxide and Aldehydes in
Ambient Air of Denver
Larry G. Anderson1'2
Pamela Wolfe2
and John A. Lanning1'2
Department of Chemistry, Box 194
2Center for Environmental Sciences, Box 136
University of Colorado at Denver
P.O. Box 173364
Denver, Colorado 80217-3364
MTBE Meeting
B-70
-------
Purpose of this Project
1. To develop techniques using ambient concentration
data that are suitable for assessing the effectiveness of
oxygenated fuels for reducing atmospheric concentrations
of carbon monoxide.
2. To collect and analyze the data necessary to assess
the effect of oxygenated fuels use on the atmospheric
concentrations of formaldehyde and acetaldehyde.
MTBE Meeting
B-71
-------
i
V)
t;
B-72
-------
>
CL
Q
O
O
1 C. iJ
I
j 1 VI w • 1
enver CAM? i
j
thiy Average |~
i
:
» 'J •
• .«ru 'o / '
X -•* «; /J;
H^u *•• ^^ >A
j
\\ ni ^i. >;£*. 7-
i
:
:
•
^i '
*» *
^^ ' '*
!A\ "
'/7u ,'/' 1
Jan-Si Jan-S3 . Jan-£5 Jan-57 Jan-S8 Jan-91 .J'an-93"
Jan-82 Jan-54 Jan-56 Jan-SS Jan-SO Jan-£2
Date
40-
35|
c
C
«
"»•
'(
30-
b. Denver CAM?
Mor.thiy Maximum 8-hr Average
Jan-cl 'Jan-SS Jan-S5 Jan-S7 Jari-S9 Jan-91 Jah-93
Jan-S2 Jan-54 Jan-65 Jan-SS Jan-SO Jan-92
n---^
L-C.LC;
Anaiys;s S5% CL I
B-73
-------
i
: ': ' ! -5 T\O"°J^" \'T"— ' ,' I
: ; <—. ~x w.- TW»>I««»* , .
McnLbiv Avers if
Jan-Si Jan-S3 Jan-£5 Jan-£7 Jan-SS Jan-91 Jan-S3
Jan-S2 Jan-S4 Jan-SS Jan-SS Jan-SO Jan-S2
Date
30-
25t
T 2°'
1 15-
O
O
b. Denver NJH
MoathJv Maziaun 8-hr A
ii
. 3
i;
' ?M' 'H' 'n' 1 jL I ' i ' /'"
,','4: ;/T, ;n; a-; . > / j /;
/Ji !| a i il| !,'»; iij ;•* i •";
finii ''-n' ***"•' »' "' .. • * '^" -J.
' '; ii'. .'f'ln ii-»i .'; x'. ^.- :i .< y ^*t'j
> n *» :}', ~ftt ; a >i< 51 .'^j .i • • <^ 11 i; 1'.
Jan-Si Jan-S3 Jan-S5 Jan-S7 Jan-SS Jan-Si
Jan-S2 Jan-S4 Jan-S6 Jan-SS Jan-SO Ja
Date
Jan-83
+ CO
Anaivsis
B-74
•fcC7o v_w
-------
a. uer.ve:
•-r—*•**—r——
cr.thiy Average
^-
I
O
0
UUUMLi-IU
.'/>
•
a: -.77
^ ^tTTuiTti tii* (i HI*\t*ti u tttIt117i IT*iiiiTnT»Mn^i7n*H
Jan-£1 Jan-£3 Jan-£5 Jan-£7 Jan-SS Jan-Si Jan-S3
an-S2 Jan-S4 Jan-£6 Jan-SS Jan-SO Jan-S2
Date
wU
t
^
o'
o
t
i
t
' i
' 1
t
t
i
i
t
"
*
-+•
h
-
;; i
t > *
[ r
{ |
1 |
i
\"> • • < • i
— r*', ! • ' • •'
1 1: i i " ^ i
'ft' ' 1;! ^* '
'/I '*;. ' i.: >
; -J/J' •*
: .'/T. JJ
! i * '"*
' 'U:» a-
b. Denve
Monthly
r
I
"(•• -A
: -/iv. :/, ;/
"• J !i & !!•*
!* Mi; i/ :; n
*(»»'' * '-^- »«!'•• ' ' * «
*» >*' »l» '''H; ;/'lU iii'T'- •*<'
\ii ^V^' nlj^^HAj^'^ ^
^ jtf ^ \^ ^
f
r
i
' Ccrnacfe
.^sximurn £-hr Average
j
i
f
j
i
•A
^
;' j< •'
' i ' 'ih ' ' ; ^
; ^ ^. i^-^ M 4. ^
V
%:'
w*'^
' ':
1
1,
i
t
y t
/
/
•
, If* k
^^
Jan-Si Jan-£3 Jan-£5 Jan-£7 Jan-£S Jan-Si Jan-
Jan-£2 Jan-64 Jan-£5 Jar.-SS Jan-SC Jan-S2
Date
- CO
Anaivsjs
B-75
-------
>
O
O
O
Jan-£1 Jan-£3 Jan-S5 Jan-£7 Jan-8S Jan-Si Jan-93
Jan-82 Jan-S4 Jan-S6 Jan-£S • Jan-SO Jan-22
Dale
25-
20:
15-
10-4
•(
'(
M
b. Bouider
Mcnrhlv Misiaua S-hr Averss«
! )
11
11
i )
Jan-£1 Jan-£3 Jan-£5 Jan-£7
Jan-52 Jan-£4 Jan-8S Ja
Jan-Si Jan-93
Jan-92
p.—
jf. : —
•*• CO
~~
C^ /C
B-76
-------
»:
M
! '
i j j
i
r;
1 3
I 3
*i -
i a
i\;j.7;f.v •ra*T.a-.j,/n» ^/i) ;
K/^;if'T; VktrV^^f^-T^'"-
1-4/iK^ l^/ il j^'l^*"
v" | i^ \xv'!'i5'' \,-;
o
Jan-Si .Jan-£3 Jan-£5 Jan-£7 Jan-£9 'Jan-91
Jan-£2 Jan-S4 Jan-£5 Jan-£S u
Date
Jan-33
iMiiJJtutitii
Jan-£1 Jan-S3 Jan-£5 Jan-£7 Jan-£9 Jan-91 Jan-93
Jan-£2 Jan-54 Jan-£5 Jar,-£S Jan-SC Jan-92
* CO
A i«»
~--
CT^ ,C
B-77
-------
CAMP (avg)
G? Trend: cubic
is* Qxy Parameter: not significant (+ 0.5% )
sr-Power of the Test: .83
Carriage (avg)
«3» Trend: linear (downward)
«5> Oxy Parameter: not significant (- 6% )
*3- Power of the Test: .55
National Jewish Hospital (avg)
*& Trend: quadratic (downward)
& Oxy Parameter: not significant (-7% )
«s* Power of the Test: .58
Boulder (avg)
*? Trend: quadratic (downward)
•s* Oxy Parameter: not significant (-4% )
*& Power of the Test: .44
Colorado Springs (avg)
«5* Trend: linear (downward)
•&• Oxy Parameter: not significant ( + 12% )
«• Power of the Test: .39
B-78
-------
CAMP (max)
53* Trend: cubic
S3* Oxy Parameter: not significant ( - 2% )
& Power of the Test: .43
JSP Continued Exceedances Likely
Carriage (max)
BS* Trend: linear (downward)
53* Oxy Parameter: not signi'ffoant ( - 10% )
•3- Power of the Test: .33
55* Continued Exceedances Likely
National Jewish Hospital (max)
is? Trend; linear (downward)
55- Oxy Parameter: not significant ( -10% )
Power of the Test: -.33
Boulder (max)
& Trend: linear (downward)
*? Oxy Parameter: not significant (-11%)
*? Power of the Test: .21
Colorado Springs (max)
c? Trend: linear (downward)
*3" Oxy Parameter: not significant ( + 1.5% )
«• Power of the Test: .31
B-79
-------
Denver CAMP Data. Comparison of
Monthly Average, 8 am and 6 pm Fits
12-
f *
-,n :•.. * ^
1U i rT~~l*~ si£
81; i_ __j , __j . . ;j . ^
^^ f i i ij«*t r.i i t i! ..r..x i
1
'a 6-i=r
o JA^'^/SS^F^ifflk-^'lS ^*L "7i" '*'
° 4%r " ^
-n r • \| ^ n T"~n
IS^c -Si* Be 3 CD,
~~ l^ I3i *•
/2 •««;" rsr- Wf •*••:.
0-
Jan-81 Jan-83 Jan-85 Jan-87 Jan-8S Jan-Si Jan-93
Jan-82 Jan-84 Jan-86 Jan-88 Jan-SO Jan-92
Date
-^- Month
PM
B-80
-------
Diurnal Effects on the Results of Oxygenated Fuels
CAMP Data - Monthly Averaged Carbon Monoxide
Oxy Fuels Effect R2
All Hours +0.5% .90
8 - 9 am - 2.0 % .87
6 - 7 pm + 6.5 % .95
Morning period - more warm vehicle emissions
CO decrease attributable to oxygenated fuels
Evening period - more cold start vehicle emissions
CO increase attributable to oxygenated fuels
We believe that CAMP shows less benefit from
oxygenated fuels use because a larger fraction of its CO
emissions are due to cold starts, as compared to the
other sites studied.
MTBE Meeting
B-81
-------
Denver Formaldehyde Concentration Data
Max, Min, and Average Concentrations
3U-
£ «*-
CL 35
•3O —
Q) 30
TJ
"° 20-
03
£
0
^ -m-4
TO
5_
n-
1
••>
•H
|l
"tf
1
-
"fc.
~~
1 !
i
-•^
i
~ i
-1 1
i
-1.
M
t
.
?!
ii
•
T
T
i
1
t i
Tl
i *
•
j,
~~
Jul-87 Jul-88 Ju!-89 JuI-90 Jul-S1 Jui-S2
Jan-S8 Jan-89 Jan-SO Jan-S1 Jan-52 Jan-93
Year
B-82
-------
HCHO, CO and Correlations
Winter 1991-92 Diurnal Behavior
t1t2t3t4t5t6 t1t2t3t4t5t6 t1t2t3t4t5t6
Dec Jan Feb
HCHO • CO
B-83
-------
Trend in winter HCHO, CH3CHO and
correlation coefficient
J-F
-0.75 .2
c
0)
o
£
03
O
O
c
to
•75
0.25 g
O
N 89 - F 90 N 91 - F 92
N88-F89 N90-F91 N 92 - F 93
Winter
HCHO • CH3CHO -*- Corr Coef
B-84
-------
Trend in winter formaldehyde to
acetaldehyde ratio
J-r-F88
N89-F90 • N 91
N88-F89 - N90-F91
Winter
F92
N 92 - F 93
B-85
-------
CONCLUSIONS
1. There is a strong downward trend in CO at all Colorado
monitoring sites studied.
2. This downward trend is site dependent. •
3. Oxygenated fuels have had no significant effect on the .ambient
CO at any of the five Colorado monitoring sites studied.
4. The effect of oxygenated fuels on ambient CO was_ found to be
a reduction of 5 - 11% at three of these sites. At the downtown
Denver site the effect was even smaller.
5. We believe that the benefits of oxygenated fuels use are
reduced in downtown Denver due to the greater importance of
cold-start emissions, as compared to the other monitoring
locations studied.
6. During the winter motor vehicles are a major source of HCHO
in the Denver metropolitan area.
7. There has been a significant increase in both formaldehyde and
the formaldehyde-to-acetaldehyde ratio since the beginning of
oxygenated fuels use (as the oxygen content of the fuel
increased).
8. Based upon emissions data, both formaldehyde and the
formaldehyde-to-acetaldehyde ratio are expected to increase
with oxygenated fuel use.
9. But the effect of oxygenated fuels use on ambient HCHO has
not yet been firmly established.
M75E Meetinc
B-86
-------
Acknowledgements
Faculty Associates:
Jeffrey Boon
James Koehler
Robert R. Meglen
Students:
Charles M. Machovec
Philip Anderson
Andrea Lewis
Neil Anderson
Kerry Grant ;
Heather Spurgeon
John Koeppe
Eugene Gorman
Zahra Nadji
Regina A. Barrel!
Joyce Miyagishima
Financial Assistance:
University of Colorado at Denver
Auraria Higher Education Center
Global Change and Environmental Quality
Program, University of Colorado
MTBE Meeting
B-87
-------
-------
APPENDIX C
SESSION THREE: SPEAKER ABSTRACTS
AND PRESENTATIONS
C-1
-------
AUTHOR(S): Ted Johnson (IT Corp.)
TTTLJE: SERVICE STATION EXPOSURES
ABSTRACT
The compound methyl tertiary butyl ether (MTBE) is routinely added to gasoline during
the winter driving season to reduce carbon monoxide (CO) emissions from motor vehicles in CO
non-attainment areas. MTBE is also added to gasoline during other seasons to increase octane
rating. In 1992, the U.S. Environmental Protection Agency (EPA) began receiving complaints of
headaches, nausea, and other symptoms following alleged wintertime exposures to MTBE. In
early 1993, EPA began planning a series of clinical research studies to investigate the validity of
these claims. To properly design these studies, EPA required estimates of typical air
concentrations of MTBE that motorists and attendants may experience during refueling at service
stations that dispense gasoline containing MTBE. EPA also expressed interest in determining
typical MTBE concentrations at the property boundaries of these service stations.
In response to these needs, the American Petroleum Institute (API) funded a field study
in which IT Air Quality Services (ITAQS) measured ambient MTBE concentrations at 10 service
stations in the New York metropolitan area. The stations included:
1. Two full-service stations with Stage li vapor recovery controls on a commuting
route near East Brunswick, New Jersey;
2. Three self-service stations with Stage I! vapor recovery controls in Westchester
County, New York; and
3. Five self-service stations without Stage II vapor recovery controls in Fairfield
County, Connecticut.
The selection of full-service stations in New Jersey was mandatory, as self-service stations are
not permitted in that state.
Each station was monitored on a different day between April 7,1993, and April 23,1993.
The monitoring activities at each station were conducted during two 4-hour periods, nominally
8 a.m. to 12 a.m. and 2 p.m. to 6 p.m. Four-hour canister and impinger samples were collected
at four perimeter locations (north, east, south, and west) and one pump location at each station,
in customer breathing zones at the New York and Connecticut stations, and in attendant
breathing zones at the New Jersey stations. In addition, 4-hour charcoal tube samples were
collected in the breathing zones of all stations. These samples were analyzed for MTBE, BTEX
(benzene, toluene, ethylbenzene, xylene), and formaldehyde.
Continuous CO measurements were made in the pump area of each station using a
Metrosonics pm-7700 monitor. Organic vapor analyzers (OVA) were used to continuously
monitor total hydrocarbon (THC) concentrations in the pump areas and breathing zones. These
measurements were made to identify individual refueling events that could not be distinguished
in the 4-hour samples collected by the canister samplers.
Field personnel monitored meteorological parameters, gasoline composition (oxygenate
content, Reid vapor pressure, BTEX), and gasoline sales and deliveries during each sampling
period. Personnel also noted the time each vehicle was refueled and conducted regular counts
C-2
-------
of traffic on nearby roadways. Gasoline pumping activities were continuously recorded by a
stationary video camera.
Research findings presented on July 27,1993, at the MTBE Workshop were limited to the
results of analyzing (1) MTBE data collected by canisters and (2) continuous THC data collected
by OVA. The principal findings are summarized below.
1. Mean and maximum 4-hour average MTBE concentrations generally decrease
from breathing zone to pump island to perimeter, suggesting that refueling
activities are the principal source of MTBE that is measured at service stations.
2. MTBE concentrations are generally lower at stations with State II vapor controls.
3. Mean 4-hour MTBE concentrations are below 1 ppm at breathing zone and pump
island locations and below 0.02 ppm at the station perimeters.
4. Maximum 4-hour MTBE concentrations are below 2.6 ppm at breathing zone and
pump island locations and below 0.2 ppm at station perimeters.
5. The canister breathing zone measurements may underestimate actual breathing
zone concentrations during fuel dispensing by station-specific factors ranging from
1 to 3. Most factors fall between 1.0 and 1.4.
During the ITAQS service station study, a research team headed by Dr. Paul Lioy collected air
samples in the passenger compartments of automobiles during the typical home-to-work
commutes. Automobiles in this companion study were refueled at stations included in the service
station study. Breathing zone MTBE concentrations measured by Lioy's team during these
refueling events were generally comparable to breathing zone measurements made by the ITAQS
team.
C-3
-------
^^ ~"
si
355
If
11
0)
Q. .2
c w
255
•c a)
o> o
E >
< *
C/)
o.
o
O
CM
.0)
C-4
-------
Service Station Monitoring Study
o
Designed to characterize MTBE air concentrations
at typical stations dispensing gasoline containing
MTBE
Monitoring Period: mid-April, 1993
10 stations monitored:
New Jersey: 2
(full-serve/Stage II)
Connecticut: 5
(self-serve/non-Stage II)
New York: 3
(self-serve/Stage II)
-------
Service Station Exposure Monitoring
Sampled:
- breathing zone
- pump island
- station perimeter
Analyzed:
- air
8- - MTBE, BTEX
° formaldehyde
0 CO, total hydrocarbons
- fuel
° oxygenate content
0 RVP
0 BTEX
Conditions:
- am/pm rush hour (4 hr samples)
-------
Service Station MTBE Concentrations
Stage I I/Full - Serve
o
INO.
Location
Br. Zone
Pump Island
Perimeter
All
4
4
16
values
N.D.
0
0
1
Min.
0.084
0.120
0.001
^UIIUWI
Max.
0.520
1.600
0.036
lUcUIUM, p
Median
0.245
0.440
0.003
prii
Geometric
Mean
0.224
0.409
0.005
Geometric
Std. Dev.
2.15
3.16
3.48
-------
mo
2
Oft*
Q)
f t ^^
C ^*
f\ u)
Oco
•
1
II 1
u _
•U 2
t u.
s ^
•——|
— . G)
^j ^M.
^ ^7J
trt ^*^
. . ^ ^
^"^ C/J
^j
C|^H
La
\f m
1
|
C
T-
•
>
I
•
1
I
n
u
•
•
.
•
1
6
^>
dx
^•'
Q)
EL_
r"
1:
.....
•. . . .
....
__
Q..g
E • rn
••V »^^^^
Ck j^-
....
....
....
f(\
*••
^^
O.
•
,
, . . t f
',','. '. irt
. . . .
aide
j
1
T—
0
g
0
•
'
•
•
(
•
•
•
•
•
•
f
*l
•
T-
C
c
c
C-8
-------
Service Station MTBE Concentrations
Stage Il/Self - Serve
2
NO.
Location
Br. Zone
Pump Island
Perimeter
All
6
6
24
vaiues
N.D.
0
0
0
Min.
0.077
0.014
0.002
vsUMuen
Max.
0.780
0.080
0.083
UclUUM, pi
Median
0.205
0.048
0.007
JIM
Geometric
Mean
0.204
0.038
0.008
Geometric
Std. Dev.
2.31
2.18
3.03
-------
0)
4V<
cu
^^-
fl\ ft\
O >
Ofli
*•*
f J \f W
^^
1
11 1
UJ if.
CQ o>
r™ f j*!
S\J w
— ,
•••
O
<3 2
25* CO
CO w/
G)
^^
0)
U)
C
T~
5
•
i>..
. , . .
....
L..—>.
....
....
. , .
I
....
. .
• •
• '•
* • s « •
• s • • • •
4
1
T-
I
O
LU
(
m
u
^
fe
1
PC
•
c
•x
"
t
!.
:
i
1
ic
i
i
O
•
O
D,
;
1
i
•
i
•
2
tan
EG
••
ft
••
0)
411
0)
CL
.
T"
C
c
c
C-10
-------
Service Station MTBE Concentrations
Non - Stage I I/Self - Serve
Location
Br. Zone
Pump Island
Perimeter
All
10
10
40
VCll UC7O
N.D.
0
1
2
Min.
0.170
0.009
0.001
Max.
2.600
1.500
0.140
Median
1.500
0.170
0.014
pi n
Geometric
Mean
0.978
0.109
0.014
Geometric
Std. Dev.
2.73
5.05
3.61
-------
ff\
"t
*- o
^^ ^
^_ «.
O/lk
V
0 (/)
c
<3 '
O^^~
41 *
ii i //%
UJ Q//
m-*^..
M
^^™
s S>
Ul
C CO
.2 5;
«•• v/
CO •
vS ^
w) c
O
w
Q) *9^
O^m
Q)
1 1
i""
•
|;«K£»!V$^
j
0
•
•ssssr
SS5.
ssa
•swar
0
fl)
%lr
"Z
0
a.
S«^
|
X
•
•
•
•
ft
Qi
C
«•
^
£
i
•
^. •«
*
i
i •
i«
i
1:
^KS
I
3MK
I
SSSft
J<
i
-------
0)
o
CO
O
X
a.
E
8|eos oi
C-13
-------
Estimation of Breathing Zone MTBE
Concentrations during Refueling Events Only
o
Separate 4 hour mean THC recordings into mean
refueling and non-refueling concentrations
Adjust calculated 4 hour mean THCs to equal canister values
Breathing zone MTBE concentration during refueling events =
adjusted mean THC values during refueling
x canister MTBE/THCs ratios
MTBE9
-------
Service Station Breathing Zone Concentrations
Station No. (type)
MTBE - ppm
Integrated
4-hr sample
During fuel
dispensing only
o
—A.
Ol
4 (Stage ll/self-serve)
5 (non-Stage ll/self-serve)
6 (non-Stage ll/self-serve)
7(
II
10 (Stage ll/self-serve)
0.10
1.20
0.17
1.50
0.16
0.1
3.9
0.21
2.06
0.15
-------
Individual Service Station Breathing
Zone Comparison during Refueling
- ppm
o. .- K, «. x During EOSHI EOSHI
Station No. (type) fue| d^pensjng Values
o 4 (Stage ll/self-serve) 0.11 0.35
O)
5 (non-Stage ll/self-serve) 0.17 0.13
6 (non-Stage ll/self-serve) 0.23
7 ( " " ) ? 0.59
10 (Stage ll/self-serve) ? 0.09
-------
Distribution of Service Station
Breathing Zone Data
o
—X
-si
12
10
O 6
<5
.0
E 4
All Service Stations
0 - 0.50 0.51 - 1.00 1.01 - 1.50 1.51 - 2.00 2.01 - 2.50 2.51 - 3.00
Concentration Interval - ppm
MTBE13
-------
Distribution of Service Station
Breathing Zone Data
o
03
HAM Stations Hsil/Full •sil/Self • non-SII/Self
0 - 0.50
0.51 - 1.00 1.01 - 1.50 1.51 - 2.00 2.01 - 2.50 2.51 - 3.00
Concentration Interval - ppm
-------
Distribution of Service Station
Breathing Zone Data
9
to
CO
CD
03
o>
.Q
£2
3
Z
1
SII/Full •Sll/Self B non-SII/Self
o - 0.50
0.51 - 1.00 1.01 - 1.50 1.51 - 2.00 2.01 - 2.50 2.51 - 3.00
Concentration Interval - ppm
MTDEI5
-------
o
Summary
Maximum and mean 4-hour average MTBE concentrations
decrease from breathing zone to pump island to
station perimeter and are lower at Stage II controlled stations.
Mean 4-hour average MTBE concentrations are below 1 ppm at
breathing zone and pump island location
and are below 0.02 ppm at the station perimeters.
Maximum 4-hour average MTBE concentrations are below
2.6 ppm at breathing zone and pump island locations
and are below 0.2 ppm at station perimeters.
Breathing zone measurements underestimate actual breathing
zone concentrations during fuel dispensing by station-specific
factors ranging between 1 and 3 for monitoring periods
analyzed to date. Most factors (4 of 5) range between 1 and 1.4.
-------
AUTHOR(S): Jack Hinton, Dr.Ph., CIH
TTTLE: OCCUPATIONAL EXPOSURES - MTBE '.....
There are five basic steps in bringing methyl tertiary-butyl ether (MTBE) to market:
• Manufacturing - producing MTBE at both chemical plants and petroleum refinery
facilities;
• Blending - introducing MTBE into motor gasolines, which includes handling both
neat MTBE and MTBE-blended fuels;
• Transportation - moving MTBE or MTBE-blended fuels via barge, tanker, railcar,
truck, or pipeline to points of distribution;
• Distribution - storing and moving MTBE-blended fuels from distribution terminals
to service stations; and
• Service Station - storing and dispensing MTBE-blended fuels to the public.
The American Petroleum Institute initiated a survey to collect and aggregate occupational
exposure data for MTBE from member companies. The data collected are typical of industry
operations are reflective of all the steps listed above, and span 11 years (May 1982 to March
1993). Further, the data are representative of all the major users and manufacturers of MTBE;
92% of the data were gathered in 1990, 50% of the data were collected in the oxyfuel winter
months, and 45% of the data were gathered during the 1992/1993 oxyfuel season.
A total of 2,038 exposure measurements were received and distributed as follows:
• 18% area samples
• 7% engineering source samples
• 12% personal samples where employees wore respiratory protection
• 63% personal samples where no respiratory protection was worn
The presentation is based on the data set of 63% personal samples where respiratory
protection was not worn. This data set is most representative of potential employee exposure
to MTBE. It represents at least the following number of employees per exposure grouping:
C-21
-------
Operation
Manufacturing
Blending
Transportation
Distribution
Service station
Other
Total
Number of Exposure
Number of Workers Measurements
881
1,800
1,489
7,705
37,753
a
49,628
365
523
641
305
41
8
1,883
"not determined
The data are representative of exposure groupings, with the exception-of the service
station category. However, the 37,753 employees in the service station category are felt to be
an overstatement of the actual number of employees with potential for job activity-related
exposure to MTBE. The majority of the 37,753 employees would be store clerks who are
responsible for collecting payment for gasoline sales and operating the "food mart" portion of the
station. Only a small fraction of this number would be employees whose job description included
dispensing fuel and vehicles service and/or repair. The 41 service station exposure
measurements are less representative than the other exposure measurement categories listed,
but they are not as "out of line" as the comparison of 41 measurements to 37,753 workers
implies.
The data are further aggregated for each operation category (manufacturing, blending,
transportation, distribution, service station) by sample duration (short-term - less than 30
minutes; task/activity - between 30 minutes and 6 hours; time-weighted-average (TWA)
workshift - between 6 hours and 9 hours; and extended workshift - greater than 9 hours) and
MTBE source (neat or fuel mixture).
The American Industrial Hygiene Association (AIHA) Workplace Environmental Exposure
Limit (WEEL) was used to assess exposure exceedances for task/activity, TWA workshift, and
extended workshifts. Because no comparable short-term exposure limit exists, the Excursion
Limit convention was borrowed from the American Conference of Governmental Industrial
Hygienists (ACGIH). This "rule of thumb" convention uses a 3-fold multiplication of the 8-hour
exposure limit value to determine an acceptable short-term exposure. This presentation
advocates neither the need nor the establishment of a short-term value equal to the 300 ppm
value used here (3 x 100 ppm AIHA - WEEL = 300 ppm Excursion Limit). The value of 300 ppm
is simply used as a convention to sort and present short-term data.
Personal exposures in the manufacturing category are less than1 10 ppm for all sample
types for both routine and maintenance operations. A single exposure of 249 ppm was reported
for a "bottle-washing" activity in a quality control lab. A review of information associated with this
value indicates this to be an atypical exposure, as the automated bottle-wash equipment is
usually controlled with exhaust ventilation; and other samples included in this data set, which are
representative of this activity, are below the 10 ppm limit reported above.
Personal exposures in the blending category are less than 100 ppm for all sample types
for both neat and fuel mixture operations, with the data predominantly being less than 10 ppm.
1The term "less than" is used for values below a stated concentration (e.g., -10 ppm) and should not be
confused with values below an analytical limit of detection, which are expressed as "below the Limit of Detection".
C-22
-------
Personal exposures in the transportation category are generally less than 50 ppm for
short-term activities associated with mixed fuel and are generally less than 200 ppm for
short-term activities associated with neat MTBE. Short-term exposures for both neat and fuel
mixtures can exceed 300 ppm (the highest value being 1050 ppm) and generally reflect barge
loading, sampling, and gauging activities or vacuum hosing associated with "pigging" (cleaning)
operations in pipelines. Activity/workshift exposures for neat MTBE activities are generally less
than 10 ppm, with occasional exposures ranging up to 711 ppm, where barge loading, sampling,
gauging, or .pigging operations occurred during the shift. Activity/workshift exposures for mixed
fuel activities are generally below 10 ppm, with no exposures seen above the 100 ppm WEEL.
Personal exposures in the distribution category are generally less than 10 ppm for the
short-term operations and are generally less than 1 ppm for activity/workshift timeframes.
Personal exposures in the service station category are generally less than 100 ppm for
short-term activities and are limited to vehicle repair or gasoline dispensing. Activity/workshift
exposures are generally less than 10 ppm. Monitored exposures for this category generally
represent full-service activities associated with dispensing fuel and garage repairs. Some
samples reflect weights and measure inspection activities and fuel dispensing pump repair.
From these data, personal occupation exposures to MTBE are generally well within the
AIHA 100 ppm WEEL. The ranges are as follows:
• 26% below the Limit of Detection
• 34% between the Limit of Detection and 1.0 ppm
• 36% between 1.0 ppm and 100 ppm and
• 4% in excess of 100 'ppm
Short-term exposures to MTBE are generally well within an excursion value of three times
the 100 ppm AIHA WEEL (300 ppm Excursion Limit). The ranges are as follows:
• 19% below the Limit of Detection
• 20% between the Limit of Detection and 1.0 ppm
• 59% between 1.0 ppm and 300 ppm and
• 2% in excess of 300 ppm
The data demonstrate that exposures in excess of 100 ppm TWA or 300 ppm short-term
occur infrequently and are generally limited to specific non-routine or "extraordinary" tasks. Once
determined, respiratory protection or other ventilation techniques are used to control exposures
in these situations.
A relative index based on geometric means (G.M.) of short-term and activity/workshift
TWA concentrations can be constructed to rank exposure potential. From this data set, the
exposure potential rankings would be:
C-23
-------
Operation
Transporting neat MTBE
Blending neat MTBE
Service station
Transporting MTBE/fuel mix
Manufacturing-maintenance
Distributing
Manufacturing-routine
Blending MTBE/fuel mix
MTBE
G.M. Short-Term
1 1 .0
5.1
4.7
3.3
1.0
0.85
0.84
0.58
(PPM)
G.M. TWA
0.24
0.58
0.77
0.13
0.14
0.13
0.06
0.10
C-24
-------
LU
CD CD
Z u-
CO
CD
D CO
CD O
CD
2 X
c c
05 O
O
O
CO
D_ O5
v. O)
o *-
§•"
It
O
C-25
-------
Occupational Exposures -MTBE
o
ro
o>
Aggregated member company MTBE exposure data
Timeframe: May, 1982 to March, 1993
16 companies responded
(All major users and manufacturers of MTBE)
-------
LJ
m
O
O
o>
c
o
o
d
CO
c
cd
Q)
CD
CO
cd
o
"•«—•
c
o
"S
4-»
CO
0)
0)
CO
C-27
-------
Occupational Exposures - MTBE
Data Variables:
o
10
00
Facility type
State
Measurement locations
MTBE source
Sample type
Job type
Month, Year
Sampling/analytical method
Control information
Operating conditions
MTBE concentrations
-------
Occupational Exposures - MTBE
Operation
# Workers
# Exposure Measurements
O
fc
to
Manufacturing
Blending
Transportation
Distribution
Service Station
Other
Total
881
1800
1489
7705
37753
not determined
49628
365
523
641
305
41
8
1883
MTBE5
-------
Occupational Exposures - MTBE
o
Data distributions: 2038 exposure measurements
18% Area samples
7% Source samples
63% Personal samples w/o resp. protection
12% Personal protection w/resp. protection
100%
MTBEB
-------
Occupational Exposure - MTBE
Manufacturing - Personal Samples
9
CJ
# Values
Operation
Routine
Maintenance/
Turnaround
Exposure
Type
Short-term
Task
8-hr TWA
Ex Shift
Short-term
Task
8-hr TWA
Ex Shift
All
33
0
82
2
14
1
12
2
ND
13
-
38
0
1
0
0
0
Ormnontratinn nnm
Min.
0.016
-
0.01
0.16
0.50
0.20
0.04
0.16
Max. Median G.M. G.S.D.
7.8 1.0 0.84 3.5
-
249 0.03 0.06 6.0
0.17
7.2 0.70 1.0. 2.6
-----
0.7 0.14 0.14 2.3
0.2 ...
MTBE7
-------
Occupational Exposures - MTBE
Manufacturing - Personal Samples
Routine Operations - Short-term
o
jtf.
! (';
1-10
c
"vi
...... o
111
E
Q.
CL
O
O
0
II
_J
--Ill
UJ
Q.
Q.
O
o
T-
X
n I
11-50 51-100 101-200 201-300
ppm
MTBE8
-------
O
CO
CO
Occupational Exposures - MTBE
R(
40
30
(/>
-------
Occupational Exposures - MTBE
Manufacturing - Personal Exposures
Maintenance/Turnaround - Short-term Exposures
o
CO
co
CD
D
8
7
6
5
2
1
0
ND
c
o
5)
o
X
ill
E
a.
Q.
§
CO
•if
LU
OL
a.
o
o
:1.0
1-10
11-50 51-100 101-200 201-300
ppm
MTBE10
-------
Occupational Exposures - MTBE
Manufacturing - Personal Samples
Maintenance/Turnaround - Activity/Workshift Exposure
O
w
01
14
12
10
8
0 6
4
2
CD
D
ND
1-10
ppm
11-50
o
o
LJ
•
51-100
MTBE11
-------
Occupational Exposure - MTBE
Blending - Personal Samples
o
G3
CD
# Values
Operation
;'
Neat
Fuel Mix
Exposure
Type
Short-term
Task
8-hr TWA
Ex Shift
Short-term
Task
8-hr TWA
Ex Shift
All
50
13
13
9
136
19
122
22
ND
1
1
5
9
47
14
78
13
Min.
0.01
0.21
0.04
0.23
0.02
0.03
0.02
0.01
loncentration, ppm
Max.
97
72
88
0.34
100
2
14
0.27
Median
2.3
1.0
2.6
0.3
0.4
0.05
0.05
0.02
G. M.
5.1
2.1
1.9
0.3
0.58
0.12
0.10
0.04
G.S.D.
5.6
5.3
9.2
1.1
9.4
3.9
4.1
2.7
MTBE14
-------
Occupational Exposures - MTBE
Blending - Personal Samples
Short-term Exposure Data
201-300
MTBE89
-------
Occupational Exposures - MTBE
Blending - Personal Samples
Neat - Activity/Workshift Exposure
9
W
09
0
ND
51-100
MTBE16A
-------
Occupational Exposures - MTBE
o
w
CD
CO
CD
3
O
*
100
80
60
40
20
Blending - Personal Samples
Mixed Fuel - Activity/Workshift Exposure
ND
<1.0
1-10
ppm
11 -50
51-100
MTBE16B
-------
Occupational Exposure - MTBE
Transport - Personal Samples
o
o
# Values
Operation
Neat
Fuel Mix
Exposure
Type
Short-term
Task
8-TWA
Ex Shift
Short-term
Task
8-TWA
Ex Shift
All
114
27
17
1
86
92
59
8
ND
4
4
1
0
4
28
14
0
Min.
0.3
0.04
0.02
0.32
0.01
0.02
0.01
0.19
Concentration, ppm
Max.
1050
700
712
-
508
59
26
4.5
Median
9.7
2.2
0.21
-
2
0.42
0.12
1.5
G. M.
11
2.3
0.24
-
3.3
0.51
0.13
1.2
G. S.D.
7.3
9.4
10.6
-
13.2
14.0
6.3
3.0
MTBE17
-------
Occupational Exposures - MTBE
O
a>
60
50
40
£30
20
10
0
NO
Transport - Personal Samples
Short-term Exposure Data
I
ui
E
a.
a.
o
o
UJ.
ui
E
a.
a.
o
o
X
o
rn.
_rzi_
<1.0 1-10 11-50 51-100 101-200 201-300 302
352
365
404
507
1050
ppm
MTBE18
-------
Occupational Exposures - MTBE
o
i
ro
Transport- Personal Samples
Neat - Activity/Workshift Exposures
1-10 11-50 51-100
ppm
540
700
711
MTBE19A
-------
Occupational Exposures - MTBE
Transport- Personal Samples
Mixed Fuel - Activity/Workshift Exposures
o
i
CO
1-10
ppm
11-50
51-100
MTBE19B
-------
Occupational Exposure - MTBE
Distribution - Personal Samples
o
Operation
All
Exposure
Type
Short-term
Task
8-hr TWA
Ex Shift
Jf" Willing Oonf^onf nflfin onm
TT vcMUco V>LM loci iiiaiiui i, LILJIII
All ND Mln. Max. Median G.M. G.S.D.
134 36 0.01 14 0.85 0.49 7.2
10 1 °-26 4 1.0 1.0 2.4
100 25 0.01 2.2 0.11 0.13 4.0
47 1 0.06 6.2 0.71 0.63 2.9
MTBE20
-------
Occupational Exposures - MTBE
o
i
01
(D
.=!
03
70
60
50
40
30
20
10
ND
Distribution - Personal Samples
Short-term Exposure Data
c
-------
Occupational Exposures - MTBE
Distribution - Personal Samples
Activity/Workshift Exposures
51-100
MTBE22
-------
Occupational Exposure - MTBE
Service Station - Personal Samples
o
Operation
Service Station
Exposure
Type
Short-term
Task
8-hr TWA
Ex Shift
7F V
All
11
5
13
11
ames
ND
2
0
0
0
^
Mln.
0.16
0.01
0.09
0.01
»unceniri
Max.
136
2.7
34
17
auon, ppm
Median
2.8
0.34
0.59
1.1
G.M
4.7
0.75
0.77
2.8
G.S.D.
11.5
2.8
4.7
4.5
MTBE23
-------
Occupational Exposures - MTBE
O)
§
o 2
0
Service Station - Personal Samples
Short-term Exposure Data
% * 5
**
ND
ET3
<1.0
1-10
11-50
ppm
51-100
c
to
'55
3
u
Ul
E
Q.
Q.
O
O
f>
II
ui
in
E
Q.
0.
101-200 201-300
MTBE24
-------
Occupational Exposures - MTBE
o
4,
(O
co
CD
Service Station - Personal Samples
Activity/Workshiff Exposures
ppm
MTBE25
-------
o
en
o
Summary
The data set is typical of industry operations
The data set is representative of employee exposures
The data set spans a 10 year period with the majority
of the data post CM oxyfuel initiation (92% since '90)
50% of the data represents oxyfuel winter months, with
45% of all data reflective of the '92 - '93 season
MTBE28
-------
Conclusions
O
01
Personal occupational exposures to MTBE are generally well
within the AIHA 100 ppm WEEL, ranging as:
u •
26% below Limit of Detection 36% between 1 & 100 ppm
34% between LOD & 1 ppm 4% in excess of 100 ppm
MTBE27A
-------
Conclusions
O
en
r\>
Short-term exposures to MTBE are generally well within an
excursion value of three times the 100 ppm WEEL (300 ppm).
19% below Limit of Detection 59% between 1 & 300 ppm
20% between LOD & 1 ppm 2% in excess of 300 ppm
MTBE27A
-------
Conclusions
O
61
CO
Data demonstrate exposures in excess of 100 ppm TWA
or 300 ppm Short-Term occur infrequently and are
generally limited to specific non-routine or
extraordinary tasks. Once determined, respiratory
protection or other ventilation techniques are used
to control exposures in these situations
MTBE27A
-------
Occupational settings ranked in order
of exposure potential would be:
o
4
Kh
G.M.
Short-term
1. Transporting Neat MTBE
2. Blending Neat MTBE
3. Service Station
4. Transporting MTBE/Fuel Mix
5. Manufacturing-Maintenance
6. Distributing
7. Manufacturing-Routine
8. Blending MTBE/Fuel Mix
11.0
5.1
4.7
3.3
1.0
0.85
0.84
i
0.58
G.M.
TWA
0.24
0.58
0.77
0.13
0.14
0.13
0.06
0.10
-------
o
en
Ol
Distribution of Occupational Exposures to MTBE
CO
CD
.=!
03
CD
E
ID
Z
1,400
1,200
1,000
800
600
400
200
0
All
Work Shift
Short-term
>Work Shift
Task-Specific
/ • / /
1-10
11-40
41-100 101-200 201-500 501-1000 >1000
Exposure Interval (ppm)
MTBE301
-------
AUTHOR(S): PJ. Lioy, C. Weisel, E. Pellizzari, J. Raymer
TITLE: VOLATILE ORGANIC COMPOUNDS FROM FUELS OXYGENATED WITH MTBE:
CONCENTRATION AND MICROENVIRONMENTAL EXPOSURES TO MTBE IN
AUTOMOBILE CABINS
There are two locations where field contact with gasoline and its constituents can lead
to a wide range of concentrations and exposure patterns: (1) the interior of an automobile during
stop/go commutes, and (2) the activities surrounding refueling practices. These
microenvironments were examined during April 1993 in two typical suburban commuting settings:
Fairfield County, Connecticut, Westchester County, New York, and Middlesex County, New
Jersey. In terms of gasoline dispensing, differences between these locations were that
Connecticut and NY had self-service refueling, and a mix of stations with Stage II and non-Stage
II vapor recovery controls. New Jersey had Stage II and operator-assisted refueling of vehicles.
The experimental design included: (1) selecting a specific commuting route in each area; (2)
selecting a specified protocol for refueling samples before, during, and after; (3) selecting four
vehicles that represent a mix of present (1992) and older (1986, 1987) automobiles; and (4)
collecting interior cabin and personal samples during refueling at self-service stations.
MTBE samples were collected using two methodologies: an active carboxen trap system,
and an active canister system. Results from side-by-side sampling with both methods and
duplicate analyses of specific canister samples indicated that our data quality objectives were
met for both sampling techniques. Both samplers were operated simultaneously in the commuter
studies and only trap samples were taken for the personal samples.
The cabin interior results for a 1 -hour run from the commuter study yielded a geometric
mean concentration of 8.2 ppb and a range of 1.2 to 160 ppb. The Connecticut commuter runs
had higher mean concentrations than the New Jersey runs, 23/;g/m3 and 16/^g/m3, respectively.
Comparisons of all paired vehicle hood (outside) and interior cabin (inside) samples indicated
that the highest differences were associated with an older vehicle (1987 Caprice). Emission
studies conducted on each vehicle appeared to indicate that there is an abnormally high
evaporative emissions factor for the Caprice which was between 8-30 times higher than the other
three vehicles. .
The refueling studies found that the highest levels of MTBE were associated with personal
refueling of vehicles at both the Stage II and non-Stage II gasoline stations (range between
13 ppb - 4100 ppb). Elevated but lower levels of MTBE were observed for the interior cabin
samples taken during refueling. The pre- and post-refueling levels were lower than the personal
and cabin interior samples during refueling. The post-refueling samples, however, were generally
higher than the values recorded before a car pulled up to the pump for vehicle refueling.
C-56
-------
VOLATILE ORGANIC COMPOUNDS FROM
FUELS OXYGENATED WITH MTBE:
CONCENTRATIONS AND MICROENVIRONMENTAL
EXPOSURES IN AUTOMOBILE CABINS
BY
PAUL J. LIOY AND CLIFFORD WEISEL
ENVIRONMENTAL AND OCCUPATIONAL HEALTH SCIENCES INSTITU1
PISCATAWAY N.J.
AND
EDO PELLIZZARI AND JAMES RAYMER
RESEARCH TRIANGLE INSTITUTE
RESEARCH TRIANGLE PARK, N.C.
-------
ACKNOWLEDGEMENTS
THE RESEARCH WAS SUPPORTED BY:
8 THE AMERICAN PETROLEUM INSTITUTE
WASHINGTON DC.
THE PROJECT DESIGN WAS DEVELOPED BY EOHSI/R.T.I. IN COLLABORATION WITH:
AN MTBE EXPOSURE WORKGROUP CONSISTING OF MEMBERS OF THE OIL/CHEMICAL
INDUSTRIES AND GOVERNMENT
-------
MTBE- OXYGENATED FUEL
MICROENVIRONMENTAL STUDY
HYPOTHESIS:
A MICROENVIRONMENT THAT CAN LEAD TO A RANGE OF EXPOSURES FROM
EVAPORATION OF MTBE IN OXYGENATED GASOLINE IS THE AUTOMOBILE
INTERIOR DURING PERIODS OF EXTENDED USE IN STOP/GO COMMUTER
TRAFFIC AND ACTIVITIES SURROUNDING REFUELING.
9
$
-------
MTBE- OXYGENATED FUEL
MICROENVIRONMENTAL STUDY
SPECIFIC AIMS:
A, TO QUANTIFY THE RANGE OF INTEGRATED CABIN CONCENTRATIONS
AND MICROENVIRONMENTAL EXPOSURE TO MTBE FOR A REPETITIVE NUMBER
OF TRIPS OVER A SPECIFIC COMMUTER ROUTE IN CENTRAL NEW JERSEY
B. TO QUANTIFY THE RANGE OF INTEGRATED CABIN CONCENTRATIONS
AND MICROENVIRONMENTAL EXPOSURE TO MTBE FOR A REPETITIVE
&
0 NUMBER OF TRIPS ON A SPECIFIC COMMUTER ROUTE IN FAIRFIELD
COUNTY, CONNECTICUT
C. TO COMPARE THE BREATHING ZONE CONCENTRATION OF DRIVERS
WHO RE-FILL THEIR GASOLINE TANK BY SELF-SERVICE vs. THE
IN-CABIN MICROENVIRONMENTAL EXPOSURE DURING ATTENDANT ASSISTED
REFUELING
-------
MTBE- OXYGENATED FUEL
MICROENVIRONMENTAL STUDY
SPECIFIC AIMS:
D. TO DETERMINE THE INCREMENTAL CABIN CONCENTRATIONS AND
MICROENVIRONMENTAL EXPOSURES OBTAINED DURING A COMMUTE DUE
TO REFUELING OF THE AUTOMOBILE vs. THE GENERAL OPERATING
CONDITIONS OF THE COMMUTE
E. TO COLLABORATE WITH INTERNATIONAL TECHNOLOGIES (IT) CORPORATION
o FOR THE INTERCOMPARISON OF SAMPLE/ANALYTICAL METHODOLOGIES list,
- THE COLLECTION AND ANALYSIS OF MTBE: SUMMA CANISTER, CARTRIDGE,:
CHARCOAL
-------
MTBE- OXYGENATED FUEL
MICROENVIRONMENTAL STUDIES
-TYPES OF SAMPLING SITUATIONS EMPLOYED IN THE FIELD EXPERIMENTS
1. PM. COMMUTE IN NEW JERSEY
2. A.M. COMMUTE IN NEW JERSEY
3. PM. COMMUTE IN CONNECTICUT
4. A.M. COMMUTE IN CONNECTICUT
5. CONNECTICUT GAS FULL SERVICE
- BEFORE
o - DURING ATTENDANT FILLING
B - AFTER
6. NEW JERSEY GAS FULL SERVICE
-BEFORE
- DURING ATTENDANT FILLING
- AFTER
7. CONNECTICUT/NEW YORK SELF SERVICE
- BEFORE
- DURING FILLING
- AFTER
8. CONNECTICUT
- PRE+ POST FULL SERVICE
- PRE+POST SELF SERVICE
-------
MTBE - OXYGENATED FUEL
MICROENVIRONMENTAL STUDY
• AUTOMOBILES USED IN STUDY
1. COMMUTER ROUTES
A. NEW JERSEY
-1987 CAPRICE
- 1992 WHITE CORSICA
B. CONNECTICUT
- 1986 MONTE CARLO
- 1992 GREY CORSICA
• GASOLINE FILL UP STUDIES
1. FULL SERVICE
- 1987 CAPRICE
- 1992 WHITE, GREY CORSICA
- 1986 MONTE CARLO
2. SELF SERVICE
- 1986 MONTE CARLO
-1992 GREY CORSICA
-------
EPA EMISSIONS
TESTS FOR COMMUTER VEHICLES
COMPONENT 86 MONTE CARLO
87 CAPRICE
92 CORSICA
#C0327W
92 CORSICA
#C0563G
HC
NO.
CO
CO,
EVAP.
(HOT SOAK)
0.82 g/mi
0.71
29.29
431.75
3.63
036 g/mi
0.62
1.54
376.60
25.55
0.08 g/mi
0.24
3.90
282.83
0.08
0.04 g/mi
035
1.05
286.25
0.16
CAR CHARACTERISTICS
1. Monte Carlo
5.0 L, V8 Engine - 4 Venturi Carb, Rear Wheel Drive
2. Caprice
5.0 L, V8 Engine - 4 Venturi Carb, Rear Wheel Drive
3. Corsica (Both)
3.1 L, V6 Engine - Multipoint Fuel Injection, Front Wheel Drive
C-64
-------
Design of Automobile Commute and Gasoline Refill
Experiments for MTBE Exposure
A. New Jersey/Connecticut Commuter Experiments (cm)
All Vehicles
C2t2
Midpoint
Start (AM/PM)
Return
Ecm = C1t1 + C2t
-------
Design of Automobile Commute and Gasoline Refill
Experiments for MTBE Exposure
C. Gasoline Refuel Exeriments - Self Service (gs)
c3t3
C2t2
Start (AM/PM)
Return
+ C2t2 + CC or Ct
44
33
Key:
1
2
3
4
Concentration in each portion of commute
Duration of exposure (samples) in each location
Microenvironmental Exposure
Interior pre-gas or midpoint
Interior post-gas or midpoint
Interior during refuel
Personal during refueling
C-66
-------
CUMMULATIVE STREET TRAFFIC
FOR EACH GAS STATION STUDIED IN NY-NJ-CONN
STATE # OF VEHICLES (8 HRS.)
NJ1
NJ2
NY4
NY9
6>
NY1.0
C3
C5
C6
C7
C8
26040
36360
15270
10410
15060
12030
8550
'11460
12300
13170
Source: API-IT
-------
MTBE- OXYGENATED FUEL
MICROENVIRONMENTAL STUDY
3. QUALITY ASSURANCE STUDIES
A. DUPLICATE SAMPLES
B. INTERCOMPARISON OF ANALYTICAL TECHNIQUES
C. INTERCOMPARISON OF SAMPLING TECHNIQUES- CANISTERS/CARTRIDGES
D. REPLICATE ANALYSIS
E. REPLICATE ANALYSES BETWEEN LABORATORIES
F. QUALITY CONTROL SAMPLES
G. FIELD AUDITS
-------
RTI Canister Comparison to EOHSI Adsorbent
For Commuting Samples
o
6>
(O
0
EOHSI Cone (M9/m3)
20 40 60 80
0.02
?
ex
ex
o
c
o
o
p 0.01
a:
0,00
O.I
i i r 1 i
/ 9
1
f
f
1 1:8 lln« \_
/ »'
1 .''ill LIn«
/ •'
f t>'
* *
1 r^. ' *
O * **
/O O .'" ^ ' '«:1 Iln«
' eft''' O
£ ,mo^ ^'' o
» * fl^Ujtx ^W
m Hff ••^..^ ^^
^^^*4. *^^
, ,''0. "
» X
/' " |
" I 1
30 0.01 0.02
100
- 80
- 60
70
o
3
n
- 40
- 20
u
EOHSI Cone (ppm)
-------
o
-g
o
RTI Canister Transferred to EOHSI Traps
Comparison for MTBE
EOHSI Trap Cone
E
ex
a.
0.20
0.16
c 0.12
o
o
u
V
.2 0.08
c
o
o
0.04
0.00
100 200 300 400 500 600 700
1 I 1 / I I t
I
1
t
1:2 line
f
t /
i ,'1:1 line -
r
1 t
1
_ f
If
1 ,'
} / 0 ,
/ • ' *>'
1 * ~
1 •' *''
i . ^2:1 line
1 •'' ^''
i * *
• x
- / * ^
^
/ . • ^
5^^ ,
/ uu
600
500
400
300
200
100
n
?i
H
^"«
o
Q
^
5°
"
rr1
o
-i
O
o
3
n
^
ID
\
3^
0.00 0.04 O.OB 0.12 0.16 0.20
EOHSI Trap Cone (ppm)
-------
o
MTBE Frequency Distribution
of Commuter Runs
Concentration
c
40
30
:s
u
i 20
tr
u
•
k»
U.
10
0
0.0
) 50 100 15
-
-
-
p]
J
7
^
\
'
',
',
''
7
',
',
/
^
;
/
i i
-
-
-
Y\ Y\ M ra M i M
00 0.014 0.028 0.0
Concentration (ppm)
-------
0.03
E
2: 0-02
c
o
o
o
c
o
O
0.00
MTBE Concentrations
During Commuter Runs
O
1
?-
New Jersey Conneticut
100
BO
O
3
O
60 a
O
•3
40
20
u
-------
MTBE Concentrations
Cabin vs Hood
Concentration
c
0.06
"?
QL
QL
~ 0.04
O
o |
CO •*-•
c
o
o
c
S 0.02
TJ
0
O
X
/•
n nn
) 40 80 120 160 2C
III!'
S
'
s
f
' 1:1 line
'
/
t
s
s
o
• /
*i~ - ° "
naBr 0 * * 0 0 °
,*^w^ i i
)0
200
160
c
o
n
o
120 E*
a
r+
^
O
^
80 ^
\
^
3
c
40
n
O 1066 Caprice
• 1002 Coriloa
V 1002 Coriloa
T 1066 Monte
Carlo
0.00 0.02 0.04
Cabin Concentration (ppm)
0.06
-------
Commuter Study
Hood-Cabin
MTBE - PPM
0 i M \/ i Geometric Geometric
bampie No. Values Mjn_ Max_ Mec|jan Mean Std. Dev.
o
Hood 104 0.0013 0.090 0.0051 0.0051 1.7
Cabin 156 0.0012 0.16 0.0068 0.0082 2.4
-------
WINTER, 1992-1993
MTBE COMMUTER EXPERIMENTS
o
•si
en
DATE
TYPE OF SAMPLE
12/7/92 INTERIOR
12/7/92 HOOD
12/7/92 INTERIOR
12/7/92 HOOD
12/15/92 INTERIOR
12/15/92 HOOD
12/15/92 INTERIOR
12/15/92 HOOD
12/15/92 INTERIOR
12/15/92 HOOD
12/15/92 INTERIOR
12/15/92 HOOD
12/24/92 INTERIOR
12/24/92 HOOD
MTBE
ug/m3
458
279
280
189
MTBE
ppm
70 0.020
75 0.021
360 0.102
1000 0.282
520 0.147
213 0.060
228 0.064
142 0.040
0.129
0.079
i
0.079
0.059
160 0.045
139 0.039
A 1QRR p| YIWIHI ITW
-------
Microenvironmental/PerBonal Exposure
During Refueling Studies NJ/NY/CT
o
10
E
a.
a.
c
o
^ 0.1
c
o
o
c
o 0.01
0.001
Stage II
c
a
b
i
n
full
1
n
o
n
a
I
• elf
Non-Stage II
T
n
full
o
n
a
I
• elf
10000
o
o
3
n
o
1000 2-
o
100
10
-------
Commuter Study
Personal
MTBE - PPM
c . M \/ i Geometric Geometric
bample No. Values Mjrh Max_ Median Mean std- Dev_
o
Stage II 3 0.35 4.1 0.57 0.93 3.7
Self-Serv.
NonStage 4 0.13 1.5 0.38 0.37 3.1
Self-Serv.
-------
CUSTOMER EXPOSURE DATA TO GASOLINE WITH 12-13% MTBE IN
PHOENIX. AZe 1990
20
18
16
14
12
M
«9
U
S3
5 10
o
O
O
8
3 3
og/m X 10 35 4
70.8
106.2
141.2
0 -
I
I I I
10 15 20 25 30 35
ppm MTBE
• Phoenix
D EOHSI Winter 1992 - in car
D EOHSI/RT! Spring 1993 -
persona!
Source: API. Clayron Environ., July 1991
C-78
-------
9
-j
ID
MTBE Concentration Before
and After Refueling
Pre Refueling Cone.(/ig/m3)
0 50 100 150 200 250
0.060
IT
o.
o.
o 0.045
c
0
0
o»
c
= 0.030
0
n
H-
0
K
J 0.015
0-
0.000
i i ii •
-
O
/
t
,
'1:1 line
x
/
*
/
/ _
O x
-
'
'
~ T V/ • -
^^Jx t
^
i i i i
250
TJ
200 S
r"f*
3J
a
-*•
c
150 SL
5
«a
o
o
100 3
o
f-*l
"\i
V3
50 3
u
0
0.000 0.015 0.030 0.045 0.060
O 1905 Cnprico
• 1992 Corsicn
V 1992 Corsicn
T 190G Monte
Carlo
Pre Refueling Cone, (ppm)
-------
Commuter Study
Refueled Runs
MTBE - PPM
0 . M ., . Geometric Geometric
bample No. Values MjrL Max> Median Mean Sld>
Before 27 0.0015 0.090 0.0069 0.0082 2.2
9 Refuel
During 22 0.0033 0.16 0.026 0.026 3.2
Refuel(Cab.)
After 27 0.0039 0.055 0.0090 0.0092 1.9
Refuel
-------
WINTER 1992-1993
MTBE GASOLINE REFUELING
DATE
TYPE OF SAMPLE
MTBE MTBE
jug/m
ppm
/4/93
PRE-FILLUP
INTERIOR-DURING FILLUP
POST-FILLUP
38 0.011
21000 5.90
4600
1.30
1/20/93 PRE-FILLUP
INTERIOR-DURING FILLUP
POST-FILLUP
88
0.025
2600 0.621
1150 0.325
-------
Log MTBE Concentration
for CT refueling
10
S
E
Q.
CL
C
o
C
V
u
C
o
o
0.1
0.01
r—
y
\
y
«.
S
V
s
M
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
1
>?
X
X
X
X
X
X
X
X
X
r
I 1
ra
\
V
\
V
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
^
X
X
X
X
X
X
X
X
X
X
X
1
4/12 4/16 4/10 4/28
Sampling Date
10'
I I Before
K\N During
Personal
After
o
o
3
O
9
3
1 O3 .2
10'
-------
O
CD
CO
E
Q.
CL
C
O
c
0
o
c
o
o
10
10
10
"1
10
~2
Log MTBE Concentration
for NY refueling
4/26 4/28
Sampling Data
10
o
o
3
O
9
•3
10s 2..
T:
«a
II B«for«
Pertonol
mi Aft.r
10* -'
u
-------
Frequency Distribution of
MTBE to Toluene Concentration Ratio
100
o
u
c
U.
n
n
Commute
j | Refuel
0 5 10 15 ' 20
Concentration Ratio of MTBE/To!uene (ppm/ppm)
-------
Preliminary Conclusions & Recommendations
• MTBE was found within the automobile microenvironment at ppb
levels during a I hour commute.
-• Personal MTBE measurements were found to be at the ppm (tng/rrr1)
levels during a 5 minute self-service gasoline refueling cycle.
• The 1 hour cabin MTBE levels during a commute exceeded those
2 of roadway air (hood samples) suggesting some retention or
evaporation of MTBE into the vehicle.
-------
Preliminary Conclusions & Recommendations
(Continuation)
The higher levels of MTBE were predominantly found within an
automobile that had higher volatilization of VOC from its
gasoline tank. This observation should be studied for a larger
fleet of these types of vehicles to determine the nature
of high-end exposures.
The in-vehicle MTBE concentrations measured in New Jersey
in the Spring- 1993 were generally lower than found in the
Winter 1992-1993 experiments. This comparison is not equivalent
because the RVP was approximately 20% - 25% lower in the spring
and different vehicles were used in each experiment.
' '•* •..-,-'•-'-''''' ^ • •. • : ' • •: '•-'*
. . v. , ( .. , ' j. J .'.',...:•' j ,| - .\ .,/'••',• • • , • '
1 The pe-rsohal exposures measurecl In the EOHSI/RTI study
-------
AUTHOR(S): Alan H. Huber
TITLE: HUMAN EXPOSURE ESTIMATES OF METHYL TERTIARY BUTYL ETHER (MTBE)
ABSTRACT
Data on ambient air quality and microenvironmental exposures (e.g., during refueling,
inside cars, in personal garages) are too limited for a quantitative estimate of population
exposures to MTBE. At best, they can be used to estimate broad ranges of potential exposures.
Because of the interest in MTBE, the present evaluation focuses on this compound, even though
any potential health effects might result from complex pollutant mixtures of which MTBE is only
one component. Furthermore, potential exposures of only the general public, not occupationally
exposed groups, were evaluated.
Figure 1 outlines the personal activities that have been considered in developing an
' annual human exposure estimate. Gasoline refueling is divided into two parts to cover both the
fill-up (1.5 fill-ups per week) and the remaining time spent in the station environment. The
distribution of hours spent in each microenvironment is based on reasonable interpretation of
available population activity studies. The greatest difficulty arose in trying to distribute the
balance of time spent in one's residence, office, or outdoors. In this example, meant to represent
one exposure scenario, the typical time one spends either at home or in a work place is relatively
large. Therefore, if there are elevated concentrations in these mtcroenvironments, they may
become the largest contributor to annual average human exposures.
There is a need to estimate both acute and chronic exposures to elucidate health risks.
A gasoline fill-up, although brief, results in the highest acute exposures because human
exposures are greatest when one is near evaporative emissions. Thus, exposures are greatest
when handling gasoline. Figures 2-13 summarize available concentrations during fill-up, at gas
stations, and in-cabin during automobile commutes. New field measurements (LJoy et a!., 1993;
Johnson, 1993) were collected in New Brunswick, NJ (two stations with full service and phase
II vapor recovery), Westchester County, NY (three stations with self service and phase II vapor
recovery) and Fairfield County, CT (five stations with self service and no phase II vapor recovery).
Ambient air quality was measured in Fairbanks AK, Stamford CT, and Albany NY (Zweidinger,
1993). Details on the study data should be obtained from the authors' presentations and reports.
The data analyses should be considered preliminary. The presentation below is provided to meet
an immediate need to present the range of MTBE concentrations in the identified
microenvironments and an annual exposure assessment with some margin of safety.
International Technologies Corporation (IT) completed a set of field measurements of
MTBE concentrations in the personal breathing zone during fill-up, at the pump island, and
around the property line of gas stations (Johnson, 1993). This study was done in coordination
with the Environmental and Occupational Health Sciences Institute (EOHSI)/Research Triangle
Institute (RTI) study (LJoy et al., 1993) at the same ten gas stations. All concentrations for the IT
study, even those in the intermittent breathing zone, were a 4-hour continuous sample. Because
the breathing zone collection was videotaped, IT will try to adjust the breathing zone
measurements to reflect a personal exposure during the fill-up. Four-hour average fence-line
MTBE concentrations were found to range from 0.018-0.234 mg/m3 (0.005-0.065 ppm). The
highest fence-line MTBE concentrations ranged from 0.36-0.5 mg/m3 (0.1 -0.14 ppm). The highest
4-hour average breathing zone and pump island MTBE concentrations ranged from 0-7-9 mg/m3
(0.2-2.5 ppm). These breathing zone concentrations are comparable to the 4-hour continuous
sample occupational concentrations in a recent National Institute for Occupational Safety and
Health (NIOSH) study (NIOSH, 1993). In the NIOSH study, the mean breathing zone MTBE
concentration for station attendants was 2 mg/m3 (0.58 ppm) with the highest concentrations
exceeding 14.4 mg/m (4 ppm). As expected, these breathing zone concentrations are lower
than reported by the Clayton Environmental Consultant study (Clayton, 1991), which collected
samples only during the refueling period. In the Clayton study, mean MTBE concentrations in
the breathing zone for 12-13% MTBE were 13 mg/m3 (3.9 ppm) with vapor recovery and
C-87
-------
30 mg/m3 (8.3 ppm) without vapor recovery. The absolute range of the MTBE concentrations
was 0.32 to 137 mg/m3 (0.088-38 ppm).
A wide range of ambient air concentrations within the breathing zone can be expected.
Ambient air concentrations measured at a gas station will be highly dependant upon wind speed
and direction. In addition, breathing zone concentrations can be dramatically influenced by how
one stands relative to the wind. A typical worst case MTBE concentration in the breathing zone
during refueling would be 36 mg/m (10 ppm) for a few minutes. However, an accidental spill
of fuel while filling the tank can dramatically increase the inhaled concentration.
Lioy et at. (1993) provides measurements of MTBE concentrations inside an automobile
during an approximate 30-minute commute and during refueling of the gas tank. A late-new
model automobile (199? Corsica) and an older-model automobile (1985 Caprice or 1986 Monte
Carlo) were assigned to each commuter route. The samples were collected in the front
passenger side of the automobile. The number of sampling runs (cases) per automobile ranged
from 14-20 for the commute and 3-5 for refueling. The driver's window was completely open
during the refueling. The average time to complete a fill-up was about 2 minutesr while the total
time at the gas station was approximately 5-10 minutes. Average in-cabin concentrations of
MTBE during the commute were found to range from 0.018 to 0.275 mg/m3 (0.005-0.075 ppm).
Average in-cabin concentrations during the fill-up ranged from 0.036 to 1.8 mg/m3 (0.1 -0.5 ppm).
In addition to the measurements inside the automobile, several measurements were collected on
the person refueling the gas tank. These concentrations were found to range from 0.7-14 mg/m3
(0.2-4 ppm). The older model automobiles were found to result in higher inside automobile
concentrations which probably reflect differences between the automobile design and "wear".
Figures 14-20 summarize air concentrations collected in Fairbanks, AK, Stamford, CT, and
Albany, NY (Zweidinger, 1993). Alaska was not in the MTBE program during the February/March
collection. Albany was not part of the MTBE program. Concentrations (except for garage/auto-
shop) in Alaska are reduced from 0.0072-0.14 mg/m3 (0.002-0.04 ppm) range to the 0.0036-0.054
mg/m3 (0.001-0.015 ppm) range after ending the MTBE program. Concentrations inside the
house were higher than outside for some cases, indicating that there may be a source of MTBE
indoors. It is possible that the residential garage may have had a source of evaporative
emissions after parking the hot car in the garage or from gasoline being stored in the garage.
Figures 21-23 summarize an EPA case study of measured evaporative emissions from an
automobile at rest after being run through the Federal Test Protocol (FTP) cycle that was
completed to provide an example herein. Approximately 0.5 grams of MTBE was emitted during
the 4-hour test. These emissions were then used as the modeled source in a 95 m3 garage
attached to a residential house. This is believed to provide a reasonable worst-case
demonstration of in-house concentrations due to a hot car parked in a closed residential garage
at 75°F, A multi-zonal mass balance model CQNTAM88 (Grot, 1991) was used to model indoor
concentrations. Peak concentrations were 2.3 mg/m3 (0.65 ppm) in the garage and 0.12 mg/m3
(0.035 ppm) in the residence. One-hour averaged concentrations in the garage ranged from 2.5-
4.3 mg/m3 (0.7-1.2 ppm), while concentration in the residence ranged from 0.072-0.32 mg/m3
(0.02-0.09 ppm). This is a worst-case situation because a newer car or cold winter temperatures
would likely have reduced evaporative emission rates resulting in lower concentrations.
Figure 24 summarizes the range of concentrations for the identified microenvironments.
The components of an annual average human exposure estimate are shown in Figure 25.
Commuting and gasoline refueling environments are clearly the most important, unless one has
significant evaporative emissions in the residential garage. The annual estimate uses the
Figure 25 values for the 4-month MTBE season and assumes that MTBE concentrations are 10%
of these values the remainder of the year. This assumption is based on belief that the amount
of MTBE in the ambient air is proportional to the amount of MTBE in the fuel (1.5% versus 15%
should allow a margin of safety). It is difficult to estimate MTBE levels during the non-oxyfuel
season because MTBE is used in premium gasolines and to a lesser extent in some regular
gasolines. These exposure values result in an annual estimate of 0.03 mg/m3 (0.0084 ppm) using
the low concentrations and 0.046 mg/m (0.013 ppm) using the high concentrations in Figure 25.
The above exposure scenario was calculated to represent a reasonable worst case exposure
C-88
-------
estimate for the working adult population after factoring-in the margin of safety. Exposure for
children is expected to be lower because children do not pump gas and spend less time
commuting in heavy traffig.
ACKNOWLEDGEMENTS
This presentation could not have been possible without the timely effort of Ted Johnson
(FTC), Paul (Joy and Cliff Weisel (EOSHI), and Kenneth Knapp and Roy Zweidinger (EPA) to share
their data files as the data are preliminarily processed. The presented analyses were prepared
within several days preceding the conference and was only possible with the assistance of Gary
Evans (EPA), John Streicher (EPA), Mike Zelenka (EPA), Azzedine Lanzari (METI), and Graham
Glen(METJ).
REFERENCES "
Clayton Environmental Consultants (1991). Gasoline Vapor Exposure Assessment for the
American Petroleum Institute. Clayton Project No. 31774.00, Jury 2,1991.
Grot, R. A. (1991). User Manual NBSAVIS COISTTAM88. Report NIST1R 4585, National Institute
of Standards and Technology, Gaithersburg, MD.
Johnson, T. (1993). Service Station Monitoring Study. (Abst. and presentation) Conference on
MTBE and Other Oxygenates, Falls Church, VA, Jury 1993.
Uoy, P. J., C. Weisel, E. Pellizzari, and J. Raymer (1993). Volatile Organic Compounds from
Fuels Oxygenated with MTBE: Concentration and Microenvironmental Exposures to
MTBE in Automobile Cabins. (Abst. and presentation) Conference on MTBE and Other
Oxygenates, Falls Church, VA, July 1993.
National Institute for Occupational Safety and*Health (1993). HETA 88-304-2326 American
Petroleum Institute, Washington, DC.
Zweidinger, R. B. (1993) Air Quality Measurements in Fairbanks, Stamford, and Albany. (Abst.
and presentation) Conference on MTBE and Other Oxygenates, Falls Church VA July
1993.
C-89
-------
PERSONAL ACTIVITIES
o
8
8760 hours/year
TIME (hours)
1. Gas Fill-up
1.5/wk @2min
other @ 10min
2. Commute/In Vehicle lohr/week
3. AutO Shop 4/yr@15min
4. Public Garage 2/day @ lomin
5. Residential Garage 2m'm/day
6. Residence 10hr/day-fweekend
7. Office 40hr/wk
8. OTHER/PUBLIC BUILDINGS 17/wk
9. Outdoors 20hr/wk
2.6
13.0
520
1
60.83
12.16
4160
2080
884
,1040
Figure 1
-------
CLAYTON FILL-UP STUDY
October - November, 1990
UN/40 H/N/40 L/Y/34 H^/6 H/N/42
Axis Legend: High or Low MTBE Fuel/ Yes or No Vapor Recovery/ Sample Size
LOW •HIGH BIMEAN
Figure 2
-------
s
2.5
Q.
a,
ill
m
h-
1.5
0.5
0
4 hr., AM sample
Fill-up Stations Data
ITC
AM Samples
34567
Station #
i
Figure 3
8
Pump
Breath Zone
10
-------
o
Fill-up Stations Data
fence-line locations
ITC
4 hr., AM Sample
u. 10
04 A
. 14
04 O
. I/I
f— , /~\ j
§ 0.1
a,
^ 0.08
m
t: 0.06
On/i
.U4
Ono
.U/i
n
o
CD
*.
5
1 4
-L- 5 4
f; I i
° ' • 1 i ——
-^- , +4-
4
••w
•••••»
^
••••••
^
/
•*«••••
^
iHHI^^
••WOT
f
^^g^,
•••<••
M^iMH
Max
— Mean
Min
t
Lof .ilions 1,2, 3, & 4, combined
456
Station #
Figure 5
7
8
9 10
-------
o
6
01
Fill-up Stations Data
fence-line locations
ITC
4 hr.,PM Sample
U. 1
Or\o
.Do
fi r\ r\o
g- O.Oo
LU
CD n r\A
^ U.U4
^
Ono
.UZ
n
-
-
-
\^
5 —
-J-
' ^
MMHM «H^BW
f
t
•MBIW
5
•MMVM
-t~
) e
•MM* a*m*>
5
•MHMM
••••^
t
m^mm
•HWM
MVM^
MMOTI
I
L
5 —
MHMB
1
MMJ»
Max
— Mean
Min
1 23456789 10
Station #
Locations 1,2, 3, & 4, combined
Figure 6
-------
In-Cabin Concentrations During Refueling
1985 Caprice
o
0.18
0.16
0.
E 0.12
CL
Q.
LU
CD
0.08
0.06
0.04
0.02
Group 7; NJ during fillup
Max
Mean
Min
RTI
EOHSI
Figure 7
-------
In-Cabin Concentrations During Refueling
1986 Monte Carlo
o
-------
In-Cabin Concentrations During Refueling
1992 Corsica (W)
o
(b
CD
U.. 1
-------
In-Cabin Concentrations During Refueling
1992 Corsica (G)
o
*
u. iz
04
,\
Onn
.Uo
QL
CL
"""*'' n HR
LU °'06
H
•^ n C\A
U.UH
Ono
.UZ
n
-
-
-
4
4
Max
— Mean
Min
t
RTI
Figure 10
Groups 11 & 13, combined; CT during fillup (full- and self-serve, resp.)
EOHSI
-------
In-Cabin Commuting Concentrations
1985 Caprice
\J.\JO
Or\7
.U (
On£j
.uo
o
—*
° *cT n r\t^
C U.UD
Q.
Q.
" — n n/i
ULJ U'U4
CD
Li n m
S ^-^^
Ono
.U/c
On-i
.U 1
n
-
-
-
14
St. Dev=0.0194
-
17
Std Qev=0.0038
'
Max
— — Mean
Min
•
i. !
EOHSI
Figure 11
RTI
-------
In-Cabin Commuting Concentrations
1986 Monte Carlo
o
_&
o
0.025
0.0225
0.02
0.0175
!. 0.015
Q.
^0.0125
CD
H 0.01
0.0075
0.005
0.0025
0
Sid. Dev=0.0043
20
Std.Dev-0.0025
Max
Mean
Min
EOHSI
RTI
Figure 12
Group 2
-------
In-Cabin Commuting Concentrations
1992 Corsica (W) & 1992 Corsica (G) - combined
o
o
ro
LU
0.035
0.03
0.025
0.02
0.015
0.01
0.005
0
35
Std.Dev.=0.0034
EOHSI
Figure 13
39
Std.Dev.=0.0051
Max
Mean
Min
RTI
-------
ALASKA MTBE
December 17-22, 1992
o
g
o
CO
AIRPORT
GAS ST.
HOUSE
POST O.
SCHOOL
0
Ambient 8-hour Sample
1
j*™'""*"''™''"^' '"W™ ™™™
HQffMIMMMMHV
£2a™
'
kv (: ii-,iy.(J. :>:.;.-. .((««!. -M4M
RMMMHMMMVIII
^^;tjj»J
^^
,..J
i ill
'
1 1 1 1
. .4 1 1 1 _
1 1 1 4—
0.01
0.02 0.03 0.04
CONCENTRATION (ppm)
Figure 14
0;05
-------
ALASKA MTBE
December 17-22, 1992.
Indoor 8-hour Sample
o
.^.
g
HOUSE
POSTO.
SCHOOL
!
i
i
!
i
i
j
i
I
1
1—f-
0
0.01 0.02 0.03 0.04
CONCENTRATION (ppm)
0.05
Figure 15
-------
ALASKA MTBE
February & March 1993
o
o
in
AIRPORT
POST O.
SCHOOL
HOUSE
GAS ST.
0
Ambient 8-hour Sample
i „! r | _i
I | i i
-i ^
0.005 0.01
CONCENTRATIONS (ppm)
0.015
Figure 16a
-------
ALASKA MTBE
FEBRUARY & MARCH 1993
AIRPORT
POSTO.
SCHOOL
HOUSE
GAS ST. £
0
Ambient 8-hour Sample
-I 4
0.05 0.1
CONCENTRATIONS (ppm)
.0.15
Figure I6b
-------
ALASKA MTBE
February & March 1993
o
I
VEHICLE
POST O.
SCHOOL
HOUSE
GARAGE
0
Indoor 8-hour Sample
0.05 0.1
CONCENTRATIONS (ppm)
Oil 5
Figure 17
-------
CONNECTICUT MTBE
April 13-14, 1993
Ambient 8-hour Sample
SCHOOL
LI. SOUND
LIBRARY
Parking
Garage
GASST
^^i Hfca*
I 1-
1 1 \ 1 1 1 1
H 1
0
0.01 0.02 0.03 0.04
CONCENTRATION (ppm)
0.05
-------
9
I
CONNECTICUT MTBE
April 13-14, 1993
Indoor 8-hour Sample
OFFICE
Parking
Garage
0
See Figure 19b for better scale
-4 1 1-
-I 4 » f-
1 -4-
0.1
0.2
0.3
CONCENTRATION (ppm)
I—I-
0.4 .0.5
Figure 19a
-------
CONNECTICUT MTBE
April 13-14, 1993
o
-±
o
OFFICE
Parking
Garage
0
Indoor 8-hour Sample
0.005 . 0.01
CONCENTRATION (ppm)
0.015
Figure 19b
-------
NEW YORK MTBE
Albany
OTHER
o
GAS ST.
0
8-hour Sample
-I (__4 j }__i 1 ( [ | 1 f
0.01 0.02 0.03 0.04 0.05
CONCENTRATION (ppm)
Figure 20
-------
E
D
i_
cn
CO
CO
0.6
0.5
0.4
0.3
0.2
0.1
0.0
Evaporative Emissions at 75F
Following FTP Cycle
Vehicle: 1 986 Monte Carlo
Carburetor Fuel System
Milage: 80,300
Fuel: Alaska II (Sun)
MTBE content: 12.1%
RVP:15.5psi
Benzene q tg/min] = 0.0007*exp(-time/60)
Benzene Emission
i , i . i i
0 20 40 GO 80, 100 120 140 160 180 200 220 240'
TIME tminutes)
Figure 21
-------
• MTBE Evaporative Emissions
Q = 1 2.9-Exp(-t/45) (mg/min)
o
_&
_i
W
£
CL
a.
S if
c
o
c
(U
u
c
o
o
LJ
m
i—
0.7 r
0.6
0.5
0.4
0.3
0.2
0.1
0.0
0 1
-*— Garage
o Kitchen
56 7 8 9 10 11 12
Time (hr)
Figure 22
-------
0.04
1 0.03
CL
c
o
.b 0.02
c
QJ
U
C
o
o
LJ 0.01
CD
1—
0.00
MTBE Evaporative Emissions
Q = 1 2.9 Exp(-t/45) (mg/min)
-
0 . o Kitchen
oo • .-..a--- Living
"™ •• 1O
o • ™
/ JU
' H
°»j1 »
" / *\«
• ®ffl
- QJ" ^"'"'HBBlHBBfflHIH
i . I , t i i . i i I i I . i • i • I '• i • I • I
0 1 2 3 45 6 7 8 9101112*
Time (hr)
Figure 23
-------
-------
|