UNITED STATES ENVIRONMENTAL PROTECTION AGENCY

WASHINGTON, D.C.  20460

OFFICE OF           

PREVENTION, PESTICIDES

AND TOXIC SUBSTANCES

DATE:		April 10, 2007

SUBJECT: 	Methyl Bromide: Phase 5 Health Effects Division (HED) Human
Health Risk Assessment For Soil, Greenhouse, and Residential/Structural
Uses.  PC Code: 053201, DP Barcode: D337288		

FROM:	Jeffrey L. Dawson, Chemist/Risk Assessor

Elizabeth Mendez, Ph.D., Toxicologist/Risk Assessor

Reregistration Branch 1

Health Effects Division (7509P)

THROUGH:	Michael Metzger, Chief

Reregistration Branch 1

Health Effects Division (7509P)

TO:		Steven Weiss, Chemical Review Manager

Special Review & Reregistration Division (7508P)

The risk assessment for the fumigant methyl bromide is attached.  It
builds upon the information that has been presented in previous
assessments including the Phase 5 assessment for the commodity (i.e.,
food) uses of methyl bromide (D304623, 3/10/06 & its Addenda D304619,
7/12/06) and the Phase 1 and 3 assessments, previous to that, which
addressed all uses of methyl bromide (D311945-1/31/05 &
D316326-6/10/05).  As indicated in D316326, methyl bromide can be used
in 5 key industry sectors (4 of which are based on a similar structural
analysis) including: (1) pre-plant soil; (2) greenhouse/potting soil;
(3) commodity; (4) industrial facility; and (5) residential.  The Phase
5 commodity assessment (D304623) directly addressed all uses related to
the commodity and industrial facility uses (i.e., all uses with food
tolerances and a few related others).  Hence, this document contains no
food or drinking water analyses nor does it address aggregate or
cumulative exposure issues since potential risks from those exposures
were addressed in D304623.  In this assessment, all other non-food uses
are addressed.  However, in many cases, the analyses that were completed
in D304623 are directly applicable to the greenhouse and residential
uses and, as such, are cited as appropriate in this document rather than
reproduce the information herein.  This assessment also incorporates
necessary modifications that were identified as a result of the comments
received related to both assessments.  Specific responses to these
comments, however, will be provided in upcoming, separate documents. 
Both of the previous documents and supporting information can be found
at   HYPERLINK "http://www.Regulations.gov"  www.Regulations.gov  under
methyl bromide.  

This risk assessment addresses both exposures in the general population
and for those occupationally exposed.  The key concern is exposure that
can occur in the general population primarily via inhalation for those
in proximity to treated fields and facilities (i.e., bystanders).  The
key difference between the results presented in this document and the
previous soil assessment is that the total applicable uncertainty factor
(which establishes the Agency's level of concern)  for acute bystander
exposures has been reduced from 300 to 30 because the previous
requirement for submission of a developmental neurotoxicity study has
been met and the results indicate no need for an additional uncertainty
factor.  Information pertaining to the selection and use of air models
for predicting off-target risks to bystanders has also been updated to
reflect the methods that have been used to develop the risk estimates
herein (i.e., based on the PERFUM model), to provide more extensive
characterization of the modeling methods and to provide further
clarification pertaining to the selection of PERFUM for this assessment
and the potential utility of other modeling systems (e.g., FEMS or
CALPUFF).



HUMAN HEALTH 

RISK ASSESSMENT

Methyl bromide

 

U.S. Environmental Protection Agency

Office of Pesticide Programs

Health Effects Division (7509P)

Jeffrey L. Dawson, Chemist/Risk Assessor

Elizabeth Mendez, Ph.D., Toxicologist/Risk Assessor

Date: April 10, 2007

HUMAN  HEALTH  RISK  ASSESSMENT

Methyl bromide

Risk Assessment Team:

Risk Assessor:					Jeffrey L. Dawson 

						Elizabeth Mendez, Ph.D.

Dietary Risk:					Felicia Fort

						Michael Metzger MS

						Toiya Goodlow

Product and Residue Chemistry:		Christine Olinger

						Michael Metzger MS

Occupational and Residential Exposure:	Jeffrey L. Dawson 

						Susan Nako

						Sherrie Kinard

Epidemiology:					Jerome Blondell, MPH, Ph.D.

						Ruth Allen, Ph.D.

						Monica  Hawkins MPH

						Hans Allender, Ph.D.

Toxicology:  					Elizabeth Mendez, Ph.D. 

						Byong-Han Chin, Ph.D.

Drinking Water Estimates:			Faruque Khan 

Table of Contents

  TOC \o "1-5" \h \z \u    HYPERLINK \l "_Toc163953987"  1.0  Executive
Summary	  PAGEREF _Toc163953987 \h  6  

  HYPERLINK \l "_Toc163953988"  2.0  Ingredient Profile	  PAGEREF
_Toc163953988 \h  10  

  HYPERLINK \l "_Toc163953989"  2.1	Structure and Nomenclature	  PAGEREF
_Toc163953989 \h  12  

  HYPERLINK \l "_Toc163953990"  2.2	Physical and Chemical Properties	 
PAGEREF _Toc163953990 \h  12  

  HYPERLINK \l "_Toc163953991"  3.0  Hazard Assessment and
Characterization	  PAGEREF _Toc163953991 \h  13  

  HYPERLINK \l "_Toc163953992"  3.1	Hazard Characterization	  PAGEREF
_Toc163953992 \h  13  

  HYPERLINK \l "_Toc163953993"  3.1.1	Database Summary	  PAGEREF
_Toc163953993 \h  13  

  HYPERLINK \l "_Toc163953994"  3.1.2	Endpoints	  PAGEREF _Toc163953994
\h  13  

  HYPERLINK \l "_Toc163953995"  3.1.3	Dose-response	  PAGEREF
_Toc163953995 \h  14  

  HYPERLINK \l "_Toc163953996"  3.1.3.1  Inhalation Exposure	  PAGEREF
_Toc163953996 \h  14  

  HYPERLINK \l "_Toc163953997"  3.1.3.2  Dermal Exposure	  PAGEREF
_Toc163953997 \h  17  

  HYPERLINK \l "_Toc163953998"  3.1.3.3  Classification of Carcinogenic
Potential	  PAGEREF _Toc163953998 \h  17  

  HYPERLINK \l "_Toc163953999"  3.1.4	Endocrine Disruption	  PAGEREF
_Toc163953999 \h  17  

  HYPERLINK \l "_Toc163954000"  3.2	Uncertainty Factors	  PAGEREF
_Toc163954000 \h  18  

  HYPERLINK \l "_Toc163954001"  3.3	Summary of Toxicological Endpoint
Selection	  PAGEREF _Toc163954001 \h  19  

  HYPERLINK \l "_Toc163954002"  4.0  Public Health Data	  PAGEREF
_Toc163954002 \h  20  

  HYPERLINK \l "_Toc163954003"  5.0  Non-Occupational Exposure
Assessment and Characterization	  PAGEREF _Toc163954003 \h  22  

  HYPERLINK \l "_Toc163954004"  5.1	Residential Bystander Exposure and
Risk Estimates	  PAGEREF _Toc163954004 \h  24  

  HYPERLINK \l "_Toc163954005"  5.1.1 	Bystander Exposures and Risks
From Known Sources	  PAGEREF _Toc163954005 \h  25  

  HYPERLINK \l "_Toc163954006"  5.1.1.1 	Methods Used To Calculate
Bystander Exposures And Risks From Known Sources	  PAGEREF _Toc163954006
\h  25  

  HYPERLINK \l "_Toc163954007"  5.1.1.2	Bystander Exposures And Risks
From Known Sources	  PAGEREF _Toc163954007 \h  43  

  HYPERLINK \l "_Toc163954008"  5.1.1.2.a:  Bystander Exposures And
Risks From Pre-Plant Agricultural Use	  PAGEREF _Toc163954008 \h  45  

  HYPERLINK \l "_Toc163954009"  5.1.1.2.b:  Bystander Exposures And
Risks From Greenhouse Use	  PAGEREF _Toc163954009 \h  60  

  HYPERLINK \l "_Toc163954010"  5.1.1.2.c: Bystander Exposures And Risks
From Residential Use	  PAGEREF _Toc163954010 \h  64  

  HYPERLINK \l "_Toc163954011"  5.1.2 	Ambient Bystander Exposure From
Regional Sources	  PAGEREF _Toc163954011 \h  67  

  HYPERLINK \l "_Toc163954012"  5.1.2.1	Exposures From Regionally
Targeted Ambient Air Monitoring	  PAGEREF _Toc163954012 \h  68  

  HYPERLINK \l "_Toc163954013"  5.1.2.2	Exposures From Urban Background
Ambient Air Monitoring	  PAGEREF _Toc163954013 \h  71  

  HYPERLINK \l "_Toc163954014"  5.2	Bystander Risk Characterization	 
PAGEREF _Toc163954014 \h  74  

  HYPERLINK \l "_Toc163954015"  6.0  Occupational Exposure	  PAGEREF
_Toc163954015 \h  78  

  HYPERLINK \l "_Toc163954016"  6.1	Pre-plant Agricultural Field
Fumigations	  PAGEREF _Toc163954016 \h  79  

  HYPERLINK \l "_Toc163954017"  6.2	Greenhouse Fumigations	  PAGEREF
_Toc163954017 \h  80  

  HYPERLINK \l "_Toc163954018"  6.3	Residential Fumigation	  PAGEREF
_Toc163954018 \h  81  

  HYPERLINK \l "_Toc163954019"  6.4	Occupational Risk Characterization	 
PAGEREF _Toc163954019 \h  82  

  HYPERLINK \l "_Toc163954020"  7.0  Data Needs and Label Requirements	 
PAGEREF _Toc163954020 \h  84  

  HYPERLINK \l "_Toc163954021"  7.1	Toxicology	  PAGEREF _Toc163954021
\h  84  

  HYPERLINK \l "_Toc163954022"  7.2	Residue Chemistry	  PAGEREF
_Toc163954022 \h  84  

  HYPERLINK \l "_Toc163954023"  7.3	Occupational and Residential
Exposure	  PAGEREF _Toc163954023 \h  84  

 

Appendices

Appendix A:	Toxicity Profile

Appendix B:	Hazard Assessment Array

Appendix C: 	Summary Air Model Descriptions

Appendix D:	PERFUM Output Summary Spreadsheets For Pre-Plant Soil
Uses1.0  	Executive Summary tc \l1 "1.0  	Executive Summary 

The Health Effects Division (HED) of EPA's Office of Pesticide Programs
has evaluated the methyl bromide database and conducted a human health
risk assessment for the reregistration of the chemical.  This assessment
begins phase 5 (public participation period) of the 6 phase public
participation process for the non-food uses of methyl bromide that have
not already been addressed in the previous risk assessment (D304623,
3/10/06) and its associated Addenda (D304619, 7/12/06).  [Note:  See   
HYPERLINK "http://www.Regulations.gov"  www.Regulations.gov  ,docket
OPP-2005-0123 for further information.]  In many cases, such as for the
non-food greenhouse and residential uses addressed herein, the
information contained in D304619 directly applies so it should be
considered a companion document for this current assessment.  

Methyl bromide is a broad-spectrum fumigant chemical that can be used as
an acaricide, antimicrobial, fungicide, herbicide, insecticide,
nematicide, and vertebrate control agent.  The most prevalent use
pattern is as a soil fumigant; however, it is also used as a structural
fumigant and for post harvest treatment of commodities.  Methyl bromide
application methods and equipment vary depending upon the setting. 
Under the accords of the Montreal protocol, methyl bromide is scheduled
for phase out; however, critical use exemptions will still be available
for use under special circumstances

Acutely, methyl bromide is a low to moderate toxicant via the oral and
inhalation routes of exposure (Toxicity Categories II and IV,
respectively).  In contrast, methyl bromide is highly irritating via
both dermal and ocular routes of exposure (Toxicity Category I). 
Neurotoxicity is the most common toxic effect for inhalation exposure,
with neurotoxic effects seen throughout the data base in all tested
species of animals.  Both acute and 90-day inhalation neurotoxicity
studies in rats showed evidence of neurotoxic effects characterized by
decreased activity, tremors, ataxia and paralysis.  Neurotoxic effects
were also seen in the chronic/carcinogenicity inhalation study in mice
(ataxia, limb paralysis, degenerative changes in the cerebellum), the
developmental inhalation study in rabbits (lethargy, right side head
tilt, ataxia), and the Developmental Neurotoxicity Study [DNT]
(decreased motor activity).   In addition, two subchronic studies showed
dogs to be the most sensitive species to the neurotoxic effects of
methyl bromide.    Risk assessment endpoints for the general population
were based primarily on neurotoxic effects.

For acute inhalation risk assessments, the developmental rabbit study
was selected since the fetal effects are presumed to occur after one
exposure.  In the case of short and intermediate risk assessments, two
subchronic inhalation toxicity studies in dogs were assessed together
for endpoint selection.  The chronic/carcinogenic inhalation study in
rats was selected for long term inhalation risk assessment. A NOAEL was
not identified in this study.  Thus, a LOAEL based on nasal lesions with
basal cell hyperplasia was used as the point of departure (POD). 
Consequently, a 3x uncertainty factor was applied for the LOAEL to NOAEL
extrapolation.

Methyl bromide has been responsible for a number of incidents involving
large clusters of people.  The need for Hazmat teams, decontamination,
and medical care make these cases significant, even though symptoms are
often minor.  Methyl bromide exposure has caused symptoms such as
headache, malaise, weakness, difficulty breathing (dyspnea), 
convulsions, and severe skin burns in many of these incidents. 
Incidents have been associated with faulty containers and application
equipment.  Methyl bromide has also been responsible for a significant
number of deaths, most involving individuals not directly involved in
the application.  Factors identified in the more serious cases included
lack of training and proper protective equipment, fumigation of tree
holes, inadvertent exposure to leaking structures or structures with
unexpected conduits or openings, and working in soil or other areas
where residues remained.  An updated review of recent methyl bromide
incidents report in three separate data sources shows that: there are
fewer methyl bromide incidents with declines in recent years
(2002-2005), possibly due to a combination of changing use patterns,
State/local regulatory changes, and/or better worker education and
outreach. There are still prevention opportunities until incidents are
eliminated, as evidenced by 413 new methyl bromide incidents in
California between 2002 and 2004.

Releases of fumigants such as methyl bromide fall into two categories. 
The first is used to address bystander exposures from single known
application sites such as area (i.e., treated farm fields) or point
(e.g., stacks from a greenhouse fumigation) sources.  The second is to
address exposures from many applications within a region (i.e., from
evaluating ambient air monitoring data).  Risks from known single
sources were evaluated using monitoring data and modeling techniques. 
Risks from ambient air were evaluated solely on the basis of monitoring
data from California.

When considering the potential risks of bystanders for single
application sites that encompass single known sources (e.g., area
sources such as farmfields and point sources such as stacks from
greenhouses) it is also important to understand that this has been an
iterative process that reflects the evolution of HED’s methodologies
for calculating the potential risks associated with fumigant use.  There
are a number of volatility studies which quantified methyl bromide
emissions from treated fields and facilities.  Many of these indicate
that there can be risks of concern associated with methyl bromide use. 
However, these data are limited in their utility because they provide
results only for the specific conditions under which the experiment was
conducted.  Therefore, to provide more flexibility, different modeling
systems for risk assessment purposes was instituted in order to develop
a better understanding of the potential for risks associated with methyl
bromide use under varied conditions.  The EPA's Industrial Source
Complex: Short-Term Model (ISCST3) was first used to develop risk
estimates for bystanders associated with Methyl bromide uses
(http://www.epa.gov/scram001/).  In response to HED’s ISCST3
methodologies for assessing pre-plant soil fumigants,  additional air
models that all at that time used ISCST3 as their core processors but
that also incorporated weather and emissions variability over time
(PERFUM, FEMS, SOFEA) were evaluated by the Agency for suitability which
included a review of each by the FIFRA SAP
(http://www.epa.gov/scipoly/sap/2004/index.htm - see Aug. & Sept.).  The
SAP concluded that each of the three models could provide scientifically
defensible results for risks associated with soil fumigation practices
and also suggested modifications and additional data that could further
refine risk estimates.  Since that time, other modeling options have
also been considered including CALPUFF and AERMOD. PERFUM has been used
to evaluate bystander risks because it provides more breadth of
appropriate information for risk managers than ISCST3 does but it is
clear that other models could be used for similar purposes.  Submissions
based on the other aforementioned models (e.g., FEMS or CALPUFF), or
other applicable and valid publicly available model, would be considered
by the Agency in the course of developing its risk management decisions
in context with the results of this assessment provided sufficient
supporting information were also provided in order to document such
analyses.  

As indicated above, PERFUM was used in this assessment to evaluate the
potential risks from methyl bromide uses because PERFUM incorporates
actual weather data and flux distributions estimates then accounts for
changes relative to the time of day and altering conditions.  It is also
capable of providing distributional outputs for varying receptor
locations and using varied statistical approaches.  At the upper
percentiles of the exposure distributions generated with PERFUM, the
results are markedly similar to those calculated with ISCST3.  The power
of using a system such as PERFUM, however, is inherent in the capability
of providing outputs that can be used to examine the range of exposures
one would expect based on the distributions it calculates.  It is also
clear that many different factors can impact the air concentrations (and
hence, risks) in proximity to sources that have been treated with methyl
bromide; these include many of the factors which have been investigated
in this assessment.  It is also important to acknowledge this issue so
that stakeholders understand that the results of this analysis can be
interpreted in many ways depending upon the factors which are
considered.  Many conclusions can be drawn but the key ones include: (1)
at the edge of the treated fields or in proximity to a treated structure
that NOAEL HECs generally are not exceeded given proper use of methyl
bromide (i.e., with no uncertainty factors applied such as for
inter-species variation from rats to humans and intra-species
variability within humans) but conversely the distance predicted for
MOEs between 10 and 30 are often times hundreds of meters for many
scenarios where the appropriate uncertainty factors have been applied;
(2) the methods used to evaluate methyl bromide exposure in this
assessment generally agree and they are based on techniques that have
been routinely used for regulatory purposes, they have also undergone a
significant level of review; (3) the sensitivity of results to changes
in key factors such as flux and meteorological conditions is generally
within a factor 3X to 5X based on the varied inputs which have been
evaluated but this could differ given a different set of inputs (e.g.,
flux from a cooler climate with high organic content soil); (4) PERFUM
is an empirically based approach so the generation of additional flux
and meteorological data would allow a broader analysis that could be
applied more specifically to other regions of the country and
application techniques; and (5) the identification of a result, per se,
for any sort of regulatory action would depend upon careful
consideration of the variability and uncertainty associated with each as
well as any particular merits of the inputs associated with each.

With regard to exposure from ambient air, HED has reached similar
conclusions to that of CDPR in that there are no imminent health
concerns from methyl bromide levels in ambient air. In this analysis,
HED considered both targeted monitoring data from high use areas during
the season of use (i.e., known as CARB data) and data meant to establish
background concentrations in urban environments (i.e., known as TAC
data).  Exposures for all durations ranging from acute to chronic were
considered from methyl bromide levels in ambient air.  Regardless of the
data considered, risks do not exceed HEDs level of concern for acute,
short- and intermediate-term exposures.  HED calculated chronic exposure
based on extrapolating seasonal CARB data because monitoring data
specifically meant to establish chronic exposure levels in high use
areas were not available.  Based on this approach, in some cases,
chronic risks exceed HEDs level of concern; however, HED believes that
these results do not pose an imminent health concern to the general
public due to the nature of the calculations.  The results do however
indicate a need for additional monitoring data for this scenario. 
Chronic exposures in urban environments were not of concern.

An extensive worker monitoring database was used for the evaluation of
the risks associated with various occupational tasks that include the
following: pre-plant field fumigation (e.g., tractor drivers, co-pilots,
shovelers and tarp venters and removers) as well as for greenhouse and
residential applications (e.g., remote applicators and aerators).  Tarp
removal operations generally occurred between 4 and 6 days after
application and most data represent exposures from fields covered with
high barrier films.

Overall, data indicate that worker risks exceed the level of concern for
all scenarios considered without respiratory protection (i.e., MOEs<30).
 If appropriate respiratory protection was used (i.e., air purifying
respirators [APRs] with a methyl bromide specific canister for field and
greenhouse uses or self contained breathing apparatus [SCBA] for
residential uses), results were mixed. When air purifying respirators
were evaluated with maximum exposure levels in order to assess acute
exposures for field and greenhouse workers, in most cases, exposures
were not reduced sufficiently to address risk concerns.  When APRs were
evaluated with mean exposure levels in order to assess short- or
intermediate-term exposures, the results were varied with all greenhouse
scenarios resulting in risks still exceeding the level of concern but
risks from most field uses not being of concern.  Risks for workers from
exposures during residential treatments were not of concern if SCBA is
used.

2.0	Ingredient Profile tc \l1 "2.0	Ingredient Profile 

Methyl bromide is a broad-spectrum fumigant chemical that can be used as
an acaricide, antimicrobial, fungicide, herbicide, insecticide,
nematicide, and vertebrate control agent.  Most use is on terrestrial
non-food use sites but other commonly treated sites include indoor food
and non-food use sites, residential settings, and commercial/industrial
facilities.  Approximately 47 million total pounds were applied annually
during the years 1990 through 1999.  Pre-plant field uses in agriculture
accounted for about 41 million pounds per year while post-harvest
commodity treatments accounted for another 4 million pounds and
structural fumigations accounted for 2.3 million pounds per year.  A
2001 update to that analysis for pre-plant soil fumigation of selected
crop (tomatoes, strawberries, onions, and selected tree fruits and
melons) for 8 major use states (CA, FL, NC, SC, MI, GA, WA, OH)
indicates that 2001 use (22.4 million pounds) was 40 percent of the 1991
baseline (56.2 million pounds) for those crops and locations. 
Strawberries, eggplant, peppers, and tomatoes are the crops with the
highest percentage of their overall acreage treated.  The average annual
percent crop treated for those crops, respectively, was 54, 43, 17, and
13 percent while the maximum percent crop treated, respectively, for
those crops was 70, 75, 19, and 21 percent.  Most crops were treated
once per year and the average application rate for crops (lb
ai/acre/application) ranged from a low of 5 lb ai/acre on cotton to a
high of 260 lb ai/acre/application on cucumbers.  Common pre-plant
agricultural field uses for various crops have maximum application rates
that range from the 200 lb ai/acre/application range up to around 430 lb
ai/acre/application (e.g., 5785-4 and 5785-42).  Very high rates such as
the 890 lb ai/acre/application are generally reserved for more
specialized applications such as tree planting scenarios.  The treatment
of perishable goods used 2,290,000 lb/year while durable good treatments
and quarantine uses accounted for 1,373,000 and 530,000 lb/year,
respectively.  The use of methyl bromide as a structural fumigant is
waning because of the availability of alternatives.  Annual use averaged
2,300,000 lb/year with facilities and food handling establishments
accounting for 755,000 lb/year; residential/museum/antique treatments
accounting for 1,373,000 lb/year; and transport vehicles accounting for
another 160,000 lb/year.  Application rates for commodity fumigations
can range from 1 to 20 lb ai/1000 ft3 but most are in the 1 to 9 lb
ai/1000 ft3 range.  Perishable commodities are generally not treated at
rates higher than 4 lb ai/1000 ft3 and typically many are treated at
about 1 lb ai/1000 ft3.  Likewise, structural fumigations can be in the
1 to 9 lb ai/1000 ft3 range.  For risk assessment purposes, a pre-plant
application rate of 430 lb ai/acre has been used since pre-plant field
applications account for the majority of methyl bromide use [see
Brom-O-Gas (5785-4 & -42) and Terr-O-Gas (5785-22), 
http://www.e1.greatlakes.com/agproduct/soil.html]  For the non-food uses
considered in this assessment (greenhouse and residential treatments)
maximum application rates of  4 lb ai/1000 ft3 and 3 lb ai/1000 ft3 were
used as the basis for risk assessment purposes (i.e., based on
Meth-O-Gas 5785-41 for rhizomes, seeds, roots, bulbs, corms and tubers
and Brom-O-Gas 5785-55 or -08 for residential uses).

Methyl bromide application methods and equipment are quite varied
depending upon the setting.  Generally, the methods and equipment fall
into three basic categories that include: (1) pre-plant agricultural
field fumigations; (2) structural fumigations to industrial, commercial
and residential sites; and (3) other specialized fumigations such as
certain tree replant uses.  Pre-plant agricultural field fumigations are
completed with highly specialized equipment that generally includes a
typical tractor outfitted with implements that are capable of carrying
and delivering methyl bromide gas then sealing the gas in the treated
soil. 

Methyl bromide has been identified as an ozone depleting chemical and
as such was scheduled for a phase-out by 2005 and it is subject to other
restrictions under the Montreal Protocol entered into by the United
States.  However, in certain situations, agronomic needs warranted its
use under the Montreal accord because of technical and economic reasons
as well as the lack of suitable alternatives.  To account for and codify
these uses, a process was established which allows for “Critical Use
Exemptions or CUEs” which are redefined on an annual basis in the
process established under the accords.  In 2005, there were 19 distinct
industry “sector” CUEs (pre-plant uses on cucurbits, tomatoes,
strawberries, etc.) that allow the United States to consume 35 percent
or so (i.e., approximately 19.7 million pounds) of the 1991 baseline
annual total amount used of approximately 56.3 million pounds.  For 2006
the number of industry “sectors” was reduced to 15 and the United
States was allowed to consume 32 percent of the 1991 baseline of methyl
bromide (i.e., approximately 18.0 million pounds).  However, for both
years the United States was allowed to use methyl bromide for quarantine
and pre-shipment uses without any controls on the amount used.

HED has closely coordinated with the California Department of Pesticide
Regulation (CDPR) during the development of this assessment since CDPR
has considerable experience and has also instituted requirements
governing methyl bromide use that are more restrictive than those
contained in current federal labels
(http://www.cdpr.ca.gov/docs/dprdocs/methbrom/mebrmenu.htm).  CDPR has
also generated a majority of the data considered in this assessment. 
CDPR assessed exposures from various types of known area sources, such
as a single treated agricultural field in proximity to residential
areas, and then instituted a number of regulations that govern the
application of methyl bromide in California for these situations.  In
California, applicators must apply for permits through local
agricultural commissioners for each use.  In each permit application,
users describe the site, crop, or harvested commodity to be treated; the
equipment to be used in the application; and any control technologies
that will be used to reduce emission such as tarps in fields.  In order
to allow methyl bromide users flexibility and reduce exposures, CDPR has
opted to use buffer zones that are determined based on the factors
included in permit applications.  This is done through a series of
look-up tables that methyl bromide users reference for their specific
permits.  These look-up tables are based on broad combinations of
application equipment and control technologies (i.e., these categories
are commonly referred to as permit conditions).  These permit conditions
serve as the basis for much of this assessment; for example, the
categories of application equipment and control technologies contained
in the CDPR permit conditions are the basis for the Agency’s
dispersion modeling as the same categories and emission rates were used.
[Note: Further discussion is provided below about how the modeled
estimations were completed.] CDPR is also considering how to reduce
nonpoint or ambient sources of exposure.  In order to do this, a
proposal is currently being considered which sets an exposure limit on a
regional basis (i.e., 6 by 6 mile townships are used).   For workers,
CPDR has also instituted restrictions for entry into treated fields and
facilities based on the time after application or aeration patterns. 
These have been included in the discussion of occupational risks for
comparative purposes.



2.1	Structure and Nomenclature tc \l2 "2.1	Structure and Nomenclature 

Table 1 provides the structures and relevant nomenclature for methyl
bromide.

Table 1: Test Compound Nomenclature



Properties	

Methyl Bromide

  SEQ CHAPTER \h \r 1 Chemical Structure	





Chemical Group	

Alkyl Bromide



Common Name	

Methyl Bromide



Molecular formula	

CH3Br



Molecular Weight	

94.94



CAS No.	

74-83-9



PC Code	

053201



2.2	Physical and Chemical Properties tc \l2 "2.2	Physical and Chemical
Properties 

A listing of the physical and chemical properties of methyl bromide is
provided in Table 2.

Table 2: Physical and Chemical Properties of Methyl Bromide



Parameter	

Methyl Bromide



Appearance	

colorless, odorless gas at normal temperatures and pressures and a
liquified gas under moderate pressure



Boiling Point	

3.6 (C



Vapor Pressure	

1400 mm Hg at 20 (C



Partition Coefficient	

(log Pow) 1.19



Solubility in Water	

1.75 g/100 mL at 20 (C



3.0	Hazard Assessment and Characterization tc \l1 "4.0	Hazard
Assessment and Characterization 

3.1	Hazard Characterization tc \l2 "4.1	Hazard Characterization 

3.1.1	Database Summary tc \l3 "4.1.1	Database Summary 

Studies available and acceptable (animal, human, general literature)

Data are available for both oral and inhalation routes and have been
used accordingly in the risk assessments (Appendix A).  The inhalation
database includes: acute neurotoxicity study in rats, developmental
neurotoxicity in rats, developmental toxicity studies in rats and
rabbits, subchronic toxicity studies in rats and dogs, and chronic
studies in mice and rats.  Four studies conducted via the oral route in
rats and dogs were also available. 

Sufficiency of studies/data

The toxicological database for methyl bromide is sufficient for risk
assessment purposes and includes a Developmental Neurotoxicity Study in
Rats conducted via the inhalation route.  

3.1.2	Endpoints

 tc \l3 "4.1.2	Endpoints 

Acutely, methyl bromide is a low to moderate toxicant via the oral and
inhalation routes of exposure (Toxicity Categories II and IV,
respectively).  In contrast, methyl bromide is highly irritating via
both dermal and ocular routes of exposure (Toxicity Category I).

Neurotoxicity is the most common toxic effect for inhalation exposure,
with neurotoxic effects seen throughout the data base in all tested
species.  Both acute and subchronic inhalation neurotoxicity studies in
rats showed evidence of neurotoxic effects of methyl bromide
characterized by decreased activity, tremors, ataxia and paralysis.  Two
subchronic studies demonstrated dogs to be the most sensitive species to
the neurotoxic effects of methyl bromide.  Neurotoxic effects were also
seen in the chronic/carcinogenicity inhalation study in mice (ataxia,
limb paralysis, degenerative changes in the cerebellum), the
developmental inhalation study in rabbits (lethargy, right side head
tilt, ataxia), and the developmental neurotoxicity study in rats
(decrease in motor activity ).  Developmental effects described as
increased incidence of agenesis of the gallbladder and fused sternebrae
were also seen in the developmental inhalation study in rabbits.  In
addition, the multi generation reproduction toxicity study in rats
revealed that methyl bromide exposure via the inhalation route resulted
in decreases in pregnancy rates and body weights (pups and adults).

Four studies conducted via the oral route are available in the methyl
bromide database.  Since methyl bromide is a gas under standard
atmospheric conditions, in dietary studies the test article was
administered micro encapsulated, with the exception of one study where
the feed was fumigated.  Effects noted after dietary exposure were
primarily decreases in body weight gain, body weight, and food
consumption.  Evidence of stomach lesions was seen in the 90 day oral
toxicity study in rats.  

Several studies in the database indicate that methyl bromide is a
genotoxic agent.  However, no indications of carcinogenesis were
observed in the rodent bioassays.  

3.1.3	Dose-response tc \l3 "4.1.3	Dose-response 

The general public may be exposed to fumigants in air because of their
volatility following application.  Specifically, fumigants can off-gas
into ambient air and can be transported off-site by wind to
non-agricultural areas.  Based on air monitoring studies, exposures can
be acute (less than 24 hours), short-term (1-30 days), intermediate-term
(1 month-6 months), and/or long-term (> 6 months) in duration.  In
addition, the U.S. population may be exposed to methyl bromide through
dietary intake.   The risk assessment associated with dietary exposure
was completed on July, 2006 (D304623, 3/10/06 & its Addenda D304619) and
will not be addressed in this document.

3.1.3.1  Inhalation Exposure tc \l4 "4.1.3.1  Inhalation Exposure 

The critical effects of methyl bromide exposure via the inhalation route
are agenesis of the gall bladder and fused sternebrae observed in the
developmental toxicity study in rabbits, neurotoxicity effects, and
nasal histopathology observed in the chronic toxicity/carcinogenicity
study in rats.  In evaluating the risks that a compound may pose to
human health after exposure via the inhalation route, different
methodologies have been historically used by the U.S. EPA and CDPR.  The
two approaches differ in their use of species-specific parameters to
derive HECs.  Therefore, the differences noted in the risk assessments
of each organization are due, in part, to their use of different
methodologies and uncertainty factors (UFs).  HED’s approach to
estimating risks due to inhalation exposure is based on the guidance
methodology developed by ORD for the derivation of inhalation reference
concentrations (RfCs) and human equivalent concentrations (HECs) for use
in MOE calculations.  An example of CDPR’s methodology, and the
species-specific parameters used in this approach can be found in the
CDPR website and their methyl bromide risk assessment, Appendix G at the
following web address
(www.cdpr.ca.gov/docs/dprdocs/methbrom/append_g.pdf).  As OPP
understands the importance to harmonize, to the extent possible, with
other regulatory agencies, this risk assessment will present HECs
derived using both methodologies.

For this risk assessment, endpoint selection will be based on the
endpoints occurring at the lowest HECs (which may or may not be the
lowest animal NOAEL) derived using the RfC methodology.  In this
methodology, different HECs may be calculated for the same experimental
NOAEL due to: 1) the different algorithms used to derive HECs for
systemic versus portal of entry effects; or 2) the time adjustments
conducted for non-occupational (commodity treatment facility bystander
or agricultural setting bystander) versus occupational exposure
scenarios.  The differences between systemic versus portal of entry
effects, arise from the use of different calculations to estimate the
inhalation risk to humans which are dependent on the regional gas dose
ratio (RGDR).  In the case of systemic versus portal of entry effects,
different RGDRs are derived for each type of toxicity.  For agricultural
bystander exposure (i.e., non-occupational) versus worker exposure
(i.e.,occupational), the differences arise because while it is presumed
that non-occupational exposure may occur 24 hours/day, 7 days/week;
occupational exposure occurs only during the course of an average
workweek (8 hours/day and 5 days/week).  For commodity bystanders (i.e.,
non-occupational) exposed as a result of commodity fumigation in
treatment facilities, it is presumed that exposure may occur during the
course of an average workweek (8 hours/day and 5 days/week) while the
treatment facility is in operation.

For further details on the critical studies used for endpoint selection
refer to the Toxicology Chapter of the reregistration eligibility
decision (RED) prepared by Dr. Paul Chin (DP Barcode: D271581,
Submission: S586801, dated March 18, 2003).  Additional information on
the methodologies used in this risk assessment and HEC arrays is
available in Appendix  B.

Acute Inhalation Exposure

In a developmental toxicity study (MRID No. 41580401), pregnant New
Zealand White rabbits (26 animals/dose) were exposed by whole body
inhalation to 0, 20, 40 or 80 ppm methyl bromide vapor for 6 hr/day on
Days 6-16 of gestation.  

The maternal NOAEL is 40 ppm (HEC = 10 ppm for agricultural bystander
exposure or 30 ppm  for occupational and commodity bystander exposure )
and the LOAEL is 80 ppm based on decreased appetite, lethargy, right
side head tilt, ataxia and lateral recumbency.

The developmental toxicity NOAEL is 40 ppm (HEC = 10 ppm for
agricultural bystander exposure, 40 ppm for greenhouse/structural
bystander and commodity bystander exposure or 30 ppm for occupational
exposure ) and the LOAEL is 80 ppm based on agenesis of the gall bladder
and increased incidence of fused sternebrae which was supported by
decreased fetal body weight (statistically not significant).  

Dose and Endpoint for Risk Assessment: HEC of 10 ppm for agricultural
bystander exposure, 40 ppm for greenhouse/structural bystander and
commodity bystander exposure or 30 ppm for occupational exposure, based
on agenesis of the gall bladder and increased incidence of fused
sternebrae.  It is presumed that these developmental effects may be the
outcome of an acute exposure thus this study is considered appropriate
for this risk assessment.  Acute and developmental neurotoxicity studies
in rats were available for consideration, however, the developmental
toxicity study in rabbits was selected since it yields the lowest HEC
(most health-protective) presumed to occur after an acute exposure.  
Although the DNT would yield a lower HEC for the effect of decreased
motor activity, this effect was not considered to be related to a single
MeBr exposure since it was observed on PND21 and not at earlier time
points (i.e., no compound-related effects changes in motor activity on
PND 13, 17).   A 30X UF defines HED’s level of concern (3X
interspecies extrapolation and 10x intraspecies variation) in accordance
with guidance provided in the RfC methodology (see section 4.2 below).

Short and Intermediate Inhalation Exposure

Short and intermediate inhalation risk assessments were based on two
subchronic inhalation toxicity studies in dog.  In a subchronic (5- to
7-week) inhalation toxicity study (MRID 43386802), methyl bromide
(tech., 100% a.i.) was administered 7 hours/day, 5 days/week to 4 beagle
dogs/sex/dose by whole body exposure at target concentrations of 0, 5,
10/150, 25, 50 or 100 ppm (actual mean 

concentrations 0, 5.3, 11.0/158.0, 26.0, 53.1 or 102.7 ppm; equivalent
to 0, 0.021, 0.043/0.614, 0.101, 0.206 or 0.399 mg/L).  The systemic
toxicity NOAEL for 5 and 7 weeks is 26 ppm (HEC =5.41 ppm for
agricultural bystanders or 22.75 ppm for occupational and commodity
bystander exposure).  The LOAEL is 53.1 ppm based on decreased activity.


In a six-week nonguideline inhalation toxicity study (MRID 45722801),
four groups of beagle dogs consisting of 4 males and 4 females/group
were administered methyl bromide (Lot No: 1010PK15A; purity: 100% a.i.)
by whole body exposure at concentrations of 0, 5.3, 10, and 20 ppm
(equivalent to 0, 1.8, 3.4 and 6.9 mg/kg/day).  The exposures were for
seven hours/day, five days/week for six weeks (total of 30 exposures).  

The NOAEL is 5.3 ppm (HEC = 1.0 ppm for agricultural bystander exposure
or 4.4 ppm commodity bystander or occupational exposure), and the LOAEL
for methyl bromide is 10 ppm based on the absence of proprioceptive
placing and the increased incidence of feces-findings (soft, mucoid
feces, and/or diarrhea).  

Dose and Endpoint for Risk Assessment:  HEC = 1.0 ppm for agricultural
bystander exposure or 4.4 ppm for commodity bystander or occupational
exposure based on the absence of proprioceptive placing and the
increased incidence of feces-findings (soft, mucoid feces, and/or
diarrhea).  This study is of the appropriate duration for these risk
assessments and yield the lowest HECs of the studies of this duration. 
An UF of 30X defines HED’s level of concern in accordance with
guidance provided in the RfC methodology (see section 4.2 below). 

Chronic Inhalation Exposure

In a chronic toxicity/carcinogenicity study (MRIDs 41213301, 42418301,
44359101), 50 Wistar (Cpb:Wu) rats/sex/dose were exposed to methyl
bromide (>98.8% a.i.) by whole body exposure at concentrations of 0, 3,
30 or 90 ppm (0, 0.0117, 0.117 or 0.335 mg/L) for 127 weeks (males) or
129 weeks (females).

No NOAEL was identified for local respiratory effects.  The LOAEL for
local respiratory irritation is 3 ppm (HEC = 0.13 ppm for agricultural
bystander exposure or 0.55 ppm for occupational and commodity bystander
exposures) based on increased incidence of basal cell hyperplasia of the
nasal cavity in both sexes.

The NOAEL for systemic toxicity is 30 ppm (HEC =5.36 ppm for
agricultural bystanders or 22.5 ppm for occupational and commodity
bystander exposures).  The LOAEL is 90 ppm based on increased mortality,
decreased body weight and relative brain weight, hemothorax, increased
incidence of thrombus, cartilaginous metaplasia, myocardial degeneration
and irritation of the esophagus and forestomach. 

Dose and Endpoint for Risk Assessment:  HEC of 0.13 ppm for agricultural
bystander exposure or 0.55 ppm for occupational and commodity bystander
exposures based on increased incidence of basal cell hyperplasia of the
nasal cavity in both sexes.  This study is of the appropriate duration
and yields the lowest HECs for this risk assessment.  Since a NOAEL was
not identified for the effect of concern (nasal histopathology) a 3X UF
for LOAEL to NOAEL extrapolation is recommended.  Thus an UF of 100X (3X
interspecies extrapolation, 10X intraspecies variation, and 3X LOAEL to
NOAEL extrapolation) defines HEDs level of concern in accordance with
guidance provided in the RfC methodology (see section 4.2 below). 

3.1.3.2  Dermal Exposure tc \l4 "4.1.3.3  Dermal Exposure 

Under proper use practices, dermal exposure to methyl bromide of any
significance is not expected based on the delivery systems used (e.g.,
soil injection or drip irrigation), packaging (i.e., pressurized
cylinders), and emission reduction technologies (e.g., tarping).  The
high vapor pressure of methyl bromide also makes significant dermal
exposure unlikely and quantifying any potential low level exposures very
difficult.  Therefore, a quantitative dermal exposure assessment has not
been completed. Though incidents resulting in skin burns have been
reported, these are typically associated with faulty containers or
application equipment and are not expected to occur in the course of a
typical MeBr application.   Since HED does not have adequate data to
quantify dermal risk, PPE for dermal protection should be based on the
acute toxicity of the end-use product as described in the Worker
Protection Standard and mitigation measures for dermal exposure
described in PR Notice 93-7.

3.1.3.3  Classification of Carcinogenic Potential tc \l4 "4.1.3.4 
Classification of Carcinogenic Potential 

At this time, methyl bromide is classified as a not likely human
carcinogen; consequently, no q1* or cancer risk quantification is
required.  

3.1.4	Endocrine Disruption tc \l3 "4.1.4	Endocrine Disruption 

EPA is required under the FFDCA, as amended by FQPA, to develop a
screening program to determine whether certain substances (including all
pesticide active and other ingredients) "may have an effect in humans
that is similar to an effect produced by a naturally occurring estrogen,
or other such endocrine effects as the Administrator may designate." 
Following the recommendations of its Endocrine Disruptor Screening and
Testing Advisory Committee (EDSTAC), EPA determined that there were
scientific bases for including, as part of the program, the androgen and
thyroid hormone systems, in addition to the estrogen hormone system. 
EPA also adopted EDSTAC’s recommendation that the Program include
evaluations of potential effects in wildlife.  For pesticide chemicals,
EPA will use FIFRA and, to the extent that effects in wildlife may help
determine whether a substance may have an effect in humans, FFDCA
authority to require the wildlife evaluations.  As the science develops
and resources allow, screening of additional hormone systems may be
added to the Endocrine Disruptor Screening Program (EDSP).  When the
appropriate screening and/or testing protocols being considered under
the Agency’s EDSP have been developed, methyl bromide may be subjected
to additional screening and/or testing to better characterize effects
related to endocrine disruption.

	3.2	Uncertainty Factors tc \l2 "	4.2	Uncertainty Factors 

When conducting inhalation risk assessments, the magnitude of the UFs
applied is dependent on the methodology used to determine the
appropriate point of departure.  This risk assessment is based on the
RfC methodology developed by ORD for the derivation of inhalation RfCs
and HECs for use in MOE calculations.  Since the RfC methodology takes
into consideration many pharmacokinetic (PK) differences but not
pharmacodynamic (PD) differences, the UF for interspecies extrapolation
may be reduced to 3x (to account for the PD differences) while the UF
for intraspecies variation is retained at 10x.  Thus, the UF when using
the RfC methodology is customarily 30x. 

Uncertainty factors may also be applied to account for LOAEL to NOAEL
extrapolations.  In the case of methyl bromide, no NOAEL was identified
for the portal of entry effects observed in the chronic/carcinogenicity
inhalation study in rats that was used for the long-term inhalation risk
assessment.  Since the effects noted at this dose level were not severe,
an uncertainty factor of 3x was applied for the LOAEL to NOAEL
extrapolation.

3.3	Summary of Toxicological Endpoint Selection tc \l2 "4.3	Summary of
Toxicological Endpoint Selection 

 Table 3: Summary of Toxicological Dose and Endpoints for Use in MeBr
Human Health Inhalation Risk Assessment



Risk Assessment*

	

Study	

NOAEL/LOAEL	

Endpoints	

HED

HECs	

CPDR 

HECs¶



Acute	

Agricultural Bystander (ambient; 24 hr exposure)	

Developmental Study in Rabbits	

NOAEL = 40 ppm

LOAEL = 80 ppm	

Developmental effects: agenesis of gallbladder, fused sternebrae	

10 ppm

UF = 30	

21 ppm

UF = 100

(child & adult)

	Greenhouse/Structural

&Commodity Bystander	

Developmental Study in Rabbits	

NOAEL = 40 ppm

LOAEL = 80 ppm	

Developmental effects: agenesis of gallbladder, fused sternebrae	40 ppm

UF = 30





Occupational 	

Developmental Study in Rabbits	

NOAEL = 40 ppm

LOAEL = 80 ppm	

Developmental effects: agenesis of gallbladder, fused sternebrae	

30 ppm

UF = 30

	

Short- and Intermediate-Term Inhalation (1 day to 6 months)

	

Ambient	

Subchronic (5- to 7-week) inhalation toxicity study - dogs 	

NOAEL = 5 ppm 

LOAEL = 10ppm	

Decreased responsiveness in females, fecal effects and eye irritation 	

1.0 ppm 

UF = 30	

0.88 ppm UF = 100 (child)

	

Occupational 	

Subchronic (5- to 7-week) inhalation toxicity study - dogs 	

NOAEL = 5 ppm 

LOAEL = 10ppm	

Decreased responsiveness in females, fecal effects and eye irritation 	

 4.4 ppm

 UF = 30

	

1.56 ppm

UF = 100

(adult)



Long-Term Inhalation (>6 months)

	Ambient	

Chronic/Carcinogenicity - rats	

No NOAEL identified. LOAEL = 3 ppm	

Nasal lesions	

0.13 ppm

UF = 100	

0.1 ppm

UF = 100 (child)

	

Occupational 	

Chronic/Carcinogenicity - rats	

No NOAEL identified. LOAEL = 3 ppm	

Nasal lesions	

0.55 ppm

UF = 100	

0.2 ppm

UF = 100

(adult)



Cancer 	

Classification: Not likely to be carcinogenic to humans

* Agricultural bystander HECs have also been applied to 24 hour
Time-Weighted-Average exposure concentrations measured from ambient air.
 All bystander assessments are non-occupational, by definition. 
Commodity bystanders are all based on 8 hour durations.

¶ Though CDPR and USEPA based their risk assessments on the same
critical studies and endpoints, different algorithms were used by each
organization to calculate HECs.  For further details, please refer to
Appendix B of this document.



4.0	Public Health Data

An analysis of incidents related to methyl bromide use that considered
data from the OPP Incident Data System (IDS), Poison Control Centers,
CDPR, and National Pesticide Information Center was completed in 2004
which was updated on March 14, 2007. Methyl bromide has a number of
different types of hazards associated with both agricultural and
structural applications.  It is often formulated with chloropicrin as a
warning agent and a sizeable number of cases result from the irritating
properties of the chloropicrin which can cause skin, eye, and
respiratory irritation which may result in tearing and cough.  Methyl
bromide is more likely to be involved when symptoms include headache,
malaise, weakness, difficulty breathing (dyspnea),  convulsions, and
severe skin burns.  Either chloropicrin or methyl bromide can be
associated with vomiting and diarrhea, though methyl bromide would
appear to be the more likely culprit if no odor is involved.  Methyl
bromide formulated with or without chloropicrin has been responsible for
a number of incidents involving large clusters of people.  The need for
Hazmat teams, decontamination, and medical care make these cases
significant, even though symptoms are often minor.  Incidents have been
associated with faulty containers and application equipment.  Methyl
bromide has also been responsible for a significant number of deaths,
most involving individuals not directly involved in the application. 
Fifteen deaths in California (1982-99) and 4 deaths reported in the
Incident Data System (all in Florida) involved burglars, residents, and
other persons ignoring posted warnings and breaking through the tented
covering.  There were two deaths, one in a California apartment and one
in an Iowa restaurant, where death occurred after the structure was
deemed safe to reenter.  Seven deaths were reported when persons in
adjacent structures were exposed to methyl bromide without any warning. 
In addition to these deaths, other cases of severe poisoning have been
associated with exposure in structures adjacent to those being
fumigated.  In California, nearly 70 percent of the poisonings were
occupational and half of those occurred in agricultural settings.  Of
the 278 cases attributed to methyl bromide from 1982 through 1999,
methyl bromide was definitely considered the causal agent in 42 percent
of cases, probable in 32 percent, and possible in 26 percent.  Factors
identified in the more serious cases included lack of training and
proper protective equipment, fumigation of tree holes, inadvertent
exposure to leaking structures or structures with unexpected conduits or
openings, and working in soil or other areas where residues remained.

Severe chronic effects, sometimes resulting in lifetime disability, have
been reported from methyl bromide poisoning.  For example, four such
cases included worker with slow cognition, depression, swings in mood,
weakness and persistent muscle pains; a case that was hospitalized for
16 months after exposure with paranoia and depression; and a case off
work for eight months due to fatigue and inability to carry out normal
work activities.  Nearly all of the chronic effects described above
resulted from heavy exposure and severe acute poisoning.  Other studies
of more moderately exposed workers did not reveal such effects.  For
example, a study by Calvert et al considered 123 structural applicators
in Florida and concluded “few health effects were associated with
methyl bromide exposure.”  (Calvert GM, Mueller CA, Fajen JM, Chrislip
DW, Russo J, Briggle T, Fleming LE, Suruda AJ, Steenland K.  (1998)
Health effects associated with sulfuryl fluoride and methyl bromide
exposure among structural fumigation workers.  Am J Public Health
88:1774-1780.)



The updated 2007 methyl bromide update is presented below:

California data: From 2002-2004, California occupational surveillance
data contained a total of 413 new incidents for methyl bromide with a
downward trend over time: ’02=391, ‘03=18 and ‘04=4 reported
incidents. [A similar downward trend over the same three-year period was
seen for chloropicrin (192 total incidents with ‘03=191, ‘04=1) and
metam sodium (449 total incidents, and ‘02=384, ‘03=61, and ‘04=4
incidents.]  The downward trend in incidents suggests the possibility of
positive results from worker outreach/education and/or CDPR regulatory
changes.  However, 1,3-Dichloropropene (Telone) incidents were seen to
increase at 101 incidents in ’04 only. This increase in a year when
other soil fumigant incidents are markedly down may be due to use
pattern changes. Reasons for the temporal patterns observed are being
further explored with CDPR incident data providers. Methyl bromide was
not covered in any of the five CA soil fumigant (’02-‘05) drift
studies with plume modeling.

NIOSH SENSOR data: Currently, twelve states report occupational
poisoning incidents to a central database. The states are CA, WA, OR,
NY, AZ, LA, TX, NM, FL, NC, MI, IA. For the time period ’98-’03
covering 5899 total incident cases, 33 incidents were for methyl
bromide, including 30 males and 3 females. States reporting are as
follows: CA=25 methyl bromide incidents, FL=2, TX=4 and WA=2.
Underreporting is a known problem with literature. Due to
underreporting, there are generally few duplicates found among the
multiple data sources. Duplicate incident cases are eliminated by
matching the exact dates, locations and other incident case details.

Poison Control Center (PCC) data: This is the only source of National
incident coverage, encompassing 61 poison center that report in a
standard format. Only PCC reports data on children, as well as
occupational and non-occupational cases, symptom severity and medical
outcome, including death.  For the time period ’92-’05, PCC reported
168 methyl bromide incidents, 77 occupational, 74 non-occupational, and
14 children. Of the 168 total incidnets 3 had unknown medical outcome.
There were no deaths reported in recent years.

6a2 and other data: A comprehensive incident data assessment for seven
soil fumigants, including methyl bromide is in preparation. The
interagency Agricultural Health Study (AHS) reported on an initial link
with methyl bromide in the prospective prostate cancer study (Alavanja
et al 2003, see   HYPERLINK "http://www.aghealth.org"  www.aghealth.org 
for text). The chemical was rapidly declining in use in Iowa and North
Carolina over the study period and preliminary indications from a
follow-up study of the next 500+ prostate cancer appears to be negative
for methyl bromide (personal communication Alavanja, el al. 2007). 

In summary, there are fewer methyl bromide incidents in recent years
(2002-2005), possibly due to a combination of changing use patterns,
State/local regulatory changes, and/or better worker education and
outreach. There are still prevention opportunities until incidents are
eliminated.



5.0 	Non-Occupational Exposure Assessment and Characterization tc \l1
"6.0 	Non-Occupational Exposure Assessment and Characterization 

Monitoring data indicate that methyl bromide volatilizes after
application to soil, facilities, and other premises and that inhalation
exposure is possible when individuals are in proximity to specific
application events.  Inhalation exposures are also possible from ambient
air levels if individuals live and/or work in regions where methyl
bromide is used.  Incidents associated with methyl bromide use
(especially when it is combined with chloropicrin which has more
distinct warning properties) also support the premise that inhalation
exposures can occur for those in proximity to a methyl bromide
application.  Dermal exposures in the general population are not
anticipated because of the volatility of methyl bromide and the fact
that all methyl bromide products are restricted use pesticides which
precludes direct dermal contact with the product because it is only a
liquid that could get on the skin prior to application.  Dermal
incidents, generally attributable to accidents or equipment failure,
have occurred in occupational populations who handle the concentrated
products in liquid form.  Contact with concentrated methyl bromide in
liquid form would not occur in the general population so the incident
pattern is consistent with the supposition that dermal exposures
estimates for non-occupationally exposed individuals are not warranted.

Risks from ingestion associated with the food uses of methyl bromide
(i.e., commodity and food handling establishments that require
tolerances) or those possible through drinking water were previously
addressed in the following:

Methyl Bromide:  Phase 5 Health Effects Division (HED) Human Health Risk
Assessment For Commodity Uses, PC Code: 053201, DP Barcode D304623,
Authors:  J. Dawson and E. Mendez, Issued:  3/10/06.

Methyl Bromide:  Addendum To Phase 5 Health Effects Division (HED) Human
Health Risk Assessment For Commodity Uses, PC Code 053201, DP Barcode
D304619 Authors:  J. Dawson E. Mendez, T. Goodlow, M. Metzger, Issued: 
7/12/06.

Based on the premise described above this non-occupational assessment is
limited in scope to inhalation exposures that could occur for a
bystander in proximity to specific application events or an individual
who can be exposed through ambient air concentrations.  Methyl bromide
is a highly volatile fumigant that is very mobile in soil and it can
also readily infiltrate all spaces within treated structures or
facilities. After application, methyl bromide typically volatilizes from
soil rapidly after application with a large portion of the total mass
being emitted in the first 24 hours.  This is illustrated by the
emissions profile from treated, tarped raised beds which a common
cultural practice for tomato and strawberry production in many states
(Figure 1).  In other types of treatments (e.g., greenhouses or
residential settings), then the emissions tend to be shaped by the
nature of the treated structure (i.e., is it a tight or leaky building?)
and how aeration is accomplished after treatment has been achieved
(e.g., aeration type – passive or active, can impact off-target
transport).  Once emitted into the atmosphere, methyl bromide is
sufficiently persistent so that exposures occurring within general
regions where methyl bromide is used (i.e., ambient exposures) have also
been addressed because methyl bromide has been routinely measured in
studies designed for the purposes of elucidating ambient levels of
airborne contaminants.

Figure   SEQ Figure \* ARABIC  1 

Because most mass of methyl bromide is rapidly emitted into the
atmosphere after field applications or during structural treatment
aerations, acute exposure scenarios are of key concern to residential
bystanders (i.e., those who are in the proximity of the emissions
resulting from a methyl bromide application).  Bystander exposure to
methyl bromide, or any fumigant for that matter, depends on two main
factors:  (1) the rate of emissions from a treated field or facility
into the atmosphere (described as flux) and (2) how those resulting
emissions are dispersed in the air over and around the treated field or
structure.  Emission rates from treated fields are affected primarily by
the amount of fumigant applied (which is proportional to the rate and
area treated), the application method and equipment used, sealing
technologies use to reduce emission levels, and the field conditions
where factors such as soil type, moisture, and amount of organic
material may impact emission rates.  Emissions from treated structures
can also be affected by several factors that can include the tightness
of the structure which is treated, the absorptive capacity of the
materials within the treated structure, the duration of the treatment,
and the nature of the aeration used to evacuate the structure after
treatment if complete.  Once methyl bromide, or any other fumigant, has
been emitted into the atmosphere, meteorological conditions, the
topography at the site, and/or the nature of the treated structure
determine how the fumigant is dispersed.  For example, if winds are high
and the atmosphere is unstable, then emitted fumigant concentrations are
more likely to be reduced because greater mixing and dispersion will
occur.  Under such conditions, the likelihood of a bystander being
exposed to a fumigant at a concentration of concern is relatively lower.
 On the other hand if winds are light and the atmosphere is stable, then
the emitted fumigant is more likely to build in concentration and be at
higher levels in proximity to the treatment area.  Topography, as well
as other factors, can also cause winds from certain directions to be
predominant which can predispose certain populations to higher exposure
levels (e.g., a school located in a valley where prevailing winds
approach it and the treatment area is upwind or a similar situation with
prevailing onshore coastal winds in California or Florida).

This section describes the potential exposure scenarios associated with
the use of methyl bromide.  These include residential bystander exposure
from two key sources including: known sources from a single application
site (i.e., point sources such as from a ventilation stack on a treated
commodity chamber during aeration or area sources such as at the edge of
a treated field) and ambient air levels that result from many
applications within a region.  There are no residential uses of methyl
bromide by homeowners so this aspect of the risk assessment focuses on
those types of exposures that may occur from professional uses of methyl
bromide that can lead to exposures in residential environments.  Section
5.1: Residential Bystander Exposure And Risk Estimates describes how
exposure and risk estimates were calculated for the general population
who may be exposed living in proximity to individual application sites
or within regions where methyl bromide use routinely occurs.  Section
5.2: Bystander Risk Characterization describes the factors that should
be considered when interpreting the results of this risk assessment.  

5.1	Residential Bystander Exposure and Risk Estimates tc \l2 "6.1
Residential Bystander Exposure and Risk Estimates 

Residential bystander exposure may occur because of emissions from
buildings and other structures or treated fields as indicated above.  A
tiered approach has been used to calculate risks from known sources from
a single application site that is based on various models, incident
information, and monitoring data. Ambient exposures have been addressed
solely using monitoring data since no predictive model and reliable
input data for this purpose are available at this time.

When considering the potential risks of bystanders for single
application known sources (e.g., a farm field or a structure), it is
important to note that they were developed based on an iterative process
that reflects an evolution in the methodologies used to calculate them. 
It is also important that results based on incidents, monitoring data,
and modeling be considered in conjunction with one another to ensure
consistency in the overall characterization of the risks associated with
methyl bromide use.  This approach is considered a tiered approach by
the Agency because each additional method allows for more predictive
capability to other use situations (i.e., it is less constrained by the
circumstances of the incident or particular field study).  There are a
number of volatility studies which quantified methyl bromide emissions
from treated fields and facilities.  However, these data are limited in
their utility because they provide results only for the specific
conditions under which the experiments were conducted.  In cases where
incidents associated with methyl bromide use occurred, that Agency has
also attempted to characterize them in a manner that explains the
particular event and how it relates to the overall risk picture for
methyl bromide but the degree of such analyses is also constrained by
the particulars of the event and the level of information that is
available with which to analyze it.  Models have also been used to
estimate potential risks from methyl bromide to bystanders under varying
conditions.  The first modeling approach was based on the deterministic
use of the Agency's Industrial Source Complex model (ISCST3) which
provides off-site air concentration estimates and the second approach
which is based on a distributional model called the Probabilistic
Exposure and Risk Model For Fumigants (PERFUM) which calculates
distances at which target concentrations are achieved at varied
percentiles of exposure.  PERFUM also can provide distributions of air
concentrations at varied distances from the perimeter of treated fields.
 It develops distributions based on 5 years of meteorological data.  It
can also probabilistically address emissions but for methyl bromide
insufficient information was available to utilize that function.

As indicated above, the use of monitoring data provide the only means
with which to assess the potential risks associated with methyl bromide
exposures in ambient air.  At the time this assessment was developed,
the only known monitoring data designed to quantify ambient methyl
bromide air concentrations were developed by the California Air
Resources Board.  These data have been used as the basis for this
assessment and have been cited as appropriate.

The potential risk related to exposures from single application area
(e.g., farmfields) or point (e.g., stacks from a treated structure)
sources for bystanders are described below in Section 5.1.1: Bystander
Exposures And Risks From Known Sources while the potential risks
associated with exposures to ambient air are described below in Section
5.1.2: Ambient Bystander Exposure From Multiple Regional Sources.  Each
section provides a description of the methods used and the results.

5.1.1 	Bystander Exposures and Risks From Known Sources tc \l3 "6.1.1 
Bystander Exposures And Risks From Known Sources 

As noted, residential bystander exposure may occur because of emissions
due to single applications from known sources such as treated fields or
structures.  The methods used to assess the exposures and risks related
to these uses are described below in Section 5.1.1.1:  Methods Used To
Calculate Bystander Exposures And Risks From Known Sources.  The results
calculated for all scenarios of interest based on the most appropriate
method for that scenario are presented in Section 5.1.1.2:  Bystander
Exposures And Risks From Known Sources.

5.1.1.1 	Methods Used To Calculate Bystander Exposures And Risks From
Known Sources tc \l3 "6.1.1.1 	Methods Used To Calculate Bystander
Exposures And Risks From Known Sources 

As indicated above, the Agency’s calculation of bystander exposures
and risks from known sources has been an iterative process based on the
ability to provide additional predictive capabilities yet consider all
possible sources of information that could be used to characterize the
overall risk picture associated with a chemical.  The interrelationship
of these factors is illustrated in Figure 2.  This approach is also
consistent with general Agency guidance on the use of air models.

 

Figure 2 

As indicated in Figure 2 above, three sources of information have been
used for assessing bystander risks.  Each source has a unique level of
predictive capability but each result has been carefully considered in
context with each other in order to develop an overall characterization
of the risks associated with methyl bromide use.  Each method is
described below along with a description of how they were used and how
they should be interpreted in the context.  Regardless of which approach
is utilized, it is clear that there can be possible human health effects
associated with the use of soil fumigant chemicals based on calculated
risk estimates.

Source Type 1: Field Level Monitoring Studies & Incident Data

	Incident Data -  Fumigants have been used for a number of years in
agricultural settings and for a variety of structural treatments. 
Throughout this history of use, there have been reported incidents where
bystanders have experienced noticeable symptoms from various fumigants
which have ranged from mild and reversible in nature to more serious
symptoms up to and including death.  When using incident reports to
characterize the results of fumigant risk assessments, the number of
incidents for a particular fumigant compared to its total use throughout
the country along with the causes and severity of major incident events
are considered.  Typically, the circumstances that led to a specific
incident are examined to the extent possible including factors such as
application method, sealing method, meteorological factors, and
possibility of misuse as well as others specific to each case.  It is
also important to examine the specific health effects experienced by
exposed individuals because they provide insight as to whether or not
the endpoints selected for risk assessment purposes have occurred under
adverse field conditions which could illustrate a level of consistency
among the approaches used to evaluate a chemical.  If not, it indicates
that further investigation is warranted if possible.  The incident
analysis that has been completed for methyl bromide is presented above
in Section 4:  Public Health Data.

In most cases, it is hard to reconstruct the exact conditions of
specific incidents which make predicting fumigant concentrations to
which bystanders were exposed during an incident extremely difficult. 
This is generally because all necessary information to reconstruct the
incident through predictive modeling is not available.  For example,
emission rates at the time of the incident specific to the way that
particular application was completed would not be measured so any
reconstruction for modeling purposes would have to utilize another
source of emissions data that would be applied to the specific incident
situation.  Additionally, without on-site meteorological monitoring
data, the same type of approach would have to be used in order to
develop weather inputs for modeling purposes.  In summary, without
site-specific emissions and weather data collected at or near the time
of an incident it is difficult to reconstruct the conditions of an
incident and therefore accurately predict concentrations for exposed
individuals.

The overall reliability of fumigant incident information can also be
circumspect because some toxic effects are not immediately recognizable
(e.g., developmental effects identified as the basis of acute bystander
risk analysis for methyl bromide - see Table 3).  Therefore, the
likelihood that incidents may be reported could be lower for fumigants
which have effects that are not immediately recognizable.  Accordingly,
incident reports will be considered in this context when characterizing
the risks associated with fumigants.

	Monitoring Studies - Field volatility studies typically measure
fumigant air concentrations produced by a single fumigant application
under specific conditions (e.g., application rate and method, area
treated, soil conditions, meteorological conditions).  In these studies,
air samplers positioned in and around a treated field continuously
sample air after the fumigant has been applied in order to quantify the
emissions from that specific field.  Sampling times can vary but
generally range from about 4 to 12 hours, so that the samples represent
the average air concentrations for the sampling intervals used. 
Usually, shorter times are used at the beginning because fumigants
generally off-gas the most within the first 24 hours after application
and shorter sampling times provide a better means for characterizing
peak emission periods that are expected to be associated with higher
exposures.  For methyl bromide, a number of monitoring studies were
considered in the development of the risk assessment.  These have been
described in detail in the previous assessments (D316326, 7/13/05).  An
example of the information that can be generated by a field monitoring
study is illustrated by Figure 3.  This figure summarizes the nature of
the application used in Field Study #8 and the sampling results at each
location around the perimeter of the treated field.  All sample
locations describe the distance and location relative to the field and
12-hour time-weighted average concentrations that resulted from the
application where sampling began with the initiation of the application.
 [Note:  Structural monitoring studies would provide a similar type of
result.  The only difference would be that the parameters of the
application and the nature of the emissions would differ.]

 

Figure 3 

Results based on using monitoring data directly from field volatility
studies for risk assessment purposes are summarized below.  There are
several limitations to this approach that should be considered in the
overall context of related methods available for calculating risks
associated with fumigant use.  Essentially, the monitoring data are both
spatially and temporally limited.  For example, data do not reflect the
values that would occur under different conditions.  Air concentrations
around treated fields are influenced by a number of factors including
how a chemical is applied, application rate, emission reduction methods
(e.g., tarps, water seals), soil conditions, and weather conditions. 
Varying weather conditions, for example, can significantly change the
air concentrations at specific sites around a treated area.  Since there
is such a large range of potential weather conditions which could exist,
it is not possible for monitoring studies to inherently capture the
entire range of potential exposures which could result.  Another example
would be that air concentrations are measured by fixed samplers
positioned at various distances and directions around the treated area,
both downwind and upwind, as well as at points in between.  This makes
it difficult to interpolate between sampler locations, if so desired, to
develop risk estimates in-between the locations.  Based on these
factors, the use of monitoring data for trends analysis is difficult
without a modeling approach.  This premise is consistent with the
general Agency approach for the use of air monitoring data related to
the air permitting process.  More information regarding the utility of
monitoring data and its limitations described above can be found in
Appendix W to 40CFR51which presents Agency policy related to the
selection and use of air models (  HYPERLINK
"http://www.epa.gov/scram001/guidance/guide/appw_05.pdf" 
http://www.epa.gov/scram001/guidance/guide/appw_05.pdf ).  Essentially,
monitoring data in this assessment were used in a manner consistent with
this guidance.

Source Type  2: Deterministic Air  Modeling

	Air dispersion modeling uses mathematical formulas to characterize how
atmospheric processes will disperse a pollutant emitted by a source. 
For the fumigants, the Agency has used dispersion models to estimate the
downwind concentration of fumigants emitted from sources such as treated
fields or structures for this purpose as is consistent with the guidance
provided in 40CFR51.  This treatment is consistent with standard model
development and implementation methods.  Dispersion models require the
categorization and/or input of data which includes:

Meteorological conditions such as wind speed and direction as well as
the amount of atmospheric turbulence (also known as the “stability
class”);

Flux rate (the mass of fumigant emitted per area per time);

Surface roughness (accounts for topography effects); and

Application Specifics (application method, sealing techniques,
application rate, field size, etc.).

The Agency maintains a Guideline on Air Quality Models (hereafter,
Guideline), which is published as Appendix W to 40CFR51
(http://www.epa.gov/scram001/guidance/guide/appw_05.pdf).   The
Guideline provides the Agency's guidance on the regulatory applicability
of air quality dispersion models in general.  In order to be included in
Appendix W, as a recommended model, models must go through an extensive
peer review and testing process.  This peer review process defines how
specific models can be used in an acceptable manner to calculate
dispersion estimates for variety of sources like point (e.g., a stack on
a building) and area sources (e.g., a fumigated field).  This assessment
was developed based on the guidance provided in Appendix W.

In producing the fumigant risk assessments, the Agency considered
various air dispersion models that are currently listed or have
previously been listed in Appendix W.  The first of these models is the
Industrial Source Complex Short Term Model (V3) (ISCST3) model which was
utilized for a number of years by the Agency to quantify the movement of
airborne pollutants for a variety of regulatory situations.  ISCST3 was
the Agency’s recommended air dispersion model up until the end of
2005.  It was also used as the sole basis for the earliest methyl
bromide assessments.  ISCST3 was replaced by the American Meteorological
Society/Environmental Protection Agency Regulatory Model (AERMOD) in
December of 2005 as the preferred air dispersion model for near-field,
steady state sources.  Both ISCST3 and AERMOD are “Gaussian Plume”
models, in which airborne concentrations are assumed to have a normal
probability distribution.  Figure 4 illustrates the basic premise of
ISCST3 and the Gaussian plume concept.  It should also be noted that
neither ISCST3 or AERMOD retain a memory of the movement of the fumigant
plume from hour to hour (e.g., they would not track changes in an
emitted plume should the wind direction change) and they do not
quantitatively address calm conditions.  For this assessment, a process
has been used where calm conditions (e.g., hours with calm wind
conditions) are dropped from calculations and a time-weighted average
result is calculated without those values.  This approach is consistent
with how ISCST3 has been historically used.  For chemicals such as
methyl bromide the impact on the calculated exposures due to handling
calms in this manner is attenuated because 24 hour time-weighted
averages are the basis for the results.  However, for chemicals where
risk estimates are based on shorter duration toxicity endpoints (e.g., 1
hour), this phenomena can significantly impact the results if the
weather data used in the assessment include a high percentage of calm
periods.  AERMOD has enhancements from ISCST3 related to how structural
releases are modeled such as improved downdraft algorithms for building
effects.  The third model that the Agency considered was CALPUFF v.5
which was recently adopted by the Agency as the preferred model for
assessing near-field air concentrations under complex meteorological
conditions.  CALPUFF is a “Gaussian Puff” model and is similar to
ISCST3 and AERMOD in that it assumes that air concentrations follow a
normal probability distribution.  Unlike the plume model, however,
CALPUFF retains a memory of the movement of the fumigant plume from hour
to hour which allows it to track emitted plumes that change direction
with shifting wind patterns.  It also has an enhanced treatment of calm
conditions relative to ISCST3 or AERMOD because it can account for the
plume being stable in calm conditions then moving again once winds pick
up instead of skipping over such conditions.  ISCST3, AERMOD, and
CALPUFF are described in more detail in Appendix C.  [Note:  There is a
yet unapproved version of CALPUFF (v6) which has not been officially
accepted by the Agency.  The major upgrade is that it can complete
sub-hourly calculations where v5 can only do calculations based on
hourly increments.  This is described as well in Appendix C.]  It should
be noted that the Agency used ISCST3 as the basis for its deterministic
assessments.  At the time the results based upon ISCST3 were developed
neither AERMOD nor CALPUFF v5 were approved models.  At this time, the
Agency has not opted to use them directly since neither can readily be
used in the distributional manner that is currently being employed by
the Agency as described below. The Agency would accept and review
submissions using these modeling platforms as they are accepted models
in the Agency Guideline as outlined in Appendix W.

Figure 4: Illustration of ISCST3 Gaussian Plume Approach

Before a modeling analysis can be done, one of the most important
parameters for ISCST3, the flux or rate of pesticide emissions from the
treated fields, buildings or structures per unit area per unit time,
must be determined.  As an example, for field applications it is usually
expressed in units of micrograms per square meter per second
(ug/m2/sec).  In essence, flux represents how quickly the pesticide
moves or volatilizes into the surrounding atmosphere from a treated
surface.  Numerous factors can influence flux rates such as application
rate, depth of soil injection, type of application (e.g., drip vs. soil
injection vs. granule application), techniques used to control emissions
(e.g., tarps), temperature, wind and weather conditions, soil type, and
others.  Three general methods are used to estimate flux from treated
fields.  These are discussed briefly below.  The first two methods
measure flux from sampling directly in treated fields, and the third is
an indirect, back-calculating method that estimates flux using samples
from downwind locations and solves for them using ISCST3.  For methyl
bromide, most flux estimates for pre-plant field applications were
completed using the indirect back-calculating method.  [Note:  For the
structural uses addressed in this document, flux estimates were
developed using a different approach.  Please refer to D304623, 3/10/06
& its Addenda D304619, 7/12/06 for further information.]

ISCST3 Flux Method 1: Chamber  The first method is a direct sampling
method for determining flux that uses emission data measured in a flux
chamber placed in a treated field.  A flux chamber is basically a box
which encloses a small defined area of a treated field, from which air
samples are obtained representing defined durations (e.g., air is pulled
through a charcoal trap collecting emitted pesticide over a continuous
length of time such as 4 hours).  Since the surface area is defined by
the area of the chamber, and the quantity of pesticide emitted per unit
time is defined by the air concentration, this method directly measures
flux.  A possible issue with flux chambers is that the conditions within
the chamber (e.g., temperature, wind, air stability) are not generally
identical to those outside the chamber in the treated field; since flux
rates can be significantly affected by these factors, flux rates
measured in these chambers may not always represent actual flux rates in
the field.  Flux chambers are not often used for estimating flux and, in
fact, no such field study data were available for use in this
assessment.

ISCST3 Flux Method 2: Aerodynamic Flux  A second direct method used is
known as the aerodynamic flux method.  In this method, air samplers are
set up in treated fields at various heights on a mast (e.g., 15, 30, 90,
and 150 cm from the ground).  Using measured air concentrations at these
various heights, a vertical gradient of concentrations can be estimated
for different time points which can be integrated across all heights to
estimate the flux rate at each time point after application.  Some
studies are available using this method to determine flux rates.

ISCST3 Flux Method 3: Indirect Back-Calculation  The method most often
used to determine flux rates is the indirect or back-calculation method.
[Note:  EPA used CDPR’s technique
(http://www.cdpr.ca.gov/docs/empm/pubs/ehapreps/eh9903.pdf).]  This
method uses measured air concentrations taken in a typical field
fumigation study in which air samplers are located at various positions
around the field.  The measured air concentrations, together with
information about weather conditions which occurred when the samples
were obtained, are used as inputs into the ISCST3.  The model assumes
that these air concentrations result from a Gaussian plume, the plume
being distributed around the treated field as a result of the wind and
weather conditions.  The model then estimates the flux rate that would
be required to emit the plume and to obtain the air concentrations
measured.  

Aside from the estimation of the flux for all application methods, there
are a number of other key inputs that must also be defined such as the
size and shape of a treated field or other sources (e.g., greenhouses or
fumigation chambers), wind speed, and atmospheric stability in order to
run ISCST3.  Atmospheric stability is a measure of how turbulent the
atmosphere is at any given time.  Stability is affected by solar
radiation, wind speed, cloud cover, and temperature, among other
factors.  If the atmosphere is unstable, then more off-field/source
movement of airborne residues is possible without a large increase in
air concentrations because the residues are carried up into the
atmosphere and moved away from the field or other source, thereby
lowering the air concentration in proximity to the field/source.  To
simplify the ISCST3 modeling process, the transport of fumigant vapors
from a source, a single wind direction, wind speed, and stability
category are used for a given period.

A range of atmospheric conditions representing the continuum from
relatively stable (low windspeed & calm) to unstable conditions (high
windspeeds & unsettled) were evaluated using ISCST3.  Under relatively
stable atmospheric conditions, the modeling produces results that
represent highly exposed individuals (i.e., ISCST3, as used for these
situations, results in exposure estimates at the upper percentiles of an
anticipated exposure distribution).  Two key inputs are the basis for
this conclusion.  First, only a constant downwind direction is
considered which would be highly unlikely in any outdoor environment. 
Secondly, the quantitative inputs used to define atmospheric conditions
are based on constant wind speed and atmospheric stability over a
particular period, which are also unlikely to occur in an outdoor
environment over a 24 hour period such as considered for methyl bromide
field uses.  Conversely, unsettled conditions may reduce risk estimates
but it is believed that even these conditions can result in conservative
estimates because wind direction is constrained to a single direction
over a particular period.

Source Type 3:  Distributional Air Modeling

The monitoring data and ISCST3 methods described above are deterministic
methods that provide results that are limited in utility.  For example,
it is difficult to extrapolate to varying distances using monitoring
data and analyses using ISCST3 which provide high-end point estimates of
exposure and risk because of the manner in which meteorological data are
input, especially for a stable atmosphere.  In response to these
methods, the pesticide industry developed three models that are
essentially pre- and post-processors for the air models described above
that incorporate the ability to complete distributional and/or
probabilistic analyses.  Each of the three has ISCST3 as their core
processor while FEMS has an option for selecting between processors
based on ISCST3 or CALPUFF (V 5 or 6).  The three models which were
developed include:  Probabilistic Exposure and Risk model for Fumigants
(PERFUM), the Fumigant Emissions Modeling System (FEMS), and the Soil
Fumigant Exposure Assessment System (SOFEA).  Each model was reviewed by
the FIFRA Scientific Advisory Panel (SAP) in 2004 during the August and
September meetings
(http://www.epa.gov/oscpmont/sap/meetings/2004/index.htm).  The SAP
concluded that each of the three models could provide scientifically
defensible estimates of the bystander exposures and risks associated
with soil fumigation practices and also suggested modifications and
additional data that could further refine risk estimates.  See Appendix
C for more details regarding each model including contact information
pertaining to how one could obtain the system.  

PERFUM and FEMS were designed specifically to take the concentration
outputs from the air dispersion models and use them to produce buffer
zone outputs in a distributional format.  [Note:  In the context of
presenting modeling results the term "buffer zone" does not refer to any
manner of regulatory decision pertaining to risk mitigation for methyl
bromide.  It refers to the distances determined based on a target
concentration defined by the HEC or Human Equivalent Concentration
adjusted by an uncertainty factor.  Different uncertainty factor values
were evaluated in this assessment to ascertain their impact upon the
predicted results.]

Recently, PERFUM was modified to also provide air concentration
information for selected distances from the perimeter of the treated
field.  PERFUM also has been modified since the SAP version in order to
be able to evaluate structural sources which were addressed in the
recent assessment for the food uses of methyl bromide (see D304623,
3/10/06 & its addenda D304619, 7/12/06 for further information).  SOFEA
was designed to calculate fumigant concentrations in air arising from
treated fields for multiple sources across entire agricultural regions. 
A generalized flowchart for these models is shown in Figure 5.  

 

Figure 5: Operational Flowchart For Distributional Models Such As PERFUM

		Selection of a Distributional Model -  The conclusions of the 2004 SAP
meetings were that all three of the distributional and/or probabilistic
modeling options were scientifically viable and represented a level of
refinement above the deterministic analyses that had been completed
using ISCST3.  For a number of reasons detailed below, PERFUM was
selected at that time to evaluate bystander risks from pre-plant soil
applications including:

PERFUM’s developers revised the model to incorporate some of the
SAP’s recommended changes in time for the Agency to use PERFUM in the
revised Phase 1 risk assessments for the soil fumigants;

PERFUM was significantly faster and more efficient to run than FEMS at
that point in time; and

PERFUM provided greater resolution than the other options on the period
of peak emission and highest potential exposure which is of key interest
to the Agency because of the acute toxicity associated with soil
fumigants.  At that time, FEMS used emissions from a single field over a
whole year so that the few days of fumigant exposure occurring after an
application were attenuated over that entire year.

It is believed that results from a distributional and/or probabilistic
model, instead of the deterministic results based on ISCST3, provide
more comprehensive information for risk managers when evaluating the
potential risks associated with pre-plant soil fumigation.  PERFUM
remains the model which has been used to develop the Agency's fumigant
assessments but it should be noted that the Agency believes that
submissions based on the other aforementioned
distributional/probabilistic models such as FEMS or SOFEA can be of
equal scientific validity and would also be evaluated and considered in
its risk management process provided all appropriate supporting
documentation were available for review (e.g., documentation of flux
rate calculations and weather data analysis).

For the other single application event exposure scenarios including
greenhouse and residential fumigations, PERFUM is the only one of the
three distributional models which can accommodate a non-field structural
source such as a treated building (see D304623, 3/10/06 & its addenda
D304619, 7/12/06 for further information).  The use of PERFUM for this
purpose, coupled with its use for evaluating pre-plant soil fumigations
also adds consistency to the Agency's fumigant analyses.

	Use of the Probabilistic Exposure and Risk model for Fumigants (PERFUM)
- PERFUM allows users to develop an understanding of the distributions
of potential bystander exposures and thus more fully characterize the
range of risks resulting to bystanders around treated fields.  In this
assessment, the PERFUM model has been used in order to calculate
differing percentiles of exposure associated with pre-plant soil
fumigation.  ISCST3 is an integral part of the PERFUM model and in fact
the basic physics and code of ISCST3 remain unchanged.  Many of the
inputs used for PERFUM are similar to those used for modeling done using
the ISCST3 model (e.g., field sizes and back-calculated flux rates). 
There are major differences with the inputs in a PERFUM analysis as
opposed to an ISCST3 assessment.  Each PERFUM analysis that is
summarized in this assessment is based on 5 years worth of
meteorological data and a flux profile specific to a unique methyl
bromide application method.  [Note:  The meteorological data were used
for both the pre-plant field uses of methyl bromide as well as the
structural uses.  Please refer to D304623, 3/10/06 & its addenda
D304619, 7/12/06 for further information pertaining to the flux values
used for the structural uses considered herein.]

Since actual meteorological data are integrated into PERFUM for each
analysis, data representative of the locations where methyl bromide use
occurs were identified and used in the analysis.  For example, major
uses occur on strawberries and tomatoes in Florida and California.  Some
use in Michigan (or elsewhere in that region) also occurs on various
crops.  As a result, the following locations and sources of
meteorological data were used in this assessment:

Bakersfield California (Source: ASOS or Automated Surface Observing
System operated by the FAA) to represent inland California locations;

Ventura California (Source: CIMIS or California Irrigation Management
Information System) to represent coastal California locations;

Flint Michigan (Source: NWS or National Weather Service) to represent
central Michigan and other upper midwest locations;

Tallahassee Florida (Source: NWS or National Weather Service) to
represent inland Florida locations; and

Bradenton Florida (Source: FAWN or Florida Automated Weather Network) to
represent coastal Florida.

In this assessment, 5 years or 1825 days of meteorological data were
considered in each calculation.  Bradenton, Bakersfield, and Ventura
data were in the range of 1997 through 2003 but Tallahassee and Flint
were in the late 1980s through early 1990s.  [Note: Please refer to the
SAP background documents for PERFUM for further information concerning
these data including how they were processed for incorporation into
PERFUM, pertinent quality control issues associated with the data, and
additional information related to their selection
(http://www.epa.gov/scipoly/sap/2004/index.htm).]  

Figure 6 provides a comparison of the distributions of daily average
windspeeds for selected stations in California and Florida that can help
characterize the deterministic assessments and different PERFUM results
for the different stations. [Note: For context, CDPR regulated methyl
bromide at 1.4 m/s windspeed.]

Figure 6:  Distribution of Daily Average Windspeeds At Selected
Meteorological Stations

Flux inputs (i.e., field volatility or emissions) for PERFUM were
categorized in a manner similar to that used for the ISCST3 analysis
described above.  In the ISCST3 analysis for methyl bromide,
calculations were based on the emission ratios developed by the CDPR
that categorized emissions based on application method and sealing
technology.  These formed the basis for the different CDPR permit
conditions for methyl bromide (Table 4).  Emission functions were used
in the PERFUM assessment as the basis for the calculations.  These were
also developed by CDPR to describe the change in flux rate over time in
order to define the duration of restricted entry intervals.  These
functions allow PERFUM results to account for changes in flux to occur
at appropriate times of day which is more reflective of actual field
situations.  Table 4 below also illustrates the analyses that were
completed using PERFUM based on different combinations of flux and
weather data.  [Note:  All emission ratios included in Table 4 are based
on the use of LDPE or HDPE films.  High barrier film use is not
reflected herein.]

Table 4: Summary Of PERFUM Analyses Completed For Methyl Bromide



Weather Station Location	

Flux Study Summary & Corresponding Emission Ratios and CDPR Permit
Conditions

	

Tarped

[ER = 0.25]	

Shallow Untarped

[ER = 0.40]	

Deep Untarped

[ER = 0.40]	

Bedded Tarped

[ER = 0.80]	

Hot Gas

[ER = 1.0]



Ventura CA	

X	

X	

X	

X	

X



Bakersfield  CA	

X	

X	

X	

X	

X



Flint MI	

X	

X	

X	

X	

X



Tallahassee FL	

X	

X	

X	

X	

X



Bradenton FL	

X	

X	

X	

X	

X



X = analysis completed

PERFUM flux profiles and categories (i.e., Permit Conditions) based on
the following information from CDPR:

To link above flux profiles to CDPR Permit Conditions, please refer to
Table 1 in the “Buffer Zone Determination” document completed under
Reg. 03-004 (effective 11/3/04)
http://www.cdpr.ca.gov/docs/legbills/03004buffer_zones.pdf. 
Determination of the appropriate Permit Condition category should be
based on the emission ratio (ER noted above) and description of the
fumigation method.

Johnson, Bruce (1999) Memorandum to Randy Segawa on Buffer Zone Duration
(November 29, 1999).

Johnson, Bruce and Segawa, Randy (2000) Memorandum to John Sanders on
Re-analysis of decline rate for Methyl Bromide flux rates and buffer
zone durations (May 31, 2000).

Johnson, Bruce (2004) Memo To Kean S. Goh, Additional Help On Air
Modeling Provided To The United States Environmental Protection Agency -
Hot Gas Flux Profile (December 28, 2004)



The Agency utilized flux inputs for modeling methyl bromide emissions
from pre-plant soil applications that were developed by CDPR as
indicated above.  The following describes the process and data upon
which the analyses to develop these profiles were based.  In January
2000, CDPR generated a memorandum outlining their recommendations for
methyl bromide buffer zones for field fumigations (“Recommendations
For Methyl Bromide Buffer Zones For Field Fumigations” Segawa, et al,
2000).  These recommendations were based on a number of air monitoring
studies dating from 1992 through 1999 that were used to define the
emission ratios which are used in the permit conditions described in
Table 6 above.  [Note:  The studies cited in the Segawa 2000 document
were also evaluated by the Agency in the previous methyl bromide risk
assessment (D316326, 6/13/05).  As such, they are available on the EPA
docket at   HYPERLINK "http://www.Regulations.gov"  www.Regulations.gov 
.]  During the development and implementation of the methyl bromide
permit conditions, CDPR also identified a need for establishing the
duration of a buffer zone.  In order to do this, it was decided that
quantifying how the rate of emissions, defined based on the
back-calculation method described above, change over time was the most
appropriate approach.  The following equation that describes the
dissipation of methyl bromide after field applications was developed as
a result of this effort (“Buffer Zone Duration” Johnson, 1999).

	Where:

		Y	=	fraction of applied mass emitted per hour;

		X	=	time in hours after the start of application;

		A 	=	constant that scales the overall function height and area;

		B	=	constant that determines where the peak is; and

		C	=	constant that determines the width of the function.  

		

This equation has other utility in that it also has been used to define
flux values for modeling purposes for each application method associated
with the permit conditions.  To estimate flux rate for an hour, the area
underneath the curve generated by this equation was used.  For each
hour, values of Y were calculated for each minute which were multiplied
by the application rate to estimate instantaneous flux rates on a per
minute basis.  The area underneath the curve for a one minute time step
was estimated using the area of a trapezoid, A=1/2h(a+b), where h=x2-x1,
and a and b are the incremental values of the instantaneous flux rate
(see Figure 7).  The areas for an hour were summed to provide an
estimate of the flux rate over that hour. 

CDPR developed profiles for each emission ratio that is described in
Table 4 above. The data which CDPR used as the basis for the flux
profile analysis is described for each application method in Table 5. 

Table 5:  Summary Of Field Study Parameters Used To Develop Methyl
Bromide Flux Profiles For PERFUM Modeling

Application Method	Study Code	Study Conductors	Study Date	County
Application Rate (lbs/total acre)	Total Acreage (acres)

Shallow, untarped, bedded 1	SE1.1	Siemer and Associates	8/19/1992
Monterey	186	19

	SE1.2	Siemer and Associates	9/24/1992	Monterey	180	15

	SE1.3	Siemer and Associates	10/27/1992	Monterey	180	15

Tarped, broadcast 1	TC199	Trical, Inc	6/30/1992	Kern	396	20

	EH9503/EH127-1	Environmental Hazards Program, DPR	10/26/1992	Monterey
235	10

Bedded, tarped 1	BR787.1C	Bolsa Research	6/24/1999	Orange	176	1

	BR787.2B	Bolsa Research	6/30/1999	San Louis Obispo	245	1

	BR787.1B	Bolsa Research	6/24/1999	Orange	177	1

	BR787.2C	Bolsa Research	6/30/1999	San Louis Obispo	174	1

	EH9503	Kim and Segawa, 1998	10/6/1998	San Louis Obispo	206	9

Deep, untarped, broadcast 2	SE2.2	Siemer and Associates	10/21/1992	Kern
396	15

	S104.2-1	Siemer and Associates	3/8/1993	Fresno	396	40

	S100B1.1	Siemer and Associates	3/13/1993	Madera	400	20

	S110.1	Siemer and Associates	10/31/1995	San Joaquin	450	7

Hot-drip gas 3	EH150-1	Environmental Hazards Program, DPR	12/11/1996
Riverside	200	25

	EH150-3	Environmental Hazards Program, DPR	1/20/1997	Kern	200	14

	EH150-4	Environmental Hazards Program, DPR	1/27/1997	Imperial	200	14

1  Flux profile and derivation described in CDPR memo entitled:  Buffer
Zone Duration Johnson, 1999

2 Flux profile and derivation described in CDPR memo entitled: 
Re-analysis of Decline Rates For Methyl Bromide Flux Rates And Buffer
Zone Determinations Johnson and Segawa, 2000

3  Flux profile and derivation described in CDPR memo entitled: 
Additional Help On Air Modeling Provided To The United States
Environmental Protection Agency – Hot Gas Flux Profile Johnson, 2004





Table 6 depicts the results of the CDPR analysis for each parameter
described in the equation above.  Subsequently, Table 7 presents the
flux rates for each application method used corrected to the maximum
application effective broadcast rate of 430 lb ai/acre which have been
calculated for methyl bromide for the first 24 hours after application
(i.e., the period of key concern).  As indicated above, the effective
broadcast rate for tarped raised beds differs from application methods
that involve 100 percent of the area of a field.  In raised bed
applications only a percentage of the total land mass is actually
treated (i.e., 60 percent for the purposes of this assessment based on
BR787.2B from Table 5) that results in an effective maximum broadcast
rate of 250 lb ai/acre for that application method.  The flux rates
included in Table 8 are those which have been used as the basis for the
PERFUM distributional modeling.  If an output file for PERFUM was
examined, these values would be included as model inputs in that file. 
However, it should be noted that the hour depicted in Table 7 reflects
the time from the start 

of application and not the hour of the day.  For PERFUM runs involving
methyl bromide, applications were assumed to begin at 9:00 am. 
Therefore, the Hour 1 flux rates shown in Table 7 appear at Hour 10
(9:00 am – 10:00 am) in the PERFUM files, with subsequent flux rates
appearing in the corresponding hours.

Table 6: Coefficients Used for Field Use Flux Rate Estimation



Application Method	Application Rate

(lb ai/acre)	A	B	C

Shallow, untarped	200	0.021113	15.5127	0.84929

Tarped	400	0.013723	15.5127	0.84929

Bedded, tarped	250	0.132867	0.602401	1.53484

Deep, untarped	400	0.016618	12.36407	1.081679

Drip, hot gas	225	1.11E-01	6.414006	0.495166



Table 7: Flux Rates (ug/m2-s), by Application Method, for Methyl Bromide
Field Use



Hour	Shallow, untarped	Tarped	Bedded, Tarped	Deep, untarped	Drip, hot
gas

1	0.31	0.2	889.64	4.42	0.15

2	7.12	4.63	868.23	33.61	29.26

3	28.53	18.54	675.28	74.48	254.32

4	60.97	39.63	537.65	112.35	703.00

5	97.73	63.52	439.43	143.49	1143.44

6	134.01	87.1	366.93	167.83	1409.25

7	167.1	108.61	311.71	186.25	1481.21

8	195.8	127.26	268.53	199.78	1412.05

9	219.75	142.83	234.03	209.39	1264.36

10	239.08	155.4	205.97	215.87	1086.08

11	254.18	165.21	182.78	219.87	907.12

12	265.5	172.57	163.38	221.92	743.25

13	273.55	177.8	146.95	222.42	601.04

14	278.8	181.21	132.92	221.71	481.74

15	281.68	183.09	120.81	220.05	383.86

16	282.59	183.68	110.30	217.65	304.74

17	281.86	183.2	101.10	214.68	241.43

18	279.78	181.85	93.01	211.28	191.10

19	276.61	179.79	85.84	207.55	151.27

20	272.56	177.16	79.47	203.59	119.82

21	267.81	174.07	73.77	199.45	95.02

22	262.52	170.63	68.66	195.2	75.47

23	256.81	166.92	64.05	190.88	60.05

24	250.79	163.01	59.89	186.53	47.88



PERFUM works by establishing a grid with receptor points around a field
built with spokes and rings then it calculates air concentrations at
each point for each day over 5 years of weather data.  The numbers of
receptors for varying sized fields is summarized in Table 8 and Figure 8
below.   The information calculated at each grid location is then used
to calculate distances in each array (or spoke protruding outwards from
the treated field in the center of Figure 8) where a target
concentration of concern is achieved.  Target concentrations are defined
by dividing the HEC by the uncertainty factor of interest for that
particular analysis.  PERFUM compiles these results for each array (or
spoke) then ultimately compiles them across all spokes and weather days
using two techniques (i.e., referred to as a “whole field” or
“maximum” buffer results which are described below).  Each receptor
corresponds to an x- and y-coordinate.  Figures 9, 10 and 11 below,
provide an example of daily PERFUM output where a contour plot has been
developed that describes the distances where a target concentration has
been achieved around the perimeter of a treated field for three distinct
weather days. Each plot pertains to one application using the same
emission rate and field size but the difference between the plots is
that each presents the results (i.e., distance where a target
concentration of concern is achieved) for a single, separate day of
varied weather conditions.

Table 8:  Receptor Points for Various Field Sizes in PERFUM

Grid Type	Field Size (Acres)	Number of Spokes	Number of Rings	Numbers of
Receptors

(Spokes * Distances)

Fine	1	96	28	2,688

	5	132	28	3,696

	10	152	28	4,256

	20	188	28	5,264

	40	232	28	6,496

Coarse	1	24	28	672

	5	33	28	924

	10	38	28	1,064

	20	47	28	1,316

	40	58	28	1,624

Note: Fine grid option was used for methyl bromide analysis.

The maximum distance used for PERFUM calculations on each spoke is
≥1440 meters.



Figure 8: PERFUM Receptor Grid

Whole field buffer results are calculated using PERFUM by compiling the
results for all arrays (i.e., using the entire perimeter) of each days
contour line outputs.  PERFUM compiles all of the locations (i.e., x and
y coordinates) along the contours in each of the plots into one
distribution and essentially produces an overall contour plot for the 5
years of weather data (see Figure 12 below).  The user can then select a
percentile of the distribution of interest (e.g., 95th percentile or
99th percentile). In essence, the “whole field” buffer results
represent the entire range of possible exposures regardless of location
relative to the treated field. 

Maximum buffer results from PERFUM are calculated by compiling only the
farthest distances from the contours produced for each weather day.  The
black dot in each plot (Figures 9 through 11) represents the maximum
distance buffer for that day which would be the only point selected for
that day used in this calculation.  PERFUM also generates these maximum
distance buffers across 5 years of weather data which is presented in
Figure 12 below.  The user can then select a percentile of the
distribution of interest (e.g., 95th percentile or 99th percentile).  In
summary, the maximum buffer results can be thought of as a way of
providing more resolution around the upper percentiles of possible
exposure.  In a physical sense, it can also possibly be applicable to
individuals who live in an area with strong prevailing winds due to
topography or other factors (e.g., in a valley or coastal situation
where on-shore winds are predominant).

Note that in Figure 12 the whole field buffer contour is within the
boundary of the maximum buffer contour.  This trend would always be
expected if the percentiles considered in each case were of the same
numerical value (e.g., 95th %tile whole and 95th%tile maximum) because
the maximum buffer distribution represents only the farthest distance
for each weather day and not all of the values as in the whole field
buffer distribution.  

Figure 9:  Example PERFUM Output - Day 1

Figure 10:  Example PERFUM Output - Day  2

Figure 11:  Example PERFUM Output - Day  3

Figure 12: Whole Field vs. Maximum Buffer Distance Example

PERFUM can generate the types of outputs discussed above for different
exposure periods from 1 to 24 hours depending on the exposure duration
and toxicity concern for the fumigant.  When the distributional results
from PERFUM are considered, they can be described by the following
statements: 

Maximum Distance Buffer:  The maximum concentration (e.g., at 95th
percentile) provides a buffer zone whereby there will not be an
exceedence for 95 percent of application days.  It follows that if a
person was at the location of the maximum concentration on any given day
they would have a 95 percent chance of being at a location with a
concentration less than the target.

Whole Field Buffer:  The whole field distribution (e.g., at 95th
percentile) provides, on average, 95 percent of the perimeter of the
buffer zone will have a concentration below the target. It follows if a
person was placed randomly onto the perimeter of a buffer zone on a
random day they would have a 95 percent chance that the concentration at
their location is less than the target.

5.1.1.2	Bystander Exposures And Risks From Known Sources tc \l4 "6.1.1.2
Bystander Exposures And Risks From Known Sources 

The risks for bystanders from various types of known sources (e.g.,
farmfields and structures) are presented in this section.  As noted
above, known sources represent either an area source from a single
application such as a treated farm field or a point source from a single
application such as a stack used to dissipate residues from a treated
structure.  

Distributional Air Modeling:   This approach is based on the PERFUM
model and is believed to provide the most refined, scientifically
defensible approach for calculating and characterizing risks because it
incorporates actual weather data and it links flux profiles to the
appropriate time of day when calculating results.  It is also based on
the proven technology of ISCST3.  PERFUM has been used to address
results from the pre-plant soil uses as well as the structural uses for
methyl bromide.



Deterministic Air Modeling:   Deterministic air modeling based on ISCST3
was completed for all methyl bromide uses.  However, since the
distributional air modeling described below is based on PERFUM that
contains ISCST3 as its core processor ISCST3 results have not been
presented herein.  It is also believed that PERFUM results represent a
refinement of the ISCST3 approach (see Phase 3 = D316326, 7/13/05 &
Phase 1 = D311945, 1/31/05 if so desired to review the ISCST3 analyses).

Monitoring Studies And Incident Data:  A series of monitoring studies
were reviewed, most of which were conducted by the CDPR.  The
information extracted from these studies data were summarized in the
previous assessments and should be referenced as appropriate (Phase 3 =
D316326, 7/13/05 & Phase 1 = D311945, 1/31/05).  The summarized results
are presented below for comparison to the modeling results.  To the
extent possible, information pertaining to methyl bromide incidents will
be obtained and summarized in order to characterize the results of the
quantitative risk assessment.

An overview of how this section is organized is presented below. 
Results have been summarized based on the industry sector.  Monitoring
data provide empirical results for calculating risks but are limited as
described above.  Conversely, PERFUM provides the most refined risk
estimates for the pre-plant agricultural uses of methyl bromide as
described above.  In addition to the results summarized herein, more
details pertaining to the monitoring data are included in the previous
risk assessments (Phase 3 = D316326, 7/13/05 & Phase 1 = D311945,
1/31/05).  

Section 5.1.1.2.a: Bystander Exposures And Risks From Pre-Plant
Agricultural Use  The recommended results summarized in this section are
based on monitoring data and the use of the PERFUM model.  The data
contained in Appendix D: PERFUM Output Summary Spreadsheets For
Pre-Plant Soil Uses are spreadsheets which summarize the PERFUM analyses
that have been completed for this scenario.  [Note:  These spreadsheets
do not contain all information related to these analyses such as input
parameters and some outputs such as monthly buffer zone distributions. 
This information is contained in the corresponding PERFUM input and
output files are available upon request.]

Section 5.1.1.2.b: Bystander Exposures And Risks From Greenhouse Use 
The recommended results summarized in this section are based on
monitoring data and the use of the PERFUM model.  

Section 5.1.1.2.c: Bystander Exposures And Risks From Residential Use 
The recommended results summarized in this section are based on
monitoring data and the use of the PERFUM model.  



5.1.1.2.a:  Bystander Exposures And Risks From Pre-Plant Agricultural
Use tc \l5 "6.1.1.2.a:  Bystander Exposures And Risks From Pre-Plant
Agricultural Use 

Exposures to bystanders from single pre-plant agricultural field
fumigation events and their associated risks, calculated using the
PERFUM modeling approach, are presented in this section.  Results based
on monitoring data are also presented for comparative purposes.  The
PERFUM modeling results included in this section represent elements of
the information presented in Appendix D.  Detailed monitoring study
information is provided in previous assessments as described above. 
Appendix D contains 26 files that provide results for the various
combinations of application methods (flux estimates) and meteorological
stations considered (see Table 5 for more details above).  These
include:

Appendix D1.  PERFUM Analysis: contains a summary of the PERFUM results
for all flux and meteorological combinations.  The analyses presented
below that compare results for different weather stations and flux types
are included in this file.

Appendix D1a-D1e.Ventura CA: 5 files that contain PERFUM results for all
field sizes, application rates, and flux profiles for Ventura California
meteorological data, also contains results for varied uncertainty factor
levels (i.e., 1, 3, 10, 30).

Appendix D2a-D2e. Tallahassee FL: 5 files that contain PERFUM results
for all field sizes, application rates, and flux profiles for
Tallahassee Florida meteorological data, also contains results for
varied uncertainty factor levels (i.e., 1, 3, 10, 30).

Appendix D3a-D3e. Flint MI: 5 files that contain PERFUM results for all
field sizes, application rates, and flux profiles for Flint Michigan
meteorological data, also contains results for varied uncertainty factor
levels (i.e., 1, 3, 10, 30).

Appendix D4a-D4e. Bradenton FL: 5 files that contain PERFUM results for
all field sizes, application rates, and flux profiles for Bradenton
Florida meteorological data, also contains results for varied
uncertainty factor levels (i.e., 1, 3, 10, 30).

Appendix D5a-D5e. Bakersfield CA: 5 files that contain PERFUM results
for all field sizes, application rates, and flux profiles for
Bakersfield California meteorological data, also contains results for
varied uncertainty factor levels (i.e., 1, 3, 10, 30).



The analyses which were completed using PERFUM are based on the 25
combinations of flux and meteorological data which are available (refer
to Table 5 above).  In addition, the impact of field size and shape,
application rates, “whole vs. maximum buffer” statistics, and target
concentrations (i.e., HECs coupled with uncertainty factor) were
evaluated.  The field sizes and shapes (N=9) that were considered
include:

1 acre (square, rectangle oriented on its side, rectangle oriented on
its end);

5 acres (square, rectangle oriented on its side, rectangle oriented on
its end);

10 acres (square);

20 acres (square); and

40 acres (square).

The maximum broadcast application rate that was considered for pre-plant
soil applications in this assessment is 430 lb ai/acre while the maximum
effective broadcast rate that has been utilized to assess raised bed
cultural production is 250 lb ai/acre.  The effective broadcast rate for
raised beds is based on 60 percent of the gross acreage being covered by
raised beds that are actually treated with methyl bromide.  In addition
to these maximum application rates, a range of other application rates
were evaluated in order to assess the impact of lowering rates including
75, 50, and 25 percent of the maximums for each use pattern.  

The impact of altering target concentrations (i.e., the combination of
HEC coupled with uncertainty factor) was also considered to allow for a
broader characterization of the risks associated with methyl bromide. 
The target concentrations that were considered in for each flux profile
and meteorological input combination (N=25) include:

[Target] = 0.33 ppm based on NOAEL HEC (10 ppm) and Uncertainty Factor =
30;

[Target] = 1.0 ppm based on NOAEL HEC (10 ppm) and Uncertainty Factor =
10; 

[Target] = 3.3 ppm based on NOAEL HEC (10 ppm) and Uncertainty Factor =
3; and

[Target] = 10.0 ppm based on NOAEL HEC (10 ppm) and Uncertainty Factor =
1 

All totaled, when varied application methods (N=5), meteorological data
(N=5), field sizes/shapes (N=5), application rates (N=4) and target
concentrations (N=4) are considered, 3600 PERFUM analyses were completed
in order to evaluate the potential risks associated with pre-plant uses
in agricultural fields.

It should be acknowledged that a myriad of micro-environmental
conditions and factors can impact how methyl bromide will volatilize and
disperse from any given treated field on a particular day.  With this
premise, it would be logical to evaluate basic factors which could
influence flux (e.g., soil type, soil temperature, percent water, etc.)
and also micro-climates (e.g., topography) and thus ultimately impact
results.  However, PERFUM cannot easily address specific changes in
these factors because it is not a First Principles Model where the
approach would be to build a predictive tool from basic fate
characteristics.  Instead, PERFUM is an empirical model which utilizes
field study and actual meteorological data to predict results and since
field study data are the basis for the PERFUM predictions it follows
that results based on empirical monitoring and those calculated with
PERFUM would be similar (see
http://www.epa.gov/scram001/guidance/guide/appw_03.pdf for additional
guidance pertaining to air model validation).

It should also be acknowledged that the nomenclature incorporated into
PERFUM uses the term “buffer zone” which equates to the distance
downwind at which a specific target concentration (i.e., combination of
HEC and UF) is met based on the desired statistical parameters.  The use
of this term does not imply any regulatory decision with regard to the
implementation of buffer zones associated with updating methyl bromide
labels.  Any required label modifications to methyl bromide will be
developed in the Agency’s extensive public participation regulatory
process.

It is clear that given the number of possible permutations of PERFUM
inputs and ways of presenting the outputs that there are many possible
approaches for interpreting the results.  The central goal, however, was
to quantify how potential risks change with factors such as application
method, distance from the treated field, percentile of exposure,
selected statistical basis (i.e., whole vs. maximum buffer approach),
application rate, and field size/shape.  Each of these factors has been
considered and very detailed results pertaining to each are available in
the appendices referenced above.  In order to summarize the analyses
which have been completed and to illustrate the general approach, a
selected number of tabular and graphical interpretations of the results
are presented below.  Most of the information presented below is based
on the Ventura California meteorological data inputs as an example.  A
similar analysis and summary could be completed based on the results for
any source of weather data.

Tables 9 and 10 present PERFUM results (i.e., predicted buffer
distances) based on Ventura California weather data and the tarped flux
emission profile for methyl bromide for 10 and 40 acre fields,
respectively.  In these tables results are presented for different
percentiles of exposure, different application rates, the nature of the
PERFUM output (i.e., maximum distance or whole field buffers), and
different uncertainty factors.  It should be noted that PERFUM analyses
were completed for an uncertainty factor = 1 but they are not included
in these tables because essentially all predicted buffer results were 0
meters.  For a 10 acre field, the maximum and whole field buffer
distances were 225 and 115 meters, respectively, at the 99th percentile
of exposure at the maximum application rate and an uncertainty factor of
30.  If any factors are reduced then predicted buffer distances change,
but in a non-linear Gaussian fashion.  For example, if all other factors
are held constant and the application rate was reduced to a more typical
215 lb ai/acre (1/2X) then distances would be 60 and 5 meters,
respectively, for maximum and whole field buffers.  Similar trends can
be observed in the results for a 40 acre field.  In a 40 acre field, the


corresponding buffer distances at the 99th percentile of exposure at
the maximum application rate and an uncertainty factor of 30 would be
555 and 330 meters, respectively.  Appendix D contains information for
all of the PERFUM analyses that were completed.  There is a summary
section of each file so similar information could be obtained for any of
these analyses if so desired.  

Table 9:  Methyl Bromide PERFUM Buffer Distributions For A 10 Acre
Square Field Based On Ventura California Weather And Tarped Flux Profile
Data

Percentiles	Max (430 lb/Acre)	75% (323 lb/Acre)	50% (215 lb/Acre)	25%
(108 lb/Acre)

	Max

Distance	Whole

Field	Max

Distance	Whole

Field	Max

Distance	Whole

Field	Max

Distance	Whole

Field

UF=30

5	0	0	0	0	0	0	0	0

10	0	0	0	0	0	0	0	0

15	0	0	0	0	0	0	0	0

20	0	0	0	0	0	0	0	0

25	0	0	0	0	0	0	0	0

30	5	0	0	0	0	0	0	0

35	5	0	0	0	0	0	0	0

40	5	0	0	0	0	0	0	0

45	5	0	5	0	0	0	0	0

50	20	0	5	0	0	0	0	0

55	30	0	5	0	0	0	0	0

60	40	0	5	0	0	0	0	0

65	50	0	5	0	0	0	0	0

70	65	0	5	0	0	0	0	0

75	80	0	25	0	5	0	0	0

80	90	0	40	0	5	0	0	0

85	110	5	55	0	5	0	0	0

90	135	5	75	5	5	0	0	0

95	165	35	95	5	5	5	0	0

97	185	65	115	20	25	5	0	0

99	225	115	150	65	60	5	5	0

99.9	290	195	195	130	85	50	5	5

99.99	315	275	220	180	105	85	5	5

UF=10

5	0	0	0	0	0	0	0	0

10	0	0	0	0	0	0	0	0

15	0	0	0	0	0	0	0	0

20	0	0	0	0	0	0	0	0

25	0	0	0	0	0	0	0	0

30	0	0	0	0	0	0	0	0

35	0	0	0	0	0	0	0	0

40	0	0	0	0	0	0	0	0

45	0	0	0	0	0	0	0	0

50	0	0	0	0	0	0	0	0

55	0	0	0	0	0	0	0	0

60	0	0	0	0	0	0	0	0

65	0	0	0	0	0	0	0	0

70	0	0	0	0	0	0	0	0

75	0	0	0	0	0	0	0	0

80	0	0	0	0	0	0	0	0

85	0	0	0	0	0	0	0	0

90	0	0	0	0	0	0	0	0

95	5	0	0	0	0	0	0	0

97	5	0	0	0	0	0	0	0

99	5	5	5	0	0	0	0	0

99.9	5	5	5	5	0	0	0	0

99.99	5	5	5	5	0	0	0	0

UF=3

5	0	0	0	0	0	0	0	0

10	0	0	0	0	0	0	0	0

15	0	0	0	0	0	0	0	0

20	0	0	0	0	0	0	0	0

25	0	0	0	0	0	0	0	0

30	0	0	0	0	0	0	0	0

35	0	0	0	0	0	0	0	0

40	0	0	0	0	0	0	0	0

45	0	0	0	0	0	0	0	0

50	0	0	0	0	0	0	0	0

55	0	0	0	0	0	0	0	0

60	0	0	0	0	0	0	0	0

65	0	0	0	0	0	0	0	0

70	0	0	0	0	0	0	0	0

75	0	0	0	0	0	0	0	0

80	0	0	0	0	0	0	0	0

85	0	0	0	0	0	0	0	0

90	0	0	0	0	0	0	0	0

95	0	0	0	0	0	0	0	0

97	0	0	0	0	0	0	0	0

99	0	0	0	0	0	0	0	0

99.9	0	0	0	0	0	0	0	0

99.99	0	0	0	0	0	0	0	0



Table 10:  Methyl Bromide PERFUM Buffer Distributions For A 40 Acre
Square Field Based On Ventura California Weather And Tarped Flux Profile
Data

Percentiles	Max (430 lb/Acre)	75% (323 lb/Acre)	50% (215 lb/Acre)	25%
(108 lb/Acre)

	Max

Distance	Whole

Field	Max

Distance	Whole

Field	Max

Distance	Whole

Field	Max

Distance	Whole

Field

UF=30

5	0	0	0	0	0	0	0	0

10	0	0	0	0	0	0	0	0

15	5	0	0	0	0	0	0	0

20	10	0	0	0	0	0	0	0

25	30	0	5	0	0	0	0	0

30	40	0	5	0	0	0	0	0

35	60	0	5	0	0	0	0	0

40	70	0	15	0	0	0	0	0

45	85	0	30	0	5	0	0	0

50	110	0	40	0	5	0	0	0

55	125	0	55	0	5	0	0	0

60	145	0	70	0	5	0	0	0

65	170	0	85	0	5	0	0	0

70	200	5	110	0	5	0	0	0

75	230	5	130	5	25	0	0	0

80	260	10	150	5	40	0	0	0

85	305	30	180	5	65	0	0	0

90	350	70	220	25	90	5	5	0

95	430	150	280	80	125	5	5	0

97	470	210	320	125	150	35	5	0

99	555	330	390	215	200	90	5	5

99.9	725	510	500	355	265	190	20	5

99.99	760	680	555	485	305	265	60	45

UF=10

5	0	0	0	0	0	0	0	0

10	0	0	0	0	0	0	0	0

15	0	0	0	0	0	0	0	0

20	0	0	0	0	0	0	0	0

25	0	0	0	0	0	0	0	0

30	0	0	0	0	0	0	0	0

35	0	0	0	0	0	0	0	0

40	0	0	0	0	0	0	0	0

45	0	0	0	0	0	0	0	0

50	0	0	0	0	0	0	0	0

55	0	0	0	0	0	0	0	0

60	0	0	0	0	0	0	0	0

65	0	0	0	0	0	0	0	0

70	0	0	0	0	0	0	0	0

75	5	0	0	0	0	0	0	0

80	5	0	0	0	0	0	0	0

85	5	0	0	0	0	0	0	0

90	5	0	5	0	0	0	0	0

95	5	5	5	0	0	0	0	0

97	30	5	5	0	0	0	0	0

99	70	5	5	5	5	0	0	0

99.9	105	65	20	5	5	5	0	0

99.99	140	125	60	45	5	5	0	0

UF=3

5	0	0	0	0	0	0	0	0

10	0	0	0	0	0	0	0	0

15	0	0	0	0	0	0	0	0

20	0	0	0	0	0	0	0	0

25	0	0	0	0	0	0	0	0

30	0	0	0	0	0	0	0	0

35	0	0	0	0	0	0	0	0

40	0	0	0	0	0	0	0	0

45	0	0	0	0	0	0	0	0

50	0	0	0	0	0	0	0	0

55	0	0	0	0	0	0	0	0

60	0	0	0	0	0	0	0	0

65	0	0	0	0	0	0	0	0

70	0	0	0	0	0	0	0	0

75	0	0	0	0	0	0	0	0

80	0	0	0	0	0	0	0	0

85	0	0	0	0	0	0	0	0

90	0	0	0	0	0	0	0	0

95	0	0	0	0	0	0	0	0

97	0	0	0	0	0	0	0	0

99	0	0	0	0	0	0	0	0

99.9	0	0	0	0	0	0	0	0

99.99	0	0	0	0	0	0	0	0



The information that is included in Tables 9 and 10 can also be
graphically presented (as can the results of any of the completed PERFUM
analysis).  Figures 13 and 14 present the maximum buffer distances based
on uncertainty factors of 30 and 10, respectively, that were calculated
using the Ventura California weather data and the tarped flux profile. 
In these graphs, buffer distance results are plotted versus the
percentile of exposure for 1, 10 and 40 acre fields at varying
application rates including the maximum, 75 percent and 50 percent rate
(i.e., 1 acre tables are not presented above for simplicity but the
information is included in the graphs for comparison).  Figures 15 and
16 are similar in nature except they present the whole field buffer
results instead of the maximum buffer results.  When reviewing the
results in Figures 13 through 16 note that the scale of the “y” axis
are similar for direct comparison.

Figure 13

Figure 14

Figure 15

Figure 16

For comparative purposes, similar graphs are presented below in Figures
16, 17, and 18 that are based on results for Ventura California weather
data but flux profiles for different application methods.  In Figure 17,
the flux profile is for raised bed applications (i.e., bedded tarped)
while Figure 18 presents hot gas application method results (i.e., also
known as drip application).  The hot gas method is the highest emitting
application method (i.e., it has the highest flux rates). As a result,
it is the one where buffer zones greater than 25 to 50 meters or so were
identified at a combined uncertainty factor of 3 (as opposed to 10 for
the others above) so these results are included as Figure 19.

Figure 17

Figure 18

Figure 19

Table 11 provides a comparison of results for selected percentiles of
exposure among flux profiles for a 40 acre field at varied application
rates and an uncertainty factor of 30.  The results essentially track
with the relative differences in emission ratios that are described
above in Table 5 with the tarped application method having the lowest
associated emission and lowest predicted PERFUM buffer results.
Conversely, the relative ranking of the hot gas and bedded tarped
application methods as having the highest and penultimate emission
ratios is also consistent with the PERFUM results provided below.  The
relative trends illustrated in Table 11 would be expected to be similar
regardless of the choice of meteorological data selected for the
analysis.  Figure 20 illustrates these trends for varied application
rates based on a combined uncertainty factor of 30.  Figure 21 provides
similar information but also provides results for a combined uncertainty
factor of 10 for comparative purposes.  As expected, the absolute values
are decreased with a lower uncertainty value so the effect appears to be
attenuated among flux types.

Table 11:   Comparison Of Results For Methyl Bromide PERFUM Buffer
Distributions Based On A 40 Acre Square Field, Ventura California
Weather Data, And All Flux Profiles At A UF=30

Percentiles	Max (430 lb/Acre)	75% (323 lb/Acre)	50% (215 lb/Acre)	25%
(108 lb/Acre)

	Max

Distance	Whole

Field	Max

Distance	Whole

Field	Max

Distance	Whole

Field	Max

Distance	Whole

Field

Flux Profile = Tarped Application Method

50	110	0	40	0	5	0	0	0

75	230	5	130	5	25	0	0	0

90	350	70	220	25	90	5	5	0

95	430	150	280	80	125	5	5	0

99	555	330	390	215	200	90	5	5

99.9	725	510	500	355	265	190	20	5

99.99	760	680	555	485	305	265	60	45

Flux Profile = Shallow Untarped Application Method

50	245	0	150	0	45	0	0	0

75	425	30	290	10	135	5	5	0

90	615	150	430	90	230	30	20	0

95	710	285	510	190	290	85	50	5

99	895	565	665	400	410	225	110	30

99.9	1185	835	855	610	520	365	150	105

99.99	1205	1130	880	810	575	500	185	170

Flux Profile = Hot Gas Application Method

50	505	25	360	10	200	5	35	0

75	705	135	510	95	305	50	85	5

90	985	315	725	235	450	135	150	25

95	1185	450	880	340	565	210	210	60

99	1440	790	1250	605	795	395	355	150

99.9	1440	1285	1440	970	1205	655	500	315

99.99	1440	1440	1440	1275	1330	845	545	455

Flux Profile = Deep Untarped Application Method

50	165	0	85	0	5	0	0	0

75	305	20	195	5	70	0	0	0

90	450	105	305	55	140	5	5	0

95	530	200	370	125	185	40	5	0

99	695	420	490	285	290	140	45	5

99.9	875	635	630	445	365	260	110	45

99.99	935	850	700	605	430	350	115	100

Flux Profile = Bedded Tarped Application Method (Note:  Effective
Maximum Broadcast Maximum Application Rate = 250 lb ai/Acre)

50	135	0	70	0	10	0	0	0

75	200	30	115	10	40	5	5	0

90	295	95	190	50	80	10	5	0

95	395	145	260	90	120	30	5	5

99	670	290	525	195	305	95	85	5

99.9	1225	590	925	440	555	285	210	85

99.99	1305	1005	955	730	565	480	230	205



Figure 20

Figure 21

In addition to the comparisons described above among flux types, a
comparison was also completed that evaluated differences concurrently
among meteorological data and flux profile (Table 12 & Figure 22).  The
relative difference based on flux profile is similar regardless of the
weather data used for the analysis.  For results based on the selection
of meteorological data, it appears that results for Bradenton Florida
have higher associated buffer distances than (in order) Ventura
California, Tallahassee Florida, Flint Michigan, and Bakersfield
California.  These results are consistent with the sensitivity analysis
completed by the model developer and presented at the 2004 FIFRA
Scientific Advisory Panel meeting
(http://www.epa.gov/oscpmont/sap/meetings/2004/index.htm).  The results
presented in Table 13 are based on a 40 acre field and an uncertainty
factor of 30 at the maximum application rate for each method.  It is
anticipated that the general trends observed in Table 12 and Figure 22
would also still apply regardless of the field size, uncertainty factor
basis, or application rate. 



Table 12:   Comparison Of Results For Methyl Bromide PERFUM Buffer
Distributions Based On A 40 Acre Square Field, All Weather Data, And All
Flux Profiles At A UF=30 And Maximum Application Rate

%tiles	Ventura California	Tallahassee Florida	Flint Michigan	Bakersfield
California	Bradenton Florida

	Max

Distance	Whole

Field	Max

Distance	Whole

Field	Max

Distance	Whole

Field	Max

Distance	Whole

Field	Max

Distance	Whole

Field

Flux Profile = Tarped Application Method

50	110	0	60	0	50	0	95	0	155	0

75	230	5	130	5	115	5	145	15	290	15

90	350	70	225	50	190	40	200	65	420	105

95	430	150	290	95	240	85	235	105	510	185

99	555	330	435	215	385	185	315	190	695	350

99.9	725	510	620	415	515	360	390	295	905	575

99.99	760	680	630	585	580	500	395	375	935	790

Flux Profile = Shallow Untarped Application Method

50	245	0	150	0	130	0	195	5	305	0

75	425	30	265	25	240	25	280	65	500	50

90	615	150	410	110	350	105	365	145	695	185

95	710	285	515	185	425	170	410	205	855	305

99	895	565	770	375	650	330	540	330	1130	555

99.9	1185	835	1050	690	905	605	640	495	1405	925

99.99	1205	1130	1065	995	960	860	705	625	1440	1340

Flux Profile = Hot Gas Application Method

50	505	25	485	20	370	5	455	65	670	35

75	705	135	705	135	495	120	605	195	950	175

90	985	315	970	315	695	265	800	335	1320	380

95	1185	450	1170	460	850	365	935	435	1435	530

99	1440	790	1440	820	1365	625	1285	665	1440	940

99.9	1440	1285	1440	1395	1440	1110	1440	1040	1440	1435

99.99	1440	1440	1440	1440	1440	1440	1440	1435	1440	1440

Flux Profile = Deep Untarped Application Method

50	165	0	105	0	90	0	135	5	220	0

75	305	20	195	10	170	10	200	40	375	35

90	450	105	310	80	255	70	265	100	530	145

95	530	200	390	140	315	120	305	145	645	235

99	695	420	605	290	490	245	400	245	865	435

99.9	875	635	810	540	680	455	500	375	1160	725

99.99	935	850	830	745	725	625	530	475	1255	995

Flux Profile = Bedded Tarped Application Method (Note:  Effective
Maximum Broadcast Maximum Application Rate = 250 lb ai/Acre)

50	135	0	135	0	100	0	85	5	170	5

75	200	30	205	35	150	25	125	35	235	50

90	295	95	280	100	205	75	175	70	315	125

95	395	145	350	150	265	115	210	100	385	175

99	670	290	500	270	415	200	320	165	475	285

99.9	1225	590	605	455	1025	380	440	285	665	440

99.99	1305	1005	650	580	1440	1245	485	440	715	600



Figure 22

In addition to the comparative analyses presented above, other factors
were evaluated relative to their possible impacts on PERFUM-based buffer
zone predictions.  These included evaluating the effect of field size
and shape on results as well as discerning if there are significant
seasonal differences in results since many fumigant use patterns are
seasonal in nature.  Figure 23 illustrates differences associated with
increasing field sizes and the results indicate that, as expected,
buffer distances increase relative to field size.  Similar trends are
observed regardless of the application rate or whether or not the
results are based on maximum or whole field buffer results.

Figure 23

Figure 24 illustrates differences based on field shape.  In this
analysis, results for a square 5 acre field and rectangular fields
(i.e., based on a 2:1 aspect ratio) oriented alternatively on
perpendicular sides were calculated.  Results were essentially similar
for all types but the “long” field orientation always provided lower
buffer estimates.  The results of this analysis may also be sensitive to
different weather conditions, site topography, and field aspect ratio
but these factors were not evaluated in more detail because their
relevance is likely more significant to specific use sites than to the
development of generally applicable buffer estimates using PERFUM.

Figure 24

The seasonal impacts of changing weather patterns have been evaluated in
every PERFUM analysis.  Table 13 below provides an example of the
outputs that are available.  In this type of analysis PERFUM compiles
distributions based on only the specific month’s worth of
meteorological data from the 5 years used for the analysis so each of
the distributions is based on 5 months instead of 5 years worth of data.
 [Note:  For comparative purposes, the corresponding 5 year distribution
for Table 13 is included in Table 9 above for whole fields, maximum
application rate with an uncertainty factor of 30 (i.e., 95th %tile = 35
m; 99th %tile = 115 m; etc.).]  It appears in this case that longer
buffer distances are predicted in the cooler winter months which may be
due to an overall trend toward a more stable atmosphere in those months
due to less convective heating and atmospheric turbulence than in the
spring and summer months.

Table 13:  Methyl Bromide PERFUM Monthly Whole Field Buffer
Distributions For A 10 Acre Square Field Based On Ventura California
Weather And Tarped Flux Profile At The Maximum Application Rate And With
An Uncertainty Factor = 30

Percentile Of

Exposure	PERFUM Monthly Buffer Distributions

	JAN	FEB	MAR	APR	MAY	JUN	JUL	AUG	SEP	OCT	NOV	DEC

5	0	0	0	0	0	0	0	0	0	0	0	0

10	0	0	0	0	0	0	0	0	0	0	0	0

15	0	0	0	0	0	0	0	0	0	0	0	0

20	0	0	0	0	0	0	0	0	0	0	0	0

25	0	0	0	0	0	0	0	0	0	0	0	0

30	0	0	0	0	0	0	0	0	0	0	0	0

35	0	0	0	0	0	0	0	0	0	0	0	0

40	0	0	0	0	0	0	0	0	0	0	0	0

45	0	0	0	0	0	0	0	0	0	0	0	0

50	0	0	0	0	0	0	0	0	0	0	0	0

55	0	0	0	0	0	0	0	0	0	0	0	0

60	0	0	0	0	0	0	0	0	0	0	0	0

65	0	0	0	0	0	0	0	0	0	0	0	0

70	0	0	0	0	0	0	0	0	0	0	0	0

75	0	0	0	0	0	0	0	0	0	0	0	0

80	5	5	5	0	0	0	0	0	0	0	5	5

85	5	5	5	5	5	5	5	5	5	5	5	5

90	25	25	20	5	5	5	5	5	5	5	5	10

95	75	65	55	40	25	15	5	15	20	40	50	40

97	110	95	85	60	45	35	30	40	45	70	75	80

99	170	145	130	105	85	70	65	75	95	125	135	130

99.9	265	240	195	175	140	155	120	120	145	175	240	195

99.99	330	290	235	200	170	185	185	160	185	200	260	275



As indicated above, monitoring data in the previous phase 3 risk
assessment for methyl bromide (D316326, 6/13/05) were identified and
summarized.  None of this information or analysis has changed since the
previous assessment.  Table 14 below provides a summary of these data
and presents margins of exposure (i.e., MOEs) calculated based on them
(i.e., MOEs > 30 are not of concern).  Note a trend for maximum
concentration estimates is not observed.  For consideration, this can be
due to many factors such as building characteristics, absorption rates
of treated materials, differences in aeration, and prevailing weather
conditions.  

  SEQ CHAPTER \h \r 1 

Table 14:  Risks From Methyl Bromide Exposure Based On Maximum 24 Hour
TWA Concentrations Based On Pre-Plant Agricultural Field Volatility Data




Distance From Treated Field

(feet) & No. Samples	Max. 24 Hour [TWA] (ppm)	Acute MOE

26 to 34 ft

(N = 206)	1.4	7

35 to 79 ft

(N = 80)	0.59	17

80 to 115 ft

(N = 27)	0.470	21

> 115 ft

(N = 90)	1.5	7



The results indicate that the potential risks which have been calculated
for methyl bromide emissions from treated fields are of concern as
calculated acute MOEs are less than 30 for all distances from the field
where data are available.  It should be noted that within each distance
category described in Table 14 that there are a number of samples.  If
MOEs were calculated for many of these they would be greater than 30 and
thus, would not be of concern.  This would not be unexpected.  If one
were to review PERFUM outputs for various analyses it would also
indicate that for many circumstances predicted buffer distances would be
small (e.g., <10 meters or so) at lower percentiles of exposure.  In
summary, monitoring data support the types of results that have been
calculated using the PERFUM model.  Additionally, the limitations
associated with the use of monitoring data can be illustrated based on
the summarized data where there are different numbers of samples in each
distance category and the distance categories themselves were defined
based on the data and not conversely which would be more appropriate. 
If so desired, the previous risk assessment and its associated
appendices that contain all of the monitoring data information can be
found at   HYPERLINK "http://www.Regulations.gov"  www.Regulations.gov 
under docket OPP-2005-0123 (i.e., key various elements include: 
OPP-2005-0123-007 – bibliography; OPP-2005-0123-008 – field
monitoring data).  [Note:  ISCST3 results are not summarized in this
document since PERFUM uses ISCST3 as its core processor and PERFUM-based
results provide more realistic estimates of risk based on actual
meteorological data.]

In conclusion, it is clear that many different factors can impact the
air concentrations (and hence, risks) in proximity to agricultural
fields that have been treated with methyl bromide; these include many of
the factors which have been investigated in this analysis.  It is also
important to acknowledge this issue so that stakeholders understand that
the results of this analysis can be interpreted in many ways depending
upon the factors which are considered.  Many conclusions can be drawn,
but the key ones include: (1) at the edge of the treated fields that
NOAEL HECs (i.e., UF=1) generally are not exceeded given proper use of
methyl bromide but conversely the distance predicted for higher
uncertainty factors (i.e., UF=30) are hundreds of meters for many
scenarios; (2) the methods used to evaluate methyl bromide exposure in
this assessment generally agree and they are based on techniques that
have been routinely used for regulatory purposes, they have also
undergone a significant level of review; (3) the sensitivity of results
to changes in key factors such as flux and meteorological conditions is
generally within a factor of 4 or 5 for those which have been evaluated;
(4) PERFUM is an empirically based approach so the generation of
additional flux and meteorological data would allow a broader analysis
that could be applied more specifically to other regions of the country
and application parameters; and (5) the identification of a result, per
se, for any sort of regulatory action (e.g., a percentile of exposure or
maximum/whole field) would depend upon careful consideration of the
variability and uncertainty as well as any particular merits of the
inputs associated with each.

5.1.1.2.b:  Bystander Exposures And Risks From Greenhouse Use tc \l5
"6.1.1.2.b:  Bystander Exposures And Risks From Greenhouse Use 

The "greenhouse" industry sector is extremely varied because of the
breadth of the facilities that are used across the country and because
of the nature of the products that are produced.  As a result, some
clarification is required in order to interpret the results presented
below.  Certainly, in common "greenhouse" operations, many types of
containerized ornamental plants and vegetable starter sets are produced
in either closed structures that will be referred to as "greenhouses" or
in other related nursery operations such as small fields, or in what are
commonly known as "shade" houses (i.e., essentially fields with an
overhead sunblock of some sort typically a semi-transluscent black shade
cloth).  In the latter type of operation, cultural practices related to
methyl bromide use are essentially identical to the pre-plant field uses
described above in Section 5.1.1.2.a except they typically occur on a
smaller scale (e.g., 1 acre applications or less).  As a result, for
risk estimates related to these types of use patterns, please refer to
the results presented above for pre-plant soil uses and smaller sized
fields.  

For the actual greenhouse operations considered in this assessment, a
maximum application rate of  4 lb ai/1000 ft3 was used based on the
Meth-O-Gas label (EPA Reg. No. 5785-41) for rhizomes, seeds, roots,
bulbs, corms and tubers.  A process has been defined, based on the
PERFUM model described above, for characterizing the off-site
dissipation of methyl bromide that has been released either
intentionally during aeration after treatment is complete or that leaked
from a structure during treatment.  Aeration can be accomplished using
several techniques that employ different devices and/or locations
relative to the treated structure.  In summary, aeration is
predominantly accomplished using either portable industrial fans that
are typically rated to move air volumes up to 5000 CFM (i.e., cubic feet
per minute) or much larger static ventilation systems that can move much
larger quantities of air.  

Based on the observations above pertaining to "greenhouse" uses, it
follows that the methods used to quantify potential risk estimates in
the recently completed commodity assessment (i.e., that addressed the
food uses) for methyl bromide are applicable to this industry.  In both
cases, a structure is treated and the methyl bromide is held in
accordance to a "CxT" (i.e., concentration by time) table in order to
achieve the desired level of efficacy; then the material is typically
actively aerated from the treated structure using various methods.  This
type of operation would occur regardless of whether or not food
commodities were in the structure.  The only elements which may differ
would be the amount retained because of the absorptive capacity of the
materials in the structure being treated.  Food use or not, insufficient
information is available to quantify this phenomenon for more than a
handful of situations so it is believed that the analysis and methods
developed and used for the commodity assessment are directly applicable
to the greenhouse use pattern.  Identification of the applicable results
will depend on how individual fumigation events occur.  In the risk
management strategy for addressing food use patterns the Agency decision
was to allow for flexibility among users depending upon the aeration
method used which is also anticipated for this use pattern.  These
methods and applicable results are presented in detail in the following
documents:

Methyl Bromide:  Phase 5 Health Effects Division (HED) Human Health Risk
Assessment For Commodity Uses, PC Code: 053201, DP Barcode D304623,
Authors:  J. Dawson and E. Mendez, Issued:  3/10/06.

Methyl Bromide:  Addendum To Phase 5 Health Effects Division (HED) Human
Health Risk Assessment For Commodity Uses, PC Code 053201, DP Barcode
D304619 Authors:  J. Dawson E. Mendez, T. Goodlow, M. Metzger, Issued: 
7/12/06

A coding error in the calculation algorithms of the PERFUM model was
identified that pertained to the minimum stack aeration scenario.  The
results summarized below are based on the corrected version of PERFUM. 
It should also be noted that there have been comments pertaining to the
method used to evaluate the PPQ aeration method (i.e., placing an
exhaust fan and vent in the middle of a parking lot attached to the
treated structure by a flexible large hose).  In essence, the issue
pertained to accounting for the orientation of the output orifice and
two options were considered (i.e., referred to as the "vertical stack"
or "horizontal stack").  It was determined that the "horizontal stack"
approach is the most appropriate, consequently, results based on this
method are presented below.  Additionally, comments were made pertaining
to the averaging times used as the basis for both the time-weighted
average air concentrations and the HEC for structural types of uses. 
Based on these comments, the values for exposure and the HEC have been
matched for risk calculation purposes (i.e., both are based on 6 hours).
 Active aeration or leakage during treatment are thought to occur over
short periods so exposure estimates are based on 1 hour emissions
followed by 5 hours of zero emissions in order to reflect anticipated
actual occurences and to reflect an appropriate time basis calculating
risks.  At times, greenhouse and nursery operators also tarp quantities
of potting soil and treat it with methyl bromide.  It is believed that
the no stack scenario (akin to no active aeration) is the most
representative for determining risks reflective of this use pattern.  In
addition, there are times when orchard and tree replanting activities
occur in order to replace whole blocks or individual trees that may have
died due to disease or other factors.  These should not be considered a
greenhouse operation.  If whole blocks of trees are being replaced,
cultural practices related to methyl bromide use are essentially
identical to the pre-plant field uses described above in Section
5.1.1.2.a except they typically occur on a smaller scale (e.g., 1 acre
applications or less).  If single (or a few) trees are being replaced, a
handheld injector is typically used.  No emissions data are available
for this pattern, however, it is believed that this technique presents
minimal risks for bystander populations because of the low amount used.

Table 15 presents a summary of some the results of the PERFUM analyses
that were completed to assess events during treatment (from leakage)
associated with the greenhouse uses of methyl bromide.  The commodity
assessment documents D304623 and D304619 detail the PERFUM iterations
that were completed which could apply to greenhouse treatments.  The
scenarios presented in Table 15 have been selected for illustrative
purposes because they represent many typical greenhouse settings and
they also represent the maximum application rate, uncertainty factor of
30, and typical to high end emissions during treatment (i.e., 1 and 10
percent leakage of the nominal application concentration).  Additional
PERFUM output files which can be provided for review of other situations
if so desired.



Table 15: PERFUM Methyl Bromide Buffer Distances (meters) During
Greenhouse Treatment Based On UF30, 4 hour Exposure Duration,

4 lb/1000 cubic feet Application Rate, Varied Structure Size & Varied
Percent Mass Released

Aeration Type	Percentile	1,000 Cubic Feet	10,000 Cubic Feet	50,000 Cubic
Feet	100,000 Cubic Feet



10% Mass Release	1% Mass Release	10% Mass Release	1% Mass Release	10%
Mass Release	1% Mass Release	10% Mass Release	1% Mass Release

Maximum Buffer Distances

No Stack	95	0	0	10	0	110	0	200	0

	99	0	0	20	0	125	0	225	0

	99.9	0	0	25	0	130	0	230	0

Whole Field Buffer Distances

No Stack	95	0	0	0	0	0	0	10	0

	99	0	0	0	0	40	0	75	0

	99.9	0	0	10	0	105	0	190	0



Table 16 presents a summary of some the results of the PERFUM analyses
that were completed to assess aeration events associated with the
greenhouse uses of methyl bromide.  The commodity assessment documents
D304623 and D304619 detail the thousands of PERFUM iterations that were
completed which could apply in some regard to the greenhouse industry,
depending upon the size and configuration of the facility.  The
scenarios presented in Table 16 have been selected for illustrative
purposes because they represent many typical greenhouse settings and
they also represent the maximum application rate, uncertainty factor of
30, and typical to high end emissions after a treatment (i.e., 75 and 95
percent of the nominal application concentration).  Additional PERFUM
output files which can be provided for review of other situations if so
desired.

Table 16: PERFUM Methyl Bromide Buffer Distances (meters) For All
Greenhouse Aeration Processes Considered Based On UF30, 4 hour Exposure
Duration, 4 lb/1000 cubic feet Application Rate, Varied Structure Size &
Varied Percent Mass Released

Aeration Type	Percentile	1,000 Cubic Feet	10,000 Cubic Feet	50,000 Cubic
Feet	100,000 Cubic Feet



95% Mass Release	75% Mass Release	95% Mass Release	75% Mass Release	95%
Mass Release	75% Mass Release	95% Mass Release	75% Mass Release

Maximum Buffer Distances

Minimum Stack

1  xch/min	95	0	0	0	0	0	0	0	0

	99	0	0	0	0	0	0	0	0

	99.9	0	0	0	0	0	0	0	0

No Stack	95	30	20	205	175	545	470	835	725

	99	35	25	225	195	600	520	930	805

	99.9	40	30	235	200	625	535	975	830

Portable Stack

1 xch/min.	95	0	0	0	0	0	0	0	0

	99	0	0	0	0	0	0	0	0

	99.9	0	0	0	0	0	0	0	0

PPQ

1 xch/min.	95	5	0	20	15	0	0	105	0

	99	10	10	30	25	70	60	140	105

	99.9	15	10	70	60	160	140	170	170

Whole Field  Buffer Distances

Minimum Stack

1  xch/min	95	0	0	0	0	0	0	0	0

	99	0	0	0	0	0	0	0	0

	99.9	0	0	0	0	0	0	0	0

No Stack	95	0	0	10	5	25	25	30	25

	99	0	0	70	60	195	170	290	255

	99.9	30	20	195	165	520	450	810	700

Portable Stack

1 xch/min.	95	0	0	0	0	0	0	0	0

	99	0	0	0	0	0	0	0	0

	99.9	0	0	0	0	0	0	0	0

PPQ

1 xch/min.	95	0	0	10	5	15	10	25	0

	99	0	0	20	15	50	45	80	0

	99.9	5	5	30	25	70	60	105	20



As indicated above, monitoring data have been integral in the
development of this risk assessment.  In the previous phase 3 risk
assessment for methyl bromide (D316326, 6/13/05), all of the monitoring
data that were evaluated were identified and summarized.  None of this
information or analysis has changed since the previous assessment. 
Table 17 below provides a summary of these data and presents margins of
exposure (i.e., MOEs) calculated based on them (i.e., MOEs > 30 are not
of concern).  Note that in some cases, air concentrations increase as
the number of days after application increase.  This appears incongruous
but could be due to a number of factors including monitoring was
completed at different facilities which could impact results, the
absorptive properties of the facilities and treated material could
retain methyl bromide until days after treatment, and wind patterns
could have shifted to a particular sample collection location.

  SEQ CHAPTER \h \r 1 

Table 17:  Risks From Methyl Bromide Exposure Based On Maximum 24 Hour
TWA Concentrations Using Greenhouse Volatility Data



Distance Downwind From Treated Greenhouse

	Days After Application	Maximum 24 Hour [TWA]

(ppm)	Acute MOE

50 feet	0	0.287	35

	1	0.292	34

	2	0.295	34

	3	1.870	5

150 feet	0	0.150	67

	1	0.127	79

	2	0.112	89

	3	0.820	12



The results indicate that the potential risks which have been calculated
for methyl bromide emissions from treated greenhouses can be of concern
as calculated acute MOEs in two cases out of  8 are less than 30.  If
one were to review PERFUM outputs for various analyses it would also
indicate that for many circumstances predicted buffer distances would be
small (e.g., <10 meters or so) at lower percentiles of exposure or
depending  upon the aeration method.  In summary, monitoring data
support the types of results that have been calculated using the PERFUM
model.   If so desired, the previous risk assessment and its associated
appendices that contain all of the monitoring data information can be
found at   HYPERLINK "http://www.Regulations.gov"  www.Regulations.gov 
under docket OPP-2005-0123 (i.e., key various elements include: 
OPP-2005-0123-007 – bibliography; OPP-2005-0123-009 – greenhouse
monitoring data).

5.1.1.2.c: Bystander Exposures And Risks From Residential Use tc \l5
"6.1.1.2.e: Bystander Exposures And Risks From Residential Use 

The residential sector represents a small percentage of the overall use
of methyl bromide.  The maximum application rate used for this
assessment is 3 lb ai/1000 ft3 based on the Brom-O-Gas label for
residential uses (EPA Reg. No. 5785-55 or -08).  A process has been
defined based on the PERFUM model described above for characterizing the
off-site dissipation of methyl bromide that has been released either
intentionally during aeration after treatment is complete or that leaked
from a structure during treatment.  Aeration can be accomplished using
several techniques that employ different devices and/or locations
relative to the treated structure.  In summary, aeration is
predominantly accomplished using portable industrial fans that are
typically rated to move air volumes up to 5000 CFM (i.e., cubic feet per
minute).  [Note:  As described with the greenhouse use above, the
analysis for residential treatments is based on the previous commodity
assessment and the related changes.  For more information refer to the
above.]  In this evaluation, home volumes range from 10,000 to 50,000
cubic feet for small to large single family houses while a 100,000 cubic
foot volume is intended to represent a multi-dwelling unit.

Table 18 presents a summary of some the results of the PERFUM analyses
that were completed to assess events during treatment (from leakage)
associated with the residential uses of methyl bromide.  The commodity
assessment documents D304623 and D304619 detail the PERFUM iterations
that were completed which could apply to residential treatments.  The
scenarios presented in Table 18 have been selected for illustrative
purposes because they represent many typical residential settings and
they also represent the maximum application rate, uncertainty factor of
30, and typical to high end emissions during treatment (i.e., 1 and 10
percent leakage of the nominal application concentration).  Additional
PERFUM output files which can be provided for review of other situations
if so desired.

Table 18: PERFUM Methyl Bromide Buffer Distances (meters) During
Residential Treatment Based On UF30, 4 hour Exposure Duration,

3 lb/1000 cubic feet Application Rate, Varied Structure Size & Varied
Percent Mass Released

Aeration Type	Percentile	10,000 Cubic Feet	25,000 Cubic Feet	50,000
Cubic Feet	100,000 Cubic Feet



10% Mass Release	1% Mass Release	10% Mass Release	1% Mass Release	10%
Mass Release	1% Mass Release	10% Mass Release	1% Mass Release

Maximum Buffer Distances

No Stack	95	0	0	30	0	80	0	165	0

	99	0	0	40	0	90	0	180	0

	99.9	0	0	45	0	95	0	190	0

Whole Field Buffer Distances

No Stack	95	0	0	0	0	0	0	5	0

	99	0	0	0	0	25	0	60	0

	99.9	0	0	30	0	70	0	155	0



Table 19 presents a summary of some the results of the PERFUM analyses
that were completed to assess aeration events associated with the
residential uses of methyl bromide.  The commodity assessment documents
D304623 and D304619 detail the thousands of PERFUM iterations that were
completed which could apply in some regard to the residential uses
depending upon the size of the structure being treated and the type of
aeration used.  The scenarios presented in Table 19 have been selected
for illustrative purposes because they represent many typical
residential settings (i.e., small to large, multi-unit structures) and
they also represent the maximum application rate, uncertainty factor of
30, and typical to high end emissions after a treatment (i.e., 75 and 95
percent of the nominal application concentration).  Additional PERFUM
output files which can be provided for review of other situations if so
desired.

Table 19: PERFUM Methyl Bromide Buffer Distances (meters) For All
Residential Aeration Processes Considered Based On UF30, 4 hour Exposure
Duration, 3 lb/1000 cubic feet Application Rate, Varied Structure Size &
Varied Percent Mass Released

Aeration Type	Percentile	10,000 Cubic Feet	25,000 Cubic Feet	50,000
Cubic Feet	100,000 Cubic Feet



95% Mass Release	75% Mass Release	95% Mass Release	75% Mass Release	95%
Mass Release	75% Mass Release	95% Mass Release	75% Mass Release

Maximum Buffer Distances

Minimum Stack

1  xch/min	95	0	0	0	0	0	0	0	0

	99	0	0	0	0	0	0	0	0

	99.9	0	0	0	0	0	0	0	0

No Stack	95	170	145	295	255	455	395	705	605

	99	185	160	330	280	505	435	780	675

	99.9	195	165	340	290	520	450	810	695

Portable Stack

1 xch/min.	95	0	0	0	0	0	0	0	0

	99	0	0	0	0	0	0	0	0

	99.9	0	0	0	0	0	0	0	0

PPQ

1 xch/min.	95	20	15	40	35	60	50	90	75

	99	30	25	50	45	75	65	115	100

	99.9	40	30	65	55	90	80	140	120

Whole Field  Buffer Distances

Minimum Stack

1  xch/min	95	0	0	0	0	0	0	0	0

	99	0	0	0	0	0	0	0	0

	99.9	0	0	0	0	0	0	0	0

No Stack	95	5	5	15	15	25	20	25	25

	99	60	50	110	95	170	145	245	215

	99.9	160	135	285	245	440	375	680	580

Portable Stack

1 xch/min.	95	0	0	0	0	0	0	0	0

	99	0	0	0	0	0	0	0	0

	99.9	0	0	0	0	0	0	0	0

PPQ

1 xch/min.	95	5	0	10	5	10	5	15	10

	99	15	10	30	25	45	35	65	60

	99.9	20	15	40	35	60	50	90	75



As indicated above, monitoring data have been integral in the
development of this risk assessment.  In the previous phase 3 risk
assessment for methyl bromide (D316326, 6/13/05), all of the monitoring
data that were evaluated were identified and summarized.  None of this
information or analysis has changed since the previous assessment. 
Table 20 below provides a summary of these data and presents margins of
exposure (i.e., MOEs) calculated based on them (i.e., MOEs > 30 are not
of concern).



Table 20:  Risks From Methyl Bromide Exposure Based On Maximum 24 Hour
TWA Concentrations Using Residential Fumigation Volatility Data

N	Number Monitors 

Not Detected	Timing and Distance from Structure	Maximum 24 Hour [TWA]
(ppm)	Acute MOE

56	11	During Fumigation

10 ft	1.5	7

56	2	During Aeration

10 ft	0.611	16

35	0	During Aeration

50 to 100 ft 

(Inside Neighboring Houses)

	0.077	130



The results indicate that the potential risks which have been calculated
for methyl bromide emissions from treated residences can be of concern
as calculated acute MOEs in two cases are less than 30.  If one were to
review PERFUM outputs for various analyses it would also indicate that
for many circumstances predicted buffer distances would be small (e.g.,
<10 meters or so) at lower percentiles of exposure or depending  upon
the aeration method.  In summary, monitoring data support the types of
results that have been calculated using the PERFUM model in some
circumstances.  Additionally, the limitations associated with the use of
monitoring data are also explicit based on the nature of the data.  If
so desired, the previous risk assessment and its associated appendices
that contain all of the monitoring data information can be found at  
HYPERLINK "http://www.Regulations.gov"  www.Regulations.gov  under
docket OPP-2005-0123 (i.e., key various elements include: 
OPP-2005-0123-007 – bibliography; OPP-2005-0123-012 – residential
monitoring data).

5.1.2 	Ambient Bystander Exposure From Regional Sources tc \l3 "6.1.2 
Ambient Bystander Exposure From Multiple Regional Sources 

Ambient levels of methyl bromide are generally not attributable to
specific application events, rather contributions may occur from
multiple sources within a region.  For example, it is likely that
individuals could potentially be exposed to methyl bromide if they live
in proximity to or otherwise frequent areas where significant uses occur
such as a neighborhood located around several strawberry fields in
California during the season of use.  

Exposures from ambient air were estimated from monitoring data collected
solely in California to represent conditions at a regional and
state-wide level.  The California Air Resources Board (hereafter
referred to as CARB) generated most of the data considered in this
analysis.  CARB is a widely recognized institution for these types of
programs and it is part of the California Environmental Protection
Agency.  CARB conducts air monitoring studies for various types of
chemicals throughout California.  The studies conducted by CARB can
generally be categorized as one of two types including: (1) targeted
monitoring typically completed upon request to provide information
related to specialized issues such as fumigant exposures in areas of
high use during the season of use and (2) routine monitoring for select
pollutants via established networks in order to better quantify
exposures in the general population (i.e., CARB Toxic Air Contaminant
monitoring program or TAC).  Additional data were considered that were
generated by the Alliance of the Methyl Bromide Industry (AMBI).  Review
of the AMBI data identified quality control issues in some sample
collection procedures and for this reason they are presented only for
comparative purposes.  

For ease and clarity, the Agency has opted by convention to describe the
available ambient bystander data used in this assessment as follows:

(1) “CARB Data”:  includes targeted monitoring data generated by
both CARB and AMBI focused on areas of high methyl bromide use in the
season of use  (AMBI is for comparative purposes only); and

(2) “TAC Data”:  includes data from CARB’s Toxic Air Contaminant
Network for Methyl Bromide that quantifies background levels in
non-agricultural, urban environments.  

The results associated with the CARB data are presented in Section
5.1.2.1 below while the results associated with the TAC data are
presented in Section 5.1.2.2.

5.1.2.1	Exposures From Regionally Targeted Ambient Air Monitoring tc \l4
"6.1.2.1	Exposures From Regionally Targeted Non-Point Source Ambient Air
Monitoring 

In 2000 and 2001, CDPR requested that CARB conduct a series of studies
to quantify ambient levels of methyl bromide
(http://www.cdpr.ca.gov/docs/empm/pubs/tac/requests.htm).  [Note:  An
additional study, which is not included below, was completed by CARB in
Ventura California in 2005.  This study is being repeated since it
appeared not to quantify peak emission times based on the initiation
date.]  The CDPR also requested that the Alliance of the Methyl Bromide
Industry (AMBI) conduct monitoring studies.  Because most of
California’s pesticide applications normally occur in agricultural
areas and are seasonal in nature, CARB conducts the monitoring studies
to collect data during the worst-case situation - in the areas of high
use during the season of peak use - instead of collecting samples
throughout the State. This "worst-case" information can then be used to
determine the ambient exposures of those people living near places where
pesticides are used.

For the targeted ambient air analysis, HED evaluated different durations
of exposure including single day acute exposures, short- and
intermediate-term exposures, and chronic exposures (Table 21).  Since
samples were collected 3 to 4 times per week from each station, and the
contribution of specific 

applications could not be determined, the statistics were calculated by
station and not on a regional basis (e.g., county).  Risks from acute
exposures were calculated using the maximum 24 hour TWA values measured
at each station and comparing them to the acute 24 hour HEC.  

Risks from short- and intermediate-term exposures (i.e., same HEC and
uncertainty factors apply to both durations) were calculated using the
mean of 8 weekly means calculated by DPR for samples taken over the
course of the use season and comparing them to the short- and
intermediate-term HEC.  This approach was taken in order to
statistically weigh equally each week’s contribution to the overall
seasonal mean because of differing numbers of samples in some weeks. 
Concentrations over the course of a season monitored in these studies
did not vary extensively so calculation of average concentrations for
shorter durations (e.g., 4 weeks) or even the use of an overall mean of
all samples would not expected to be significantly different than
estimates used in this assessment.  This supposition is supported
physically because these studies spanned high use seasons in high use
areas and use would not be expected to dramatically change at these
locations during use seasons.  It should be noted that the statistical
summaries of the available data were completed by DPR and that the
Agency reviewed and concurred with this approach.  There are many
possible ways to calculate exposure estimates given the available data
for completing a short- and intermediate-term assessment.  For example,
a TWA over an entire season could be calculated or weekly TWAs could be
calculated and then averaged over a season. The Agency agrees with the
CDPR use of the mean of 8 weekly means because it does not weigh results
for the number of samples collected in a week (i.e., most weeks had 4
samples but some had 3) and it does not require a data filling procedure
for the days missing each week (i.e., usually Wed., Sat., and Sun with
most applications early in the weekend because of near school issues).

Chronic exposure estimates were also calculated using the targeted
ambient data.  These calculations should be considered relatively
uncertain estimates of exposure because of a lack of monitoring studies
specifically designed for this purpose and the manner in which seasonal
estimates were extrapolated.  Specifically, short- and intermediate-term
estimates were amortized to reflect a potential for exposure of 180 days
out of each calendar year in order to calculate chronic estimates of
exposure.  This was determined based on the approximate use patterns for
methyl bromide over a year in high use areas.  This approach does
introduce the potential for significant uncertainty into the estimates,
however, the Agency views the potential for chronic exposures in high
use regions as significant and has addressed this scenario in order to
be health protective.  Because there are many uncertainties associated
with the approach used in this assessment it is difficult to determine
how these estimates either over- or under-predict actual chronic
exposures for those living in high use areas.  There are several factors
that should be considered:

Monitoring was specifically targeted toward areas of high use, this
limits the  populations for which these types chronic exposure estimates
could be applied (i.e., for those living in such regions);

More refined amortization approaches on a regional basis could be
possible with use data, especially in California, but in most regions
such data are not available; and

Targeted monitoring was conducted during selected seasons of high use,
but because the data are limited, the impacts of changing conditions
(e.g., from different pest pressures, use patterns, or extended seasons)
cannot be quantified, especially for different regions of the country
with different climates, which could lead to potentially missing higher
end exposures under some conditions.

Acute exposures (24 hour TWAs) for all of the monitoring stations
considered (i.e., 30 stations), do not exceed the Agency's level of
concern (i.e., all calculated MOEs > 30).  Results were similar for the
8 week TWA exposures (short- and intermediate-term exposures) in that
risks calculated based on results from none of the monitoring stations
exceed the level of concern.  For both durations of exposure, risks were
orders of magnitude less than the level of concern (i.e., MOEs >30 in
many cases, some by 2 or 3 orders of magnitude).  These results should
be considered in conjunction with the fact that these studies were
deemed to be worst-case situations as described by CDPR above.  Risk
estimated from chronic exposure based on extrapolating seasonal CARB
data were also calculated because monitoring data specifically meant to
establish chronic exposure levels in high use areas were not available. 
Based on this approach, in some cases, chronic risks exceeded the level
of concern (i.e., MOEs < 100 for 13 of 46 stations/years); however, it
is believed that these results do not pose an imminent health concern to
the general public due to the nature of the calculations as described
above.  CDPR also reached similar conclusions that risks resulting from
exposure to ambient air were of minimal concern.

Table 21: Results of 2000 Through 2002 California Ambient Monitoring In
High Use Areas During Season Of Use

CA.

County	Data Source	Site	Dates &

Mon.

Days (N)	Maximum 24 Hr. TWAs (ppb)	Acute MOE!	8 Week TWA  (mean of
means) (ppb)	Short and Intermediate-Term MOE2	Amortized 180 days (ppb)
Calculated Chronic MOE3

Kern	CARB	ARB	7/10-9/1, 2000 (25)	0.996	10040	0.189	5291	0.09	966



	6/30-8/31, 2001	0.31	32258	0.12	8333	0.06	1521



SHA	7/10-9/1, 2000 (26)	3.52	2841	0.792	1263	0.39	230



CRS	7/10-9/1, 2000 (24)	14.2	704	2.16	463	1.07	84



	6/30-8/31, 2001	33.50	299	2.49	402	1.23	73



MVS	7/10-9/1, 2000 (26)	0.487	20534	0.092	10870	0.05	1984



	6/30-8/31, 2001	0.23	43478	0.08	12500	0.04	2281



VSD	7/10-9/1, 2000 (26)	0.247	40486	0.099	10101	0.05	1843



	6/30-8/31, 2001	0.23	43478	0.08	12500	0.04	2281



MET	7/10-9/1, 2000 (26)	0.224	44643	0.084	11905	0.04	2173



	6/30-8/31, 2001	0.25	40000	0.07	14286	0.03	2607



ARV	6/30-8/31, 2001	0.22	45455	0.07	14286	0.03	2607

Ventura	AMBI

(CDPR Stats Used)	SHA	8/15-10/10, 2001	2.94	3401	0.50	2000	0.25	365



	7/10-8/31, 2002 (31)	5.77	1733	0.58	1724	0.29	315



ABD	8/15-10/10, 2001	0.44	22727	0.18	5556	0.09	1014



	7/10-8/31, 2002 (30)	3.44	2907	0.76	1316	0.37	240



UWC	8/15-10/10, 2001	4.35	2299	0.82	1220	0.40	223



	7/10-8/31, 2002 (26)	13.17	759	2.22	450	1.09	82



PVW	8/15-10/10, 2001	3.17	3155	0.56	1786	0.28	326



	7/10-8/31, 2002 (32)	9.51	1052	1.62	617	0.80	113

Santa Barbara	AMBI

(CDPR Stats Used)	PLN	8/23-10/9, 2001	2.69	3717	0.93	1075	0.46	196



EDW	8/23-10/9, 2001	11.15	897	1.32	758	0.65	138



AGC	8/23-10/9, 2001	1.16	8621	0.28	3571	0.14	652



BLO	8/23-10/9, 2001	4.55	2198	0.73	1370	0.36	250



SLO	8/23-10/9, 2001	1.12	8929	__	__	__	__

Monterey	CARB	SAL	9/11-11/3, 2000 (31)	7.91	1264	1.29	775	0.64	141



	9/8-11/7, 2001	9.25	1081	1.38	725	0.68	132



OAS	9/11-11/3, 2000 (31)	1.84	5435	0.387	2584	0.19	472



CHU	9/11-11/3, 2000 (31)	2.41	4149	0.644	1553	0.32	283



	9/8-11/7, 2001	1.84	5435	0.56	1786	0.28	326



LJE	9/11-11/3, 2000 (30)	24.0	417	3.79	264	1.87	48



	9/8-11/7, 2001	14.49	690	2.82	355	1.39	65

	AMBI

(CDPR Stats Used)	BBC	9/4-10/26, 2002 (32)	6.28	1592	2.08	481	1.03	88



MAQ	9/4-10/26, 2002 (32)	4.53	2208	1.12	893	0.55	163

Santa Cruz	CARB	PMS	9/11-11/3, 2000 (31)	30.8	325	7.68	130	3.79	24



	9/8-11/7, 2001	21.08	474	2.99	334	1.47	61



SES	9/11-11/3, 2000 (31)	16.4	610	2.60	385	1.28	70



	9/8-11/7, 2001	5.31	1883	1.22	820	0.60	150



MES	9/8-11/7, 2001	36.64	273	5.51	181	2.72	33



SCF	9/8-11/7, 2001	0.74	13514	Not Sampled	__	__	__

	AMBI

(CDPR Stats Used)	WAT	9/4-10/26, 2002 (30)	16.38	611	3.79	264	1.87	48



FRM	9/4-10/26, 2002 (31)	14.00	714	2.62	382	1.29	70



CPW	9/4-10/26, 2002 (30)	11.12	899	2.06	485	1.02	89



SCF	9/4-10/26, 2002 (7)	0.69	14493	NA	--	--	--





Background site sampled only during 2 nonconsecutive weeks

1.  	Acute MOE based on maximum 24 hr. TWA.

2.  	Short term  and intermediate term MOE are based on 8 wk. TWA (i.e.,
mean of weekly means).

3. 	Chronic MOE based on short-/intermediate-term exposures amortized
for 180 days exposure per year.



5.1.2.2	Exposures From Urban Background Ambient Air Monitoring tc \l4
"6.1.2.1	Exposures From Regionally Targeted Non-Point Source Ambient Air
Monitoring 

In 2002, CARB added methyl bromide to its list of contaminants for which
it routinely screens in its TAC program (see
http://www.cdpr.ca.gov/docs/empm/pubs/tac/monitoring.htm).  The location
of these monitoring stations, however, shifted from a potential
“worst-case”, in-season use situation as described in section
5.1.2.1 above to the following:

“The ARB has a network of stations that routinely monitor California's
air for a variety of pollutants such as ozone, particulate matter,
metals, and other toxic air contaminants. In 2002, ARB began monitoring
for two pesticides, Methyl Bromide and 1,3-dichloropropene, every 12
days at approximately 20 stations in primarily urban areas throughout
the State.”  

The following should also be considered (see
http://www.arb.ca.gov/aqd/toxics/toxuses.html):

“The toxics sampling network was designed to produce a statewide
annual average to support the determination of a statewide risk
assessment. Where fewer than 12 continuous months of data are present,
we believe that it is seldom appropriate to calculate an annual average.
Most of the toxic substances show some seasonal variation, and some
substances differ by as much as two orders of magnitude between the high
and low periods of the year. If a month's data are missing, the
calculated average could be radically different from the real average,
the average that would have been calculated had the missing month's data
been available.”

TAC monitoring sites are located throughout California in urban
environments that include urban areas such as Long Beach, Burbank, Los
Angeles, Fremont, Fresno, San Francisco and San Jose.  The statistical
summaries of the 2002/2003 CARB monitoring data are provided in Table
22.  They were taken directly from 
http://www.arb.ca.gov/adam/toxics/statepages/mbrstate.html. 
Additionally, state-wide summaries from 2004 and 2005 are included in
Table 22.  [Note:  Station-specific results for 2004 and 2005 are also
available but do not differ significantly from the earlier years (i.e.,
they do not alter the interpretation of the risks associated with urban
background levels of methyl bromide).  As such, they have not been
included in Table 22 for brevity.]

Maximum values at each station were compared to acute HECs to estimate
acute MOEs.  Short- and intermediate-term risks were estimated by
comparing means to the short- and intermediate-term HECs.  Means were
selected for this analysis because they appear in most cases to be
heavily influenced by the typical 6 to 8 week use season based on the
relative contributions of a relatively small number of samples and that
medians for most stations were reported as the level of detection. 
Medians from each location were used to calculate chronic MOEs.  True
chronic exposures (continuous exposures >6 months) in and around most of
the monitored sites probably do not occur because the limit of detection
(½ LOD or 0.015 ppb) has been reported as the median for approximately
75 percent of the stations for each year where there are data.  These
monitoring data indicate that exposure patterns track with seasonal use;
therefore, shorter duration exposures are more prevalent which reflects
the seasonal use of most methyl bromide in California.

No exposure levels reported by through the TAC program exceed the level
of concern for any duration of exposure including acute (all MOEs >30),
short-, intermediate-term (all MOEs >30), or chronic (all MOEs > 100)
exposures in an urban environment.

Table 22:  Results of 2002 Through 2005 California Ambient Monitoring
For Methyl Bromide In Urban Areas



Site	

Year	

N	

Results of Annual Methyl Bromide Monitoring (ppb)



	

Maximum	

Acute MOE!	

Mean	

Short and Intermediate-Term MOE2	

Median	

Chronic MOE3



Statewide	

2002	

440	

0.91	

10989	

0.042	

23810	

0.015	

607

	

2003	

503	

0.90	

11111	

0.040	

25000	

0.015	

607

	2004	503	1.1	9090	0.036	27778	0.015	607

	2005	510	0.69	14493	0.036	27778	0.015	607



Azusa	

2002	

27	

0.14	

71429	

0.041	

24390	

0.03	

303

	

2003	

28	

0.16	

62500	

0.036	

27778	

0.015	

607



Burbank	

2002	

30	

0.14	

71429	

0.031	

32258	

0.015	

607

	

2003	

26	

0.10	

100000	

NR	

--	

0.015	

607



Calexico	

2002	

29	

0.11	

90909	

0.020	

50000	

0.015	

607

	

2003	

30	

0.33	

30303	

0.036	

27778	

0.015	

607



Chula Vista	

2002	

29	

0.06	

166667	

0.021	

47619	

0.015	

607

	

2003	

28	

0.05	

200000	

NR	

--	

0.015	

607



El Cajon	

2002	

28	

0.06	

166667	

0.020	

50000	

0.015	

607

	

2003	

30	

0.05	

200000	

0.021	

47619	

0.015	

607



Los Angeles	

2002	

21	

0.14	

71429	

NR	

--	

0.03	

303

	

2003	

29	

0.10	

100000	

0.032	

31250	

0.015	

607



Long Beach	

2002	

25	

0.11	

90909	

0.035	

28571	

0.015	

607

	

2003	

27	

0.13	

76923	

0.035	

28571	

0.04	

228



Riverside	

2002	

25	

0.13	

76923	

NR	

--	

0.015	

607

	

2003	

30	

0.10	

100000	

0.028	

35714	

0.015	

607



Simi Valley	

2002	

26	

0.91	

10989	

0.101	

9901	

0.05	

182

	

2003	

31	

0.90	

11111	

0.120	

8333	

0.015	

607



Bakersfield	

2002	

29	

0.22	

45455	

0.058	

17241	

0.04	

228

	

2003	

29	

0.88	

11364	

0.080	

12500	

0.04	

228



Chico	

2002	

29	

0.14	

71429	

0.026	

38462	

0.015	

607

	

2003	

31	

0.15	

66667	

0.022	

45455	

0.015	

607



Fremont	

2002	

27	

0.05	

200000	

0.018	

55556	

0.015	

607

	

2003	

30	

0.11	

90909	

0.019	

52632	

0.015	

607



Fresno	

2002	

30	

0.19	

52632	

0.049	

20408	

0.015	

607

	

2003	

31	

0.19	

52632	

0.055	

18182	

0.05	

182



Roseville	

2002	

29	

0.11	

90909	

0.021	

47619	

0.015	

607

	

2003	

31	

0.03	

333333	

0.016	

62500	

0.015	

607



San Francisco	

2002	

15	

0.08	

125000	

NR	

--	

0.015	

607

	

2003	

31	

0.015	

666667	

0.015	

66667	

0.015	

607



San Jose - 4th Street	

2002	

8	

0.09	

111111	

NR	

--	

NR	

--



San Jose - Jackson St.	

2002	

6	

0.05	

200000	

NR	

--	

NR	

--

	

2003	

31	

0.23	

43478	

0.031	

32258	

0.015	

607



Stockton	

2002	

27	

0.90	

11111	

0.144	

6944	

0.05	

182

	

2003	

30	

0.48	

20833	

0.088	

11364	

0.04	

228



Mexicali - Mexico	

2002	

19	

0.10	

100000	

NR	

--	

0.015	

607

	

2003	

17	

0.07	

142857	

NR	

--	

0.015	

607



Rosarito - Mexico	

2002	

25	

0.05	

200000	

NR	

--	

0.015	

607

	

2003	

30	

0.14	

71429	

0.027	

37037	

0.015	

607

1.  	Acute MOEs based on maximum concentrations.

2.  	Short term and intermediate term MOE are based on the mean
concentrations.

3.  	Chronic MOEs are based on the median concentration.

5.2	Bystander Risk Characterization tc \l2 "6.2	Bystander Risk
Characterization 

It is believed that the data and methodologies used in the development
of this assessment represent the state-of-the-science relating to
pesticides that can be characterized as fumigants.  However, it is clear
that there is an ongoing evolution relating to the types of data that
could be used to complete such assessments in the future.  Essentially,
all data that were currently available were used herein but those data
clearly have limitations related to overall quality, as well as temporal
and spatial limitations.  It is also clear that the PERFUM modeling
framework provides significant amounts of information appropriate for
risk managers to consider but that there are other systems that could be
considered as robust for the same types of analyses.  As indicated
above, submissions based on other viable modeling frameworks would be
considered for risk management purposes.  

Some of the limitations and considerations that have been identified
that should be considered in the interpretation of these results
include:

All of the data used for this analysis have been generated in
California; however, methyl bromide is used in many regions of the
country.  In fact it is possible that most methyl bromide use occurs in
the Southeast but no emissions data are available for specifically
modeling emissions from that region or any other region where methyl
bromide is used.  As a direct result of a lack of additional data, the
results presented herein are based solely on California data and
agricultural practices.  Factors such as soil type, solar radiation
levels, or farming practices themselves may impact the overall amounts
of methyl bromide emitted and the rate at which it is emitted over time,
thus buffer outputs predicted using PERFUM could be impacted but it is
not possible to quantify this sensitivity at this point.  PERFUM is not
a first-principles model (i.e., it cannot predict results for
incremental changes in soil conditions parameters such as soil
temperature or percent moisture).  Instead, PERFUM is an empirical model
that is calibrated to specific emissions profiles that then serve as the
basis for predicted results.  This is a very common modeling approach
when first-principles models are not available.  Additionally, the flux
profiles that were used as the basis for the PERFUM results in this
assessment were defined based on what is known as the back-calculation
method as opposed to direct reading or aerodynamic flux approaches.  At
this point, the Agency treats results based on all of these methods
similarly as there is no information in the applicable literature to
suggest that any are inherently biased to over- or under-prediction of
flux.  Flux inputs for the greenhouse and residential uses are based on
the same approaches as D304623. 

Data quality associated with the calculation inputs also needs to be
considered.  For most of the data, it is believed that they are of
reasonable quality for risk assessment purposes.  A significant number
of studies were generated by CARB and CDPR.  Because of the rigorous
quality control and quality assurance protocols employed by these
Agencies, the resulting data are generally believed to be of high
quality.  Additionally, the industry group AMBI also generated a
significant amount of data.  Some technical issues have been identified
with those data (i.e., problems with completeness and flow controller
issues in some data) but they have been included for comparative
purposes.  AMBI-based results are generally similar to those observed
based on the other data which indicates no major impacts based on data
quality.

The premise of the PERFUM-predicted buffer zones is based on a
developmental effect which by definition, implies that a pregnant woman
is at a location where the time-weighted average concentration of methyl
bromide can be of concern.  There are several probabilities related to
an exposure event which must be met for such developmental effects to
occur.  These include:

An exposed individual must be female and at a critical phase of the
pregnancy for the effect to occur;

An exposed individual has to be in proximity to a methyl bromide
application/aeration event for a sufficient duration for the effect to
occur - there are 3 key factors to consider for this element including:

that the types of applications considered in this assessment are either
seasonal or infrequent which limits the number of possible adverse
exposure events, 

time-activity data indicate that many parts of the population move from
site to site on a daily basis (e.g., to go to work and back) which
limits the overall number of possible adverse exposures events, and 

time-activity data indicate that most individuals spend a majority of
their daily time indoors and for intermittent exposure sources such as
this it is known that being indoors typically reduces exposures to
contaminants relative to outdoor air but the PERFUM results do not
account for this exposure reduction factor. 

A multi-faceted approach was used to evaluate risks using monitoring
data, incidents, and the distributional model, PERFUM.  Monitoring data
have temporal and spatial limitations as has been discussed above. 
Incident data are informative but their use for characterization
purposes is also limited because the exact conditions that lead to
specific incident events are difficult, if not impossible, to quantify
thus making any reconstruction of their circumstances limited unless
blatantly obvious such as a railcar accident.  However, for methyl
bromide, most incidents in the general population are believed to be
associated with a significant equipment failure, atmospheric inversion,
or misuse of some sort, intentional or not.  Because the endpoint of
concern associated with methyl bromide is developmental in nature, it is
highly likely that individuals in the general population would not
associate such health issues with a previous exposure to methyl bromide.
 For this reason, it is possible that methyl bromide incidents in the
general population could be under-reported.  Effects that would be
likely to be experienced in the general population after a methyl
bromide incident would be irritant effects from its companion chemical,
chloropicrin, for many applications.  Symptoms experienced such as
shortness of breath or chest tightness are not readily attributable to
methyl bromide exposure.

PERFUM modeling results (or any distributional model for that matter)
can provide risk managers with much needed information about the range
of risks expected in the general population.  At this time, policy
development is ongoing with regard to defining how appropriate
selections of PERFUM outputs can be defined for risk management purposes

It is believed that PERFUM provides the most refined estimates of risk
because it can consider actual weather data and also integrate flux
distributions in order to develop distributional estimates of buffer
distances and concentrations at various distances from a source.  PERFUM
uses ISCST3 as its core processor which is an existing Gaussian plume
technology that has been utilized for air permitting by the Agency for
many years (see Technology Transfer Network Support Center for
Regulatory Air Models at http://www.epa.gov/scram001/tt22.htm#isc). 
Several issues need to be considered related to the modeling analysis
which was completed herein.  These include:

It has been assumed that there is a linear relationship between
application rate and flux, but this assumption has not been validated
with emissions data conducted in similar conditions but at different
application rates.  The California Department of Pesticide Regulation
has used this assumption in previous fumigant assessments.

The treatment of calm periods (wind speeds below 1 m/s) in PERFUM/ISCST3
is also an uncertainty.  PERFUM runs the ISCST3 model in the
“regulatory default option” (the default setting for ISCST3), which
includes the use of the calms processing routine as is described in
Agency guidance.  The calms processing routine for wind speeds below 1
m/s essentially ignores any hourly sequence in the calculations that
meets this criteria.  This approach can possibly skew results for
shorter averaging times because an analysis period that contained
several calm hours would be dominated by any period where there was a
windspeed above 1m/s.  This is a common approach in Gaussian plume
modeling.  PUFF-based models such as CALPUFF have meander algorithms
that can account for calm conditions by accounting for static or near
static plume conditions and representing such events in the results. 
Whether or not buffer estimates are enhanced or under-reported as a
result of this phenomenon depends upon the nature of the weather data
used for the calculations.  Preliminary analysis related to this issue
do not indicate significant differences when hourly calculation steps
are used especially when 24 hour time weighted average exposures are
calculated.  If less than hourly steps (e.g., minute by minute
calculations such as in CALPUFF v6) are used, the effect is attenuated
because the relative percentage of calm periods in the available weather
data seems to be diminished.



The PERFUM analyses completed for this assessment are based on the
assumption that an application has an equal probability of occurring
each day out of the 5 years of weather data.  This method does not take
into account the seasonal use patterns of fumigants in different regions
of the country.  Table 14 above provides an example of monthly
distributional results which could be examined if so desired for every
PERFUM output.  The result for each month is based on 5 months worth of
weather data instead of 5 years when all months are considered.  It
should also be noted that the selection of the sources of weather data
for this assessment, as mentioned above, represent a range of mean
windspeed values as described in the SAP document for PERFUM.  The
locations of the Florida and California stations were intentionally
selected based on this range and their coastal and inland locations.

Different field sizes and aspect ratios were considered in this
assessment (most fields were square in shape for this analysis).  As
field size increases so do predicted buffer zones which is similar to
what is noted based on increases in application rates.  Field aspects
were also examined and the orientation did impact results although it is
difficult to ascertain any general prediction based on this analysis
since field orientations relative to prevailing wind directions will
vary from site to site or region to region.

The use of a maximum 40 acre field in the risk assessment may possibly
understate potential exposure received by bystanders near treated fields
that are larger.

PERFUM was recently modified to also produce distributions of
concentrations at various receptor ring distances from the edge of a
treated field or source.  This capability was added near the completion
timeframe of this assessment.  As appropriate, this capability will be
utilized in the development of risk management decisions.  

Several factors also need to be considered in the interpretation of the
results associated with the assessment of exposures from ambient air. 
It is clear from the characterization of the data provided by CARB that
some data represent highly targeted monitoring in a region during the
season of use and others represent urban background levels.  Because of
this, the results should be considered representative for the state of
California for those types of situations.  However, California has a
number of restrictions and systems in place where the overall goal is to
reduce environmental emissions from fumigant use.  Consequently, it is
difficult to quantify how the results presented above may apply to other
regions of the country that do not have these types of programs in
place.  

6.0	Occupational Exposure tc \l1 "9.0	Occupational Exposure 

In this assessment, a number of chemical-specific monitoring studies
were available to evaluate the exposures of applicators as well as those
otherwise involved in that process (e.g., co-pilots, shovelmen, or
greenhouse workers).  Likewise monitoring data were also available to
assess the possible exposures that could occur after application events
such as for tarp cutters and tarp removers.  [Note:  Tarp cutter and
tarp remover data were generated in a period 4 to 6 days (or greater)
after an application and most data represent exposures from fields
covered with high barrier films.]  The monitoring data were not adjusted
based on application rates because these data represent cultural
practices that are allowable under current methyl bromide labels even
though for many situations they could represent conservative estimates
of exposure.  This is because they were collected at higher application
rates than is currently typical.  Another consideration is the equipment
that was used because most of the data were generated in the 1980s and
early 1990s.  Since that time, California DPR has restricted methyl
bromide uses to types of equipment intended to minimize exposures. 
However, this evolution is not evident in all other areas of the country
so the data, even though they were generated years ago for the most part
in California, can still be considered to be representative of many
methyl bromide domestic uses.  

Overall, data indicate that worker exposure levels exceed the level of
concern for all scenarios considered when no respiratory protection is
used.  The use of either air purifying respirators (APRs) or self
contained breathing apparatus (SCBA) was also considered with varied
results.  The use of an APR reduces exposure levels by a factor of 10
and the use of SCBA reduces exposure levels by a factor of 10,000. 
[Note: There are commercially available APR cartridges that have been
evaluated and recommended for reducing exposure levels, see the
technical bulletin 146 for cartridge 60928 at www.3M.com for more
information.  Protection factors are derived based on individuals who
have been trained, fit tested, and medically cleared for respirator use
but are generally applied for risk assessment purposes.]  Respirators
would be the most practical personal protective equipment choice for
reducing exposures for most workers.  The use of engineering controls is
also considered a viable option for reducing exposures.  However, much
of the field monitoring data used for this analysis already reflect the
use of varied levels engineering controls such as tarps, tractor cabs,
deep injection, or other devices including fans in proximity to drivers,
Nobel plows, or various compaction methods.  

When APRs were evaluated with maximum exposure levels in order to assess
acute exposures, for most activities, exposures were not reduced
sufficiently to address risk concerns.  When APRs were evaluated with
mean exposure levels in order to assess short-/intermediate-term
exposures, the results were varied with all greenhouse scenarios
resulting in risks still exceeding the level of concern but risks from
most field uses not being of concern.  The use of SCBA is not normally
deemed to be a viable option for agricultural workers or for uses in
greenhouses due to the difficulty in handling and maintaining these
devices as well as the cost to implement them as an exposure reduction
tool.  However, in certain situations such as residential treatments, it
is believed that SCBA may represent a viable option for reducing
exposures for a limited number of workers.  Risks from exposures during
residential treatments were not of concern if SCBA is used.

Modeling has also been used to predict air concentrations at the edge of
a treated field to determine if air concentrations have diminished to
levels that are not of concern (e.g., use of concentration outputs from
PERFUM at varied ring distances from field perimeter).  This
information, coupled with available monitoring data can be useful for
defining how long the conduct of various post-application tasks or entry
into previously treated areas should be restricted.  Current
requirements for entry of post-application workers into previously
treated fields are dictated by the Worker Protection Standard as
described in PR 93-7.  For methyl bromide, such workers are excluded for
48 hours after treatment.  PERFUM outputs representing air
concentrations 5 meters from the treated field edge support the current
48 hour requirement for soil uses if tarps are not disturbed.  However,
tarp cutter and remover activities require APR use.  Greenhouse and
residential uses require active aeration to achieve levels that are not
of concern (e.g., current labels specify a 5 ppm target concentration). 
Generally, this approach seems appropriate but the 5ppm target may be
altered based on risk management considerations.

The occupational tasks commonly associated with the use of methyl
bromide along with the corresponding risks are described below for each
use sector considered.  [Note:  Risks from chronic exposures have not
been calculated because methyl bromide use is highly seasonal.  Methyl
bromide is not expected to be used every working day for more than 6
months, for commercial applicators or large scale growers based on
available information.  Additionally, in smaller scale production,
applications are thought to be infrequent because growers would often
times just treat their own facilities.]  Sections 6.1 through 6.3
provide risk estimates for each use pattern considered in this
assessment while Section 6.4 describes the issues that should be
considered when interpreting these results.

6.1	Pre-plant Agricultural Field Fumigations tc \l2 "9.1	Pre-plant
Agricultural Field Fumigations 

Several tasks were identified that are associated with the application
of methyl bromide to agricultural fields.  The data for each scenario
varied but essentially all available data for each was considered in
this assessment.  It should also be noted that for tarp cutters and tarp
removers, these activities were monitored mainly between 4 and 6 days
after application and in most cases the tarps were referred to as high
barrier materials. The tasks considered in this assessment include:

a) First Tractor Driver,

b) Co-pilot

c) Second Tractor Driver

d) Shovelman

e) Irrigation Worker

f) Tarp Cutter

g) Tarp Remover

Table 23 indicates that if no respiratory protection is considered,
risks exceed the level of concern for all scenarios and durations except
for the second tractor drivers where risks were acceptable even without
respiratory protection.  As a result, the impact of the use of an APR
(PF 10) with the recommended cartridge was evaluated to determine how
their use would impact worker risks.  Acute risks for all other
application scenarios with the APR still exceed HED’s level of concern
(with MOE <30).  Risks were not of concern for all tarp removal
operations if an APR is used (i.e., MOEs are 63 and 231).  Short- and
intermediate-term risks without respiratory protection were of concern
except for second tractor drivers as with the acute duration.  However,
short- and intermediate-term risks with the APR are not of concern for
the majority of tasks  (i.e., MOEs >30) except for tractor drivers and
co-pilots where the MOEs are 18 and 11, respectively.  [Note:  Table 23
has been modified from the Phase 3 assessment based on comments
submitted by the Methyl Bromide Industry Panel (MBIP) pertaining to the
specific data used for this analysis.  Essentially, many of the comments
are correct and the Agency has adjusted the values in Table 23
accordingly.  The MBIP comments are available in the methyl bromide
docket at the following:    HYPERLINK "http://www.Regulations.gov" 
www.Regulations.gov  under docket number EPA-HQ-OPP-2005-0123-0127.]

As indicated above, restricting entry into previously treated fields can
be evaluated using a modeling approach or based on the tarp removal
monitoring data.  It appears that a the emission profile (based on
preliminary PERFUM outputs) indicates that concentrations will decrease
to levels that are not of concern, even at high percentiles of exposure
3 days after treatment.  This indicates that the current 48 hour
restriction is adequately protective.  However, if operations occur
which disturbs the tarp of a treated field it is likely that air
concentrations will be at levels of concern and that respirator use
would be required to reduce them to levels not of concern.

Table 23: Methyl Bromide Worker Exposure Associated With Pre-Plant
Agricultural Field Fumigation

Scenario	N	Acute Risk Summary	Short/Intermediate-term Risk Summary



Maximum Monitored [mebr]

(ppm)	Acute MOE With No Respiratory Protection	Maximum Monitored [mebr]
With Air Purifying Respirator (PF10) 

(ppm)	Acute MOE With Air Purifying Respirator (PF10)	Average Monitored
[mebr]

(ppm)	Short-term MOE With No Respiratory Protection	Maximum Monitored
[mebr] With Air Purifying Respirator (PF10) 

(ppm)	Short-term MOE With Air Purifying Respirator (PF10)

1st Tractor Driver	82	38.1	<1	3.81	8	2.5	2	0.25	18

Co Pilot	92	47.4	<1	4.74	6	4.1	1	0.41	11

2nd Tractor Driver	3	0.02	1500.0	0.002	15000	0.015	293	0.0015	2933

Shovelman 	67	12.4	2	1.24	24	0.95	5	0.095	46

Irrigation	20	20.4	2	2.04	15	1.2	4	0.12	37

Tarp Cutter	7	4.8	6	0.48	63	0.67	7	0.067	66

Tarp Remover	22	1.3	23	0.13	231	0.48	9	0.048	92

Total number of monitoring events used for this analysis = 293

MOE = Margin of Exposure, Level of concern is MOE<30

Acute MOE = (30 ppm HEC/maximum [mebr])

Short/Intermediate-term MOE = (4.4 ppm HEC/mean [mebr])

	

6.2	Greenhouse Fumigations tc \l2 "9.2	Greenhouse Fumigations 

Several tasks were identified that are associated with the application
of methyl bromide in greenhouse operations.  The data for each scenario
varied but essentially all available data for each was considered in
this assessment.  The tasks considered in this assessment include:

a) Greenhouse Applicators

b) Venters

c)  Greenhouse Workers

Table 24 indicates for greenhouse applicators and workers, risks were of
concern for all scenarios and durations considered even when respiratory
protection such as an PF10 APR is used (i.e., all MOEs <30).  Risks for
greenhouse venters were of concern if no respirator is used but are not
of concern if an APR is used regardless of the duration of exposure
considered.

Determining when entry into previously treated greenhouse structures is
not of concern is important in the greenhouse industry because of the
many hand labor-based practices that are required to produce plant
materials in these settings (e.g., transplanting).  Currently,
greenhouse uses require active aeration to achieve levels that are not
of concern (e.g., current labels specify a 5 ppm target concentration). 
Generally, this approach seems appropriate but the 5ppm target may be
altered based on risk management considerations.  No further
quantitative assessment has been completed for these scenarios because
results will be specific to the structures which are treated.



Table 24: Methyl Bromide Worker Exposure Associated With Greenhouse
Fumigation

Scenario	N	Acute Risk Summary	Short/Intermediate-term Risk Summary



Maximum Monitored [mebr]

(ppm)	Acute MOE With No Respiratory Protection	Maximum Monitored [mebr]
With Air Purifying Respirator (PF10) 

(ppm)	Acute MOE With Air Purifying Respirator (PF10)	Average Monitored
[mebr]

(ppm)	Short-term MOE With No Respiratory Protection	Maximum Monitored
[mebr] With Air Purifying Respirator (PF10) 

(ppm)	Short-term MOE With Air Purifying Respirator (PF10)

Greenhouse Applicators 	11	2000	<1	200	<1	330	<1	33	<1

Greenhouse Venters	13	5.8	5	0.58	52	0.92	5	0.092	48

Greenhouse Workers	17	200	<1	20	2	25	<1	2.5	2

Total number of monitoring events used for this analysis = 41

MOE = Margin of Exposure, Level of concern is MOE<30

Acute MOE = (30 ppm HEC/maximum [mebr])

Short/Intermediate-term MOE = (4.4 ppm HEC/mean [mebr])

	

6.3	Residential Fumigation tc \l2 "9.5	Residential Fumigation 

Several tasks were identified that are associated with the application
of methyl bromide in residential operations (e.g., house preparation,
application and aeration activities).  The data for each scenario varied
but essentially all available data for each was considered in this
assessment.  However, the data were limited so the the tasks considered
in this assessment include:

Outside Worker

Inside Worker

Table 25 indicates for all scenarios and durations considered, risks
without respiratory protection exceed the level of concern for all
durations of exposure (i.e., all MOEs <30).  As a result, HED evaluated
how the use of a SCBA (PF 10,000) would impact worker risks since it is
believed that these represent a possible risk mitigation measure for
workers involved in residential fumigations.  SCBA use reduces risks to
levels where they are not of concern for all durations of exposure
(i.e., MOE>30).  

Determining when entry into previously treated residential structures is
not of concern is important not to displace residents for extended
periods of time.  Currently, residential uses require active aeration to
achieve levels that are not of concern (e.g., current labels specify a 5
ppm target concentration).  Generally, this approach seems appropriate
but the 5ppm target may be altered based on risk management
considerations.  No further quantitative assessment has been completed
for these scenarios because results will be specific to the structures
which are treated.



Table 25: Methyl Bromide Worker Exposure Associated With Residential
Fumigation

Scenario	N	Acute Risk Summary	Short/Intermediate-term Risk Summary



Maximum Monitored [mebr]

(ppm)	Acute MOE With No Respiratory Protection	Maximum Monitored [mebr]
With SCBA (PF10000) 

(ppm)	Acute MOE With SCBA (PF10000)	Average Monitored [mebr]

(ppm)	Short-term MOE With No Respiratory Protection	Maximum Monitored
[mebr] With SCBA (PF10000) 

(ppm)	Short-term MOE With SCBA (PF10000)

Outside Worker	12	57	<1	0.0057	5263	22	<1	0.0022	2000

Inside Worker 	4	982	<1	0.098	306	357	<1	0.036	122

Total number of monitoring events used for this analysis = 16

MOE = Margin of Exposure, Level of concern is MOE<30

Acute MOE = (30 ppm HEC/maximum [mebr])

Short/Intermediate-term MOE = (4.4 ppm HEC/mean [mebr])



6.4	Occupational Risk Characterization tc \l2 "9.5	Residential
Fumigation 

There are several issues that should be considered when interpreting the
results of this risk assessment.  Compared to most occupational
assessments, the data used to complete this assessment are plentiful in
that 293 chemical-specific monitoring events were considered in this
analysis.  Most data were generated by either CDPR or investigators with
significant experience monitoring for methyl bromide using standard
capture methods.   

However, it should be considered that much of the data used in this
assessment are 10 years old or older and may not necessarily reflect
absolute state-of-the-art cultural practices in many areas of the
country.  For example, modern application controller systems, which are
becoming more widespread, are highly sophisticated programmable devices
that control applications in real-time by accounting for micro-changes
in application conditions such as changing line pressures or tractor
speeds.  It is possible that these systems could reduce exposure levels
but their impact is not reflected in the data upon which this assessment
is based.  Other examples of practices that may not be captured include
state-of-the-art shank designs that are intended to reduce emissions or
other emission reduction approaches such as high barrier films or
compaction methods.

In some cases, varied engineering controls were used by those being
monitored so they are inherent in the resulting exposure levels. 
Examples of such devices include fans located near the tractor driver to
dilute potential methyl bromide exposures.  Varying shank depths or soil
sealing methods could also reduce exposures but again the impacts of a
single factor are difficult to ascertain based on the available data. 
Rather than to attempt to further delineate the data based on these
factors, the data were grouped as presented above recognizing that each
task category presented above probably represents a range of exposures
that could occur for that task based on varied available equipment. 
Alternatively, if an attempt was made to quantitatively assess each
possible scenario the numbers of monitoring events for each category
would sometimes be very low (i.e., in some cases, 1 or 2 events) which
minimizes the ability to compare results among scenarios.  In summary,
the data used for this assessment should be considered acceptable for
risk assessment purposes and that, if anything, current agricultural
practices and systems would likely lead to lower exposure estimates than
those presented above.  This is likely because the intent of the
evolving cultural practices over recent years has been to provide for
more efficient and accurate application events in a manner that still
achieves efficacy and to also reduce overall emissions and exposures.

The results of the occupational risk assessment should also be
considered in the context of the available incident data.  In many
cases, occupational incidents involve skin sensitization or even
chemical burns which are consistent with the toxicity profile for methyl
bromide (i.e., it is a Category 1 acute dermal and eye irritant – see
Appendix A).  For this reason, loose fitting clothing is stipulated on
methyl bromide labels so as not to trap residues close to the skin. 
Irritation effects are also often noted as described above which are
difficult to differentiate from cholorpicrin exposure which is often
companion applied with methyl bromide.  As indicated above the
bystanders, it is also not surprising that other types of incidents are
not observed more frequently because some of the symptoms that could be
expected from exposure to methyl bromide may not be readily attributable
to those exposures.

[Note:  If so desired, the previous risk assessment (D316326, 7/13/05)
and its associated appendices that contain all of the monitoring data
information can be found at   HYPERLINK "http://www.Regulations.gov" 
www.Regulations.gov  under docket OPP-2005-0123.]

7.0	Data Needs and Label Requirements tc \l1 "10.0	Data Needs and Label
Requirements 

7.1	Toxicology tc \l2 "10.1	Toxicology 

There are no additional data requirements at this time.

7.2	Residue Chemistry tc \l2 "10.1	Toxicology 

Please refer to the Phase 5 risk assessment for the food uses of methyl
bromide (D304623).

7.3	Occupational and Residential Exposure tc \l2 "10.3	Occupational and
Residential Exposure 

The assessment of occupational and residential risks associated with the
use of methyl bromide is complex.  There was a significant amount of
data available but additional data are still required in order to ensure
that exposures related to current cultural practices can be addressed in
future assessments.  Additionally, emissions data are required in order
to quantify the impact of using technologies intended to improve
efficacy and/or reduce emissions in order to develop model estimates
reflective of those technologies (e.g., impacts of high barrier films on
emissions).  

The types of data, guideline citations, and examples of the scenarios
which need to be addressed are presented below.  It should also be noted
that the general data requirements for fumigants are being evaluated. 
Factors such as application technologies, regions of use, possible
emission reduction technologies, and timing of applications will be used
to define factors such as the numbers of emissions studies which may be
required for specific use patterns, the meteorological data requirements
for modeling purposes, methods for flux calculations, and possible
criteria for ambient monitoring.  Other factors may include specific
guidance for sampling methods, numbers of samples and frequency of
sampling.  Until this effort is complete, final determination of the
types of studies to address each scenario below should be made in
consultation with the Agency.  

OPPTS Guideline 835.8100 - Field volatility from soil

Volatility Studies To Determine Flux For Modeling Purposes In Major Use
Regions Of Country For Significant Application Methods (e.g., Florida or
Washington for tarped raised beds or drip irrigation)

OPPTS Guideline 875.1300 - Inhalation exposure for applicators
(outdoors)

Pre-Plant Field - (e.g., rig drivers & tenders, tarpers, tarp removers)

OPPTS Guideline 875.1400 - Inhalation exposure for applicators (indoors)

Greenhouse - (e.g., Fumigators, Media Handlers, Aerators)

Residential - (e.g., Fumigators, Tarpers, Aerators)

OPPTS Guideline 875.2500 - Inhalation exposure for postapplication
workers

Pre-Plant Field - (e.g., planters)

Greenhouse - (e.g., forklift drivers, plant maintenance workers)



Requirements For Special Studies

Meteorological data for probabilistic modeling purposes

Product Use Information By Major Use Region, Frequency, Application
Parameters (e.g., rate, acres treated, data, application equipment and
emission control technologies used)

Indoor air concentrations from residences that are in proximity of
treated areas

Ambient monitoring in high regions of use in the season of use.

 Since no effects were reported at 11 ppm during the first 5 weeks (24
exposures), exposure concentration of methyl bromide was increased from
11 ppm to 158 ppm for 6 additional days.

  Due to the limited severity of the effect, HED considered that a 3X UF
would be sufficient to extrapolate from the LOAEL to the NOAEL.

	Majewski, MS, Glotfelty, DE, Seiber, JN. 1989. A comparison of the
aerodynamic and the theoretical-profile-shape methods for measuring
pesticide evaporation from soil. Atmospheric Environment, 23:929-938

Majewski, MS, Glotfelty, DE, Kyaw Tha Paw U, Seiber, JN. 1990. A field
comparison of several methods for measuring pesticide evaporation rates
from soil. Environmental Science and Technology, 24:1490-1497.

Parmele, LH, Lemon, ER, Taylor, AW. 1972. Micrometerological measurement
of pesticide vapor flux from bare soil and corn under field conditions.
Water Air Soil Pollut. 1:433-451

 PAGE  v 

 PAGE  70 

 

 

Figure 7: Flux Rate Estimation, Area Under the Curve

Inputs -

Varied flux

& weather data

Air 

Dispersion

Model (ISC, AERMOD,

CALPUFF)

Outputs -

Distributions of

buffer zones or air concentrations

“Whole Field”

“Maximum Distance”

