facebook twitter LinkedIn flickr Member Login logo

Comments on Sulfuryl Fluoride

July 1, 2011

Office of Pesticide Programs
Regulatory Public Docket (7502P)
United States Environmental Protection Agency
Washington, DC

Sent via Federal eRulemaking Portal: http://www.regulations.gov

Re: Docket ID EPA-HQ-OPP-2005-0174

Dear Madam/Sir:

The North American Millers’ Association (NAMA) is the trade association
representing 43 companies that operate 170 wheat, oat and corn mills in 38
states and Canada. Their collective production capacity exceeds 175 million
pounds of product each day, which is more than 95 percent of the total
industry production.

These comments are submitted relative to the Environmental Protection
Agency’s (Agency) proposed resolution of objections and a stay request with
regard to sulfuryl fluoride (SF) and fluoride tolerances promulgated in 2004
and 2005 under section 408(d) of the Federal Food, Drug, and Cosmetic Act

This document will address the Proposed Order in detail, and our comments
can be summarized as follows:

  • Risk is a function of both hazard and exposure. NAMA’s comments will
    focus on exposure.
  • The contribution of milled grain products to total dietary exposure to
    fluoride is insignificant.
  • As the Agency notes in multiple locations in the Proposed Order, NAMA
    members have very limited alternatives. Methyl bromide (MB) has
    been nearly phased out and high heat is not practical for many mills.
  • It is impossible to clean a mill or milling system so thoroughly that
    there would be zero exposure of food in a regulatory environment that
    allowed SF usage without residue tolerances.
  • The Agency’s Proposed Order apparently does not contemplate
    naturally occurring background levels of fluoride present in grain, nor
    their effect on enforcement decisions should they be detected.
  • As described by the Agency, eliminating SF could result in increased
    contamination of milled grain products with insects or insect parts.
  • Residues on food products or ingredients legally treated and imported
    from any of the 16 countries where SF is approved would be illegal in
    the US.
  • Banning SF will do little or nothing to resolve the issue of high fluoride
    levels in the diets of some sub-populations.
  • Since SF presents no public health concerns, the Agency should use its
    discretionary authority to apply the de minimis treatment that EPA has
    afforded other sources of insignificant fluoride exposure.

Millers are committed to maintaining the highest standards of sanitation in
which to manufacture wholesome and nutritious milled grain foods. It
requires substantial resources to provide that level of sanitation in a modern
high-speed, high-volume manufacturing environment.

The milling industry has been a pioneer in developing and adopting
Integrated Pest Management (IPM) techniques. The use of IPM strategies has
allowed the milling industry to substantially reduce its use of chemical
fumigants, and all pesticides. This has improved our overall environmental
footprint and the environmental sustainability of our businesses. Still, despite
this progress some use of general space fumigants – either SF or MB can be

As noted in the proposal, for years the Agency has been advocating the
adoption of SF as quickly as possible as a replacement for MB. NAMA finds
the actions outlined in this proposal contrary to that advice and especially

The Agency acknowledges the findings of the National Research Council with
respect to the need to lower fluoride levels in naturally fluoridated drinking
water and private water systems. Adoption of policies to respond to these
findings would have a far greater impact on decreasing fluoride aggregate
exposure problems.

Why Fumigants Are Used
Raw grain is the miller’s only physical input. The grain arrives from the farm
field or farm bin directly, or from one commercial grain elevator or a
sequence of grain elevators. At each step in the marketing chain from the
farm to the mill the grain will be stored for some period of time. Between the
time of harvest and arrival at the mill the grain can be stored for months or
even years depending on the level of carryover stocks from previous
harvests. At a minimum, even in times with low carryover stocks grain will be
stored up to 12 months before processing.

As a biological commodity, the one thing that can be said with certainty
about grain quality is that it can never improve. The goal, therefore, is to
minimize any decline. That is the basis for deciding if and when to use

Fumigants, indeed all chemical pesticides, are by their very nature potentially
dangerous. They can be difficult to handle and store, and potentially put the
user at risk of very public criticism. So the decision whether or when to use
them is very carefully considered.

They are also expensive, not only in the cost of the chemical and its
application, but more so in lost revenues associated with closing the mill. For
example, industrial flour (sold for further food manufacturing, not packaged
for home use) currently sells for approximately $0.25/pound. An average
flour mill will grind about 1.0 million pounds of flour daily, for a gross daily
revenue of $250,000. A fumigation takes about 48 hours to complete from
start to finish, thus the lost revenues associated with fumigating a medium
size mill can total about $500,000. So the decision to use chemical pesticides
is never taken lightly. Also, since US milling industry generally operates
24/7, these lost revenues can never be made up by operating longer hours

In the marketing chain the grain is always inspected, sometimes more than
once. The inspectors may be licensed official inspectors, or not, but
regardless they always apply the quality standards of the United States Grain
Standards Act which was passed by Congress in 1916 and is administered by
the US Department of Agriculture. The official US Standards do little or
nothing to discourage the delivery to the miller of grain that will certainly
require remediation in the form of a pesticide tool.

Those standard include inspecting for live insects and insect-damaged
kernels (IDK). IDK is an important indicator of the level of hidden insect
infestation in the lot of grain. Certain insects, often in farm storage, chew
holes in wheat kernels prior to laying eggs inside the kernel. These eggs, if
not killed by chemical fumigants, will hatch and begin a new generation of
live infestation in the grain. The life cycle of most of these pests is only 28
days. So, if left unchecked, they will create a pest explosion.

NAMA response to Proposed Order
The Proposed Order contains a lengthy description of the Agency’s
assessment of the fluoride health risk. Our comments do not address risk
except in response to the described National Research Council (NRC)
analysis, conducted at the Agency’s request, that studied the issue of fluoride
in the diet and concluded that such fluorosis occurs at an “appreciable
frequency” in communities with water supplies containing at or near 4 mg/L
but that “the prevalence of severe enamel fluorosis would be reduced to
nearly zero by bringing the water fluoride levels in these communities down
to below 2 mg/L.” (Id.at 127-128).

To address the health effects issues described in the Proposed Order the US
Government could require remediation steps to address those locations with
naturally high levels of fluorine in the drinking water to bring them within the
NRC’s conclusion. This action would address the concerns outlined by the

However, with respect to estimating the human exposure contributed by
exposure to food, we have a great deal of experience in the use of fumigants
and so we can comment appropriately on that subject. Our experience
confirms the Agency’s statement that potential exposure to SF in mills has
been overstated.

Page 3426 In assessing exposure from pesticide residues in food, EPA, for
efficiency’s sake, follows a tiered approach in which it, in the first instance,
assesses exposure using the worst case assumptions that 100% of the crop
or commodity in question is treated with, or exposed to, the pesticide and
100% of the food from that crop or commodity contains pesticide residues at
the tolerance level.

Use of percent crop/commodity treated data and anticipated residue
information is appropriate because EPA’s worst-case assumptions of 100%
treatment and residues at tolerance value significantly overstate residue

NAMA response: We have no opinion as to whether the use of percent
crop/commodity treated data is appropriate from a science policy standpoint.
We do emphatically agree that, in this case, the Agency’s assumptions
significantly overstate the potential of residues, if they even exist at all. In
fact, we believe those assumptions overstate residue values so dramatically
as to render any decisions affected by the outcome of the analysis to be
deeply flawed.

For example, two different snapshot views of the potential SF residues
resulting from mill fumigation (the largest post-harvest use of fumigants)
would look like this:

Estimate A.

  1. After a mill fumigation it takes 15 minutes after re-start to achieve
    negligible residues
  2. 50 mills fumigated with SF annually
  3. 175 mills total in US
  4. 300 production days annually


175 x 300 = 52,500 industry production days annually,
which is 1.26 million production hours annually.

50 mills x 15 minutes potential exposure = 750 minutes, or 12.5 hours
aggregate potential exposure

12.5 ÷ 1.26 million = 0.000992% of the production hours could
potentially be exposed and theoretically contain residues


Estimate B

  1. average mill produces 1,000,000 lbs product per day, or 20,883 lbs in
    15 minutes
  2. 50 annual mill fumigations x 20,883 = 1,041,667 lbs potentially
    exposed annually industry-wide
  3. industry produces 175 million pounds daily
  4. 175mm lbs x 300 production days/year = 52.5 billion lbs/yr

    1,041,667 ÷ 52,500,000,000 = 0.001984% of production quantity
    could potentially contain residues

As can be seen here, whether calculating potential exposure on a production
time basis or production weight basis, the result is the same – post-harvest
use of SF does not contribute to exposure in any meaningful way.

p. 3426
Use of percent crop/commodity treated data and anticipated residue
information is appropriate because EPA’s worst-case assumptions of 100%
treatment and residues at tolerance value significantly overstate residue
values. ….This is true with food and structural fumigants such as sulfuryl
fluoride as well, especially with regard to the structural fumigant use in food
processing facilities because such use incurs infrequently and only potentially
affects a small portion of the food processed in the facility.

NAMA response: We agree, although we suggest the term small is in itself an
overstatement. It would be more precise to say the portion potentially
affected is insignificant. Further, we would also add that the level of residues,
where they exist, would be negligible or commonly below the limit of

There may be some commodities from a treated crop or commodity that
approach the tolerance value where the maximum label rates are followed,
but most generally fall significantly below the tolerance value.

NAMA response: We believe no, or close to no raw grains are treated with
SF, thus we are confident the tolerance values most certainly fall significantly
below the tolerance value even to the point of negligible or not detectable if
they exist at all. The use of SF to treat processed commodities is, while rare,
is also very important on those occasions. On occasion a tractor trailer of
bagged or boxed milled grain products must be fumigated to treat external
insects on the outside of the packages in order to prevent further damage.

Finally, EPA can consult monitoring data gathered by the FDA, the USDA, or
pesticide registrants, on pesticide levels in food at points in the food
distribution chain distant from the farm, including retail food establishments.

NAMA response: NAMA agrees. The National Agricultural Statistics Service
(NASS) Agricultural Chemical Use Program is the U.S. Department of
Agriculture’s official source of statistics about on-farm and postharvest
pesticide use and pest management practices. In the summer and fall of
2010, NASS collected data about pesticide applications on all wheat and pest
management practices for all grains handled at off-farm storage facilities.
This postharvest survey was conducted for wheat marketed from June 1,
2009 through May 31, 2010.

These results are based on 1,634 reports from operators in 14 program
states. Only 18.1 percent of post-harvest wheat was treated with any

The survey reports … other chemicals were the less commonly used class of
active ingredients. Program states reported using sulfuryl fluoride on
postharvest wheat. Due to the small number of reports of these chemicals,
data were withheld to avoid disclosing information for individual operations.

Said another way, the incidence rate of SF applications to post-harvest wheat
is so low that USDA chose not to even report it.

Report at http://quickstats.nass.usda.gov/results/9365A576-A7CE-30D7-

Additional relevant data is provided by the US Department of Agriculture’s
Agricultural Marketing Service that publishes its annual summary of the
Pesticide Data Program (PDP). Over a three-year period USDA collected 650
raw post-harvest corn samples from throughout the US. It subjected those
samples to 63,167 pesticide residue analyses; more than 99 percent of which
(62,745) showed no detectable pesticide residues of any kind. Of the
samples with positive pesticide detections, 13 different pesticides were
detected. None of the pesticides detected were SF.

PDP found at:

The similar report from a year earlier (2008) showed data from 660 corn
samples. 14 different pesticides were detected, none of which were SF.
Data at:

In 2006, 687 wheat samples were tested. 69 percent contained detectable
residues. The relatively high positive detection rate reflects the common
understanding that wheat is a food grain and thus protection from infestation
is more critical. However, all but one of the residue detections was much
lower than established tolerances and none of the detections was for SF


Page 3436
Fluoride exposure from residues of sulfuryl fluoride in food was estimated by
OPP based on usage data and residue data relevant to both sulfuryl fluoride’s
use as a direct commodity fumigant and as a structural fumigant.

NAMA response: As stated previously, we are unaware of any use of SF to
treat raw commodity. Therefore any exposure estimate that includes such
use would be flawed and misleading.

Page 3438
a. Fluoride from sulfuryl fluoride. In the exposure assessments for the 2004
and 2005 tolerance actions, EPA conducted a somewhat refined exposure
assessment of fluoride exposure in food from use of sulfuryl fluoride as both
a commodity fumigant and as a structural fumigant for food handling
facilities. Taking into account comments OPP has received from Dow
AgroSciences, OPP has further refined this aspect of the exposure
assessment. (Ref. 24). The three main refinements are:

(1) OPP used a regression analysis to estimate residue values of fluoride in
food that occur from actual use rates rather than assuming residue values as
measured under maximum application rates;

(2) OPP used a probabilistic analysis to estimate residues resulting from
possible sequential treatment of food (e.g.,fumigation of raw commodity,
incidental treatment during fumigation of structure, fumigation of the
processed commodity) rather than conservatively assuming that 100% of
food was sequentially treated; and

NAMA response: 1) We are unaware of any use of SF to treat raw commodity. 2) As described elsewhere, the residues that might conceivably be present as a result of a structural fumigation would be rare and likely lower than the limit of detection. Thus, even the Agency’s probabilistic analysis certainly overestimates residues to a very large degree.

G .Other sources of fluoride exposure. Although people are also potentially exposed to fluoride from fluoride in ambient air, fluoride dental treatments, and pharmaceuticals,among other things, OW concluded that these sources of exposure are insignificant compared to other sources of fluoride exposure. Accordingly, OPP is not including such exposures in its aggregate assessment. (Ref. 23 at 16).

NAMA response: If fluoride dental treatments are insignificant sources of
exposure, surely SF’s biannual use as a structural fumigant where food has
been largely removed and exposure is unintentional and indirect is an
insignificant source and can be excluded as a fluoride source as well for the
purpose of the Agency’s exposure assessment.

Second, this assessment does not take into account those people that
depend on private drinking water wells and not public drinking water
systems. Drinking water wells in certain portions of the United States can
have fluoride levels exceeding those used in the assessments discussed

NAMA response: We believe excluding what the Agency says are the two
largest contributors to aggregate fluoride exposure would result in a flawed
exposure assessment. We encourage the Agency to conduct a new exposure
assessment that includes all significant sources of fluoride, but excludes
insignificant sources including SF used as a structural fumigant.

Page 3442
Third, a significant portion of the U.S. population is exposed to fluoride in
water that is naturally-occurring rather than added for therapeutic purposes.

NAMA response: The naturally-occurring fluoride in drinking water
contributes, by far, the largest portion of the aggregate fluoride exposure.
Therefore, we believe public health – and the taxpayer – would be far better
served buy addressing that source, and not sources the Agency admits are

Page 3443
Second, and more importantly, the threat that fluoride poses to teeth and
bones is due to aggregate exposure to fluoride not the fluoride in food
resulting from use of sulfuryl fluoride when viewed in isolation. Use of
sulfuryl fluoride is responsible for a tiny fraction of aggregate fluoride
exposure. For example, for the most highly-exposed age groups in the
populations examined in the revised risk assessment, fluoride from sulfuryl
fluoride accounts for about 2 to 3% of aggregate fluoride exposure. Given
the aggregate level of fluoride exposure, termination of the use of sulfuryl
fluoride would not change the fact that aggregate fluoride levels would still
exceed the safe level for highly-exposed subpopulations

NAMA response: We agree with the Agency when it admits here that SF’s
contribution to aggregate exposure is tiny, although we think even that
overstates SF’s true contribution. We also agree that, for the most highly-exposed age group, terminating the uses of SF would do very little to change
their exposure profile.

Page 3443 a. Flour mills. Generally, flour mills and other food processing
facilities are fumigated two to three times per year to control insect
populations (the primary pests are the red flour beetle and the confused flour

NAMA response: Because of the MB phase-out and milling company goals to
reduce overall pesticide usage, it is now more common to fumigate one or
two times annually, with the option for a third depending on seasonal
conditions that encourage early and late development of pests that affect mill
sanitation levels.

Page 3443 a.i. To the extent facilities have an approved critical use or can
obtain methyl bromide from pre-phase-out inventories, mills will likely switch
to methyl bromide if sulfuryl fluoride uses are immediately eliminated.

NAMA response: We disagree that switching to methyl bromide will be a
likely outcome. Under the NAMA MB Critical Use Exemption, only NAMA
member companies can access that MB – non-NAMA members cannot. We
are aware of non-NAMA member companies that have been using SF for
years. Even if SF is made unavailable as a result of this Proposed Order they
would not be able to switch to MB.

Further as a result of the on-going MB phase-out, the amount of MB available
to NAMA members has been reduced about 90 percent from historical use
levels. Mills still using MB as a sanitation solution already have difficulty
finding adequate quantities of that fumigant – adding new users would force
some millers to scramble to adopt an alternative years ahead of the ultimate
Montreal Protocol schedule.

Finally, the pre-phase-out MB stocks are privately owned and the quantity is
unknown. The Agency has no authority over those stocks and cannot compel
the owner to make them available to former SF users. Mills forced to look for
non-SF sanitation solutions could not plan on having access to that MB.

Page 3444 ii. Non-chemical control. The leading non-chemical control option
for use in flour mills is temperature manipulation. Either heat or cold can be
used to destroy insect pests. Use of cooling to control pests in flour mills,
however, is unlikely because cold temperatures can damage electronic
equipment in production areas. Use of heat is a more likely option.
Temperatures of 120-130 degrees Fahrenheit will kill most stored-product
insect pests. Heat, however, would not be appropriate for mills principally
constructed of wood because heat at these levels will shrink, crack, and warp
wood. This can result in structural damage to the facility and may also render
the heat treatment ineffective due to leakage of heat from the facility.
Approximately 25% of the total number of flour mills in the United States fall
in this category. These tend to be the older and smaller mills and thus
probably represent less than 25% of mill capacity. Newer mills are generally
constructed primarily of concrete or similar materials which would be
appropriate for use with heat disinfestation techniques.

NAMA response: We agree that high heat treatments can cause structural
damage. However, NAMA disagrees that structural damage from high heat is
a phenomenon restricted to facilities of wood construction, and that such
damage would be limited to “warping and cracking” as stated in the Proposed

NAMA members report significant problems resulting from high heat
treatment of the entire spectrum of milling structures including relatively new
mills of concrete construction. Reported examples include internal walls
moving several inches during high heat treatments, buckling and cracking of
walls, and failure of the steel tie-ins that join tilt-up, pre-stressed concrete
walls to structural support members.

In addition to structural damage, in its analysis the Agency failed to consider
the effects of high heat on equipment, wiring, electrical motors, bearings,
computers and programmable controllers. Suppliers of common milling
equipment have confirmed that damage caused to their products by
exposure to high heat would not be covered by warranties as the products
were never designed to withstand such stress.

Further, insurance carriers will likely view negatively such actions that cause
stress or damage to the physical structure and other capital assets.

Even though the Agency theorizes a move to high heat treatment will impact
newer mills less than older mills, in actuality the machinery in new mills is
often more sensitive – not less – due to due to the extensive

Wooden sifter boxes predominate industry-wide


Computers and programmable controllers typical in all mills, regardless of
the mill’s construction material


Drive shafts, bearings, power drive belts and sprinkler heads all affected by
high temperatures

Drive belts and wood machinery housing

The Agency also failed to include in its analysis any mention of the potential
effects of high heat on finished products or on finished product packaging in
a warehouse or other storage area within the mill structure. Such as:

  • for bagged and packaged food in warehouses awaiting shipment, the
    effect of high heat on the package seals that are fastened with glue;
  • The effect on product integrity or the cost associated with failure of the
  • The impact on food safety and consumer attitudes about the integrity
    of food;
  • Insurance coverage for losses associated with damage to capital assets
    or losses associated with food product recall or liability for injury or

High heat cannot be used to treat packaged milled grain products. It will
damage the packaging materials, their ability to protect contents and the
quality of the food they contain. The drying influence of heat will cause
failure of paper, corrugated packaging, plastic stretch wrapping and labels,
and promote handling and quality problems like infestation and rancidity
throughout the supply chain.

Page 3444 2.a.ii. cont. Initially, use of heat will involve higher costs due to
capital investment in heaters and plant modifications. However, in the long
run, use of heat may be less costly than chemical pesticides.

NAMA response: We agree that the use of high heat treatments will increase
costs. We know this to be true. However, we strongly disagree that over time
the use of heat may actually save money versus chemical fumigants. Milling
is an industry that operates on extremely thin margins. Therefore, if the
Agency’s prediction were true, heat would already be the preferred and
dominant sanitation technology.

Subramanyam et al conducted head-to-head comparisons of sulfuryl fluoride,
methyl bromide and high heat at the 340,000ft3 Kansas State University
research and teaching flour mill in 2009 and 2010. Their work showed, after
three replications, the average cost of a high heat treatment to be $30,000
per treatment, not including the cost of heater rental or purchase. The cost
of fans necessary would exceed $50,000, which could be spread out over
several treatments.

Vendor estimates of the cost of a high heat treatment for an average size
mill (1.0 million ft3) exceed $280,000 including about $200,000 to purchase
heaters and fans. This is a very significant capital cost. There are many mills
that exceed 2.0 million ft3 and even 3.0 million ft3 with commensurate
increases in labor, heaters, fans, ductwork and fuel.

Note that none of these high heat treatment cost estimates include the cost
of labor to prepare the mill for treatment, nor the labor necessary to return
the mill to production.

Phosphine is a commonly-used food fumigant that could be used in some
portions of a flour mill; however, phosphine is highly corrosive to silver and
copper metals and their alloys and thus cannot be used in the production
areas of mills that contain electronic and electrical equipment which heavily
rely on these metals. In terms of total area, the portion of a mill devoted to
production is substantial and a failure to effectively dis-infest the production
area would quickly result in re-infestation of the entire facility. Thus,
phosphine is not an alternative to the use of sulfuryl fluoride. (Ref. 25 at 6-

NAMA response: Phosphine is an extremely important tool for addressing
infestation in stored grain, isolated storage structures and bins, and
packaged raw and processed commodities. We agree that phosphine is not a
structural fumigation alternative to SF for the reasons mentioned.
Additionally, researchers are already raising the issue of insects developing
resistance to phosphine.

For additional information see:
Zettler, J. L. 1997. Influence of resistance of future fumigation technology. In
Donahaye E. J., Navarro S., and Varnava A. (Eds). Proceedings of the
International Conference on Controlled Atmospheres and Fumigation in
Stored Products, 21-26 April 1996. Printco Ltd. Nicosia, Cyprus, pp. 445-454.

Zettler, J. L., W. R. Halliday, and F. H. Arthur. 1989. Phosphine resistance in
insects infesting stored peanuts in the southeastern United States. J. Econ.
Entomol. 82: 1508-1511.

Zettler, J. L., and G. W. Cuperus. 1990. Pesticide resistance in Tribolium
castaneum (Coleoptera: Tenebrionidae) and Rhyzopertha dominica
(Coleoptera: Bostrichidae) in wheat. J. Econ. Entomol. 83: 1677-1681.

In addition to the corrosivity and other problems with using it as a structural
fumigant, we would not want to increase the usage of phosphine in a way
that would contribute to accelerating insects’ development of resistance and
reducing phosphine’s effectiveness.

ii.Non-chemical control. The leading non-chemical control option for use in
flour mills is temperature manipulation. Either heat or cold can be used to
destroy insect pests. Use of cooling to control pests in flour mills, however, is
unlikely because cold temperatures can damage electronic equipment in
production areas.

NAMA response: Cold temperatures do slow down the growth of insects, and
for that (and other) reason farm and commercial grain storage bins are fitted
with aeration fans to cool the grain quickly after harvest. Ambient air
temperatures, particularly in the northern states, will also slow down and
prevent insect growth in winter months. But grain is routinely stored for
months prior to being shipped to mills, and insects will have either resumed
their growth or new insects have entered and re-infested the grain.

In any case, we know from experience that cold temperatures typically
attained in winter months are insufficient to make cold a fumigation
alternative. Further, fumigations typically occur in the late spring and early
fall when temperatures are ideal for insect growth.

Finally, we are unaware of any refrigeration system that exists capable of
lowering the ambient air inside a mill structure to an insecticidal
temperature. Developing such a system would likely take years and the cost
would no doubt be astronomical.

Use of heat is a more likely option. Temperatures of 120-130 degrees
Fahrenheit will kill most stored-product insect pests. Heat, however, would
not be appropriate for mills principally constructed of wood because heat at
these levels will shrink, crack, and warp wood. This can result in structural
damage to the facility and may also render the heat treatment ineffective
due to leakage of heat from the facility. Approximately 25% of the total
number of flour mills in the United States fall in this category. These tend to
be the older and smaller mills and thus probably represent less than 25% of
mill capacity.

NAMA response: We agree that high heat can kill pests and also that it
presents technical problems for mills with any wood construction or wood-
containing machinery. However, we disagree that the technical problems
would not be present in mills with no wood construction.

Mill with all-wood construction

Newer mills are generally constructed primarily of concrete or similar
materials which would be appropriate for use with heat disinfestation

NAMA response: We are not aware of the basis for the Agency’s distinction
between wood and concrete mills construction as the deciding factor in
choosing heat treatment. Relatively new NAMA member mills constructed of
non-wood materials have experienced significant technical problems
associated with high heat treatments including the phenomenon of concrete
walls moving several inches over the course of the treatment. Also, in mills
with tiled walls on the inside of the mill to eliminate insect harborages and to
facilitate more efficient cleaning, millers have witnessed the tiles falling off
the walls during and after the heat treatment.

Simply put, concrete mills were not designed nor constructed with the goal of
withstanding high heat treatments. It is possible that some may
coincidentally be able to withstand high heat, but many will not. We believe
insurance carriers may take a very negative view of subjecting mills,
regardless of their construction material, to high heat given the potential for
significant risk of failure.

For those mills that can conceivably tolerate high heat treatments, cost will
still be a significant consideration. We know of a mill, about 25 percent
smaller than average, for which conversion to high heat required a capital
investment exceeding $1.0 million. Unlike many nations, the U.S. does not
protect or subsidize its milling industry. And as the Agency’s own economic
analyses show, milling is an industry with very low profit margins. Therefore,
such capital expenditures – which add nothing to efficiency, output, quality
or safety – are simply out of reach for many millers.

New mill construction is often cost prohibitive. Therefore, when current
milling capacity is insufficient, the best economic choice is often to simply
expand existing structures. Portions of many US mills date from 100 years or
more, and have been added onto multiple times throughout the years. Even
mills that are relatively new construction by US standards – from the 1950s,
60s and 70s – were never constructed with high heat capabilities in mind.

Mill with multiple construction materials including concrete block, corrugated
steel panels and slip-form concrete.

Those additions and renovations often incorporated different construction
materials based on the optimal return on the investment at the time of the
mill construction and expansion. Each of these construction materials will
have its own unique physical response to heat. They will expand at different
rates and to different degrees upon heat-up, as well as upon cool-down.
Also, the varying insulating properties of the construction materials will
present technical challenges.

For example, a mill may be constructed of concrete or brick, but the attached
warehouse may be constructed of steel panels. The concrete and brick will
hold the heat far more efficiently than the steel panels. So during a high heat
treatment, thermodynamics dictate that heat will flow to the warehouse
where it is lost to the outside environment due to the poor insulating
properties of steel.

Mill with wood, steel, brick and concrete construction

Mill constructed of wood on top of stone block walls.

Mill with concrete, steel and brick construction

Mill constructed of wood and corrugated steel.

Initially, use of heat will involve higher costs due to capital investment in
heaters and plant modifications. However, in the long run, use of heat may
be less costly than chemical pesticides.

NAMA response: We agree that the use of high heat treatments will increase
costs. However, we strongly disagree that over time the use of heat may
actually save money versus chemical fumigants. Milling is an industry that
operates on extremely thin margins. Therefore, if the Agency’s prediction
were true, heat would already be the preferred and dominant sanitation

Switching to heat will also require transition time for the industry. Not only
will mills have to purchase (or rent) heaters but modifications may be
necessary to the mill to insure that heat is evenly distributed. Individual mills
will have to go through a trial and error process to determine how the
heating technique can be effective in each unique facility. Because
disinfestations are commonly needed only two to three times per year, mills
are likely to need an extended transition time to implement the technology
effectively. If chemical alternatives are not available during that timeframe,
processed food contaminated with insect parts and waste due to failure of
initial attempts at heat disinfestation will have to be destroyed. (Ref. 25 at

NAMA response: We agree that transition time will be necessary should the
Agency move forward with the tolerance revocation despite its own
conclusion that doing so will do little or nothing to reduce fluoride exposure in at-risk subpopulations. However, we disagree strongly with the three year
phase-out described in the Propose Order.

In the transition from MB to SF, fumigant using industries are already in the
seventh year of MB Critical Use Exemptions, and that is a transition from one
chemical method to another chemical method with basically identical
application procedures. Transitioning from a chemical method to a largely
unknown high heat method will require at least that long and perhaps longer.

iii. Product removal. A third option that combines chemical and non-chemical
control would be complete removal of all food from a facility before
fumigation with sulfuryl fluoride. Currently, the sulfuryl fluoride label requires
that food in facilities be minimized prior to fumigation. Only food that is not
practical to remove may remain during the fumigation. Removal of food is
also essential to the efficacy of sulfuryl fluoride. However, if all food is
removed such that use of sulfuryl fluoride would not result in fluoride
residues in food, no pesticide tolerance would be needed for this use and
aggregate exposure to fluoride would not be increased.

NAMA response: It is true that to optimize the effectiveness of any
treatment, chemical or non-chemical, small quantities of product in the
milling system must be removed to the greatest extent practicable. However,
this kind of cleaning should in no way be assumed to be capable of removing
100 percent of the product. Some intermediate or finished milled grain
product will always be present within the milling system. Thus, some minute
quantities of material will be inadvertently be exposed to the SF and any
residues resulting would be violative.

The following photos show product residue that is typical of all mills.

Page 3445 5. Harm to health. Other beetles have been associated with
gastrointestinal illness and discomfort. (Ref. 32 and 33). Contamination also
could include food-borne pathogens that cause disease, such as E. coli or
Salmonella, introduced by flies that would no longer be controlled by sulfuryl
fluoride. (Id.)

NAMA response: The presence of E. coli or Salmonella is extremely rare in
dry milled grain products, and we are unaware of any example of such an
occurrence as a result of being introduced by flies.

It is not clear why the Agency cites references 32 and 33 as support for the
statement that contamination with food-borne pathogens could be introduced
by flies. The first of the cited references, an FDA consumer update about an
infant formula recall, makes no mention whatsoever of foodborne pathogens.
The hazard cited in that consumer update is potential irritation of the
gastrointestinal tract in babies, leading to stomach upset or food refusal due to contamination with insect pieces or larvae. This is a physical hazard
apparently stemming from significant deficiencies in basic food safety
sanitation, which bears little relevance to the issue at hand. The second
cited reference, an article in FDA Consumer, makes no mention of
Salmonella, and E. coli is mentioned only in connection to cockroaches.
Cockroaches are extremely rare in a low moisture manufacturing
environment like flour milling.

Page 3446 If both sulfuryl fluoride and methyl bromide are unavailable, or
supplies are limited, there is likely to be some disruption of the food supply
as to the affected commodities and/or there is a greater likelihood of
contaminated food being released for public consumption.

NAMA response: The Agency cites no evidence to support its conclusion that
there would be a greater likelihood of contaminated food being released for
public consumption, and we are aware of no such evidence. The milling
industry takes very seriously its responsibility to produce safe milled grain
foods in a sanitary environment. Consumers expect it and both federal and
state law require it. There are multiple levels of quality control in place to
prevent and detect contamination. These requirements and controls are
designed to ensure that contaminated food presenting a risk of injury is not
distributed in commerce, and they are not contingent on the availability of
any single sanitation tool.

Although the unavailability of sulfuryl fluoride and methyl bromide could
require adjustments in quality control systems and will increase costs of
production, there is no basis to conclude that such unavailability would result
in a greater likelihood of contaminated food being released for public

If both sulfuryl fluoride and methyl bromide are unavailable, or supplies are
limited, there is likely to be some disruption of the food supply as to the
affected commodities and/or there is a greater likelihood of contaminated
food being released for public consumption. The extent of disruption and/or
contamination varies based on the type of processing facility and the
commodities involved. For newer flour mills and other food processing
facilities (i.e., ones made principally of concrete), use of heat should
eventually be a successful alternative to sulfuryl fluoride.

NAMA response: As described elsewhere the transition period for mills, even
newer ones of concrete construction, will be long and difficult. A critical
question is just how long it will take for heat to eventually be a successful
alternative. It has taken more than six years for SF to be adopted as even a
partial MB alternative using nearly identical and well-known procedures.
Logically it will take longer to transition to heat which is a very different and
less well-known treatment.

Older processing facilities constructed mainly of wood may have no options
other than to cease production unless Dow AgroSciences seeks and obtains a
registration amendment for sulfuryl fluoride that insures that sulfuryl fluoride
is used in a manner not resulting in residues in food.

NAMA response: The flour milling industry operates on a 24/7 basis. There is
little or no excess production capacity that could be tapped to make up for
mills forced to cease production as a result of the Agency’s Proposed Order.
Milling is a strategic asset in the nation’s food supply chain, and an Agency
action that will knowingly force some mills to close is unacceptable.

It is impossible to ensure that any fumigant can be used in a manner not
resulting in residues in food at some level. Milling managers have already
stated that they would not use SF without residue tolerances even if the label
were changed “in a manner not resulting in residues in food.” The liability to
the milling company would be considered too great in the event that residues
were detected in a product at any non-zero level, no matter how small or

Following are more photos showing typical product residues in milling
machinery and systems.

2. Sulfuryl fluoride’s contribution to fluoride exposure. Use of sulfuryl fluoride
results in a minimal contribution to fluoride exposure. Elimination of sulfuryl
fluoride does not solve, or even significantly decrease, the fluoride aggregate
exposure problems identified earlier.

NAMA response: Again, if the action in the Proposed Order will not solve, or
even significantly decrease the aggregate exposure then why is the Agency
proposing it? Why isn’t the Agency using its authority to apply a de minimis

2.Tolerances for commodities where there is little to no use of sulfuryl
fluoride: 90 days. EPA’s analysis and information from Dow AgroSciences
indicate that sulfuryl fluoride is not currently used in significant amounts, if
at all, on numerous commodities for which direct fumigation is allowed under
the sulfuryl fluoride registration. EPA is proposing a termination of tolerances
associated with these uses 90 days from the effective date of the order.
Ninety days should be sufficient for all affected parties to come into
compliance with the revised situation. Tolerances in this category are: barley,
bran, postharvest; barley, flour, postharvest; barley, grain, postharvest;
barley, pearled barley, postharvest; cattle, meat, dried; cheese; coconut,
postharvest; coffee, bean, green, postharvest; corn, field, flour, postharvest;
corn, field, grain, postharvest; corn, field, grits, postharvest; corn, field,
meal, postharvest; corn, pop, grain, postharvest; cotton, undelinted seed,
postharvest; ginger, postharvest; grain, aspirated fractions, postharvest;
grape, raisin, postharvest; herbs and spices group 19, postharvest; hog,
meat; millet, grain, postharvest; nut, pine, postharvest; nut, tree, Group 14,
postharvest (revised to cover only walnuts, postharvest); oat, flour,
postharvest; oat, grain, postharvest; oat, groat/rolled oats; peanut,
postharvest; pistachio, postharvest; sorghum, grain, postharvest; triticale,
grain, postharvest; vegetable, legume, group 6, postharvest; wheat, bran,
postharvest; wheat, flour, postharvest; wheat, germ, postharvest; wheat,
grain postharvest; wheat, milled byproducts, postharvest; wheat, shorts,

NAMA response: As described elsewhere, little or no raw grain is fumigated
with SF, so we would have no objection to a 90-day termination of residue
tolerances for raw grain.

However, some milled grain products may be fumigated with SF, even
unintentionally. For example, a mill that manufactures retail packages of
flour will have a warehouse attached to the mill. The warehouse may be
fumigated at the same time as the mill structure. If it became necessary for
the warehouse to be emptied prior to fumigation, this would add significantly
to the cost of operations.

Also, bins of flour within the mill structure may also receive incidental
fumigation. Here too, if all storages were required to be emptied the industry
standard Just-In-Time delivery capabilities would be damaged, as would the
efficiency of the vertically integrated supply chain.

4.Tolerances for commodities receiving residues from incidental treatment during structural fumigation—3 years. The situation for foods requiring tolerances as a result of incidental treatment from structural fumigations is more complicated.

NAMA response: We agree the situation is more complicated; very
complicated. For that reason, the three year transition for commodities
receiving residues from incidental treatment during structural fumigation
must be at least six years. It has taken longer than that to complete only a
partial transition from MB to SF.

Different types of facilities will face different hurdles in transitioning from
sulfuryl fluoride to other methods of pest control. For most facilities, use of
heat may prove an adequate pest control strategy. However, implementation
of heat technology is not expected to be seamless and the availability of
sulfuryl fluoride as a backup to avoid potential disruption or contamination is
important. OPP expects that, after the first year, use of sulfuryl fluoride in
these facilities will be the exception rather than the rule as the technology
comes online and facility operators gain experience with it. In other words,
sulfuryl fluoride would only be used when difficulties arise in perfecting the
use of heat technology in individual facilities. Given the cost of sulfuryl
fluoride treatment, facility operators, having invested in heat technology, will
have a strong incentive to avoid use of sulfuryl fluoride unless absolutely
necessary. A relatively short transition period may be appropriate for these

NAMA response: We are certain that millers will attempt to transition, but
that will take years to complete. However, we are equally certain that the
FDA, and consumers, will offer no flexibility in complying with food sanitation
regulations, nor would we want any. Our reputation for manufacturing
wholesome milled grain products in a sanitary environment is non-
negotiable. Therefore, we strongly believe that during this transition period
we must have access to SF as a backup should a high heat treatment result
in less than total control, which is almost certain. Thus, we strongly advocate
a minimum transition period of six years should the Agency move forward
with the Proposed Order.

Page 3448
Thus, to some degree, owners of wooden food handling facilities face the
most serious consequences of any producer group and, due to their relatively
large share of the market, there could be similarly serious consequences for
the public. For that reason, EPA is proposing termination of tolerances
associated with these uses 3 years from the effective date of the order. To
insure that this extended transition period will not encourage owners of
concrete facilities to maintain the status quo, EPA plans to pursue
registration modifications for sulfuryl fluoride that differentiate between
sulfuryl fluoride use in concrete and wooden structures. EPA’s goal would be
to allow sulfuryl fluoride use in concrete facilities for a period no longer than
necessary to accomplish the transition to heat technology.

NAMA response: Again, as stated earlier, the Agency’s attempt to distinguish
between facilities of wood construction and those of concrete is arbitrary and
not grounded in technical merit. Both types of facilities will struggle to
transition to high heat for the same reasons and for different reasons. The
machinery, motors, wiring and computers will be the virtually the same – or
identical – in both types of facilities.

The Agency has not described how much wood would be required for the
Agency to determine a mill is qualified for the longer transition period. Mills
vary in their percentage of wood construction. Therefore; all mills should be
granted the same transition period, and it should be a minimum of six years.


The docket on the Proposed Order includes a January 7, 2011 agency
memorandum “Assessment of Impacts on Flour Mill Operators of a Stay in
Sulfuryl Fluoride Food Tolerances.” There are important points in that
document that should be addressed also.

Page 6/22 Exclusion and Sanitation Sanitation is divided into two general
types: macro-sanitation and micro-sanitation.

NAMA response: We have not heard these terms before, although the
memorandum authors quote Heaps and Mason, both of whom are known to
the milling industry. But the description of so-called micro-sanitation leads us
to wonder if the Agency personnel didn’t misunderstand Heaps and Mason.
EPA discusses the practice of micro-sanitation in a mill with toothbrushes,
pipe cleaners and cotton swabs. The following photographs of typical mill
interiors should help make clear why such micro-sanitation is impractical and
of no help in either preparing a mill for fumigation or helping to eliminating
the need for fumigation altogether.

Mill A

Mill B

Mill C

Mill D

Mill E

Mill F

Mill G

Mill H

Mill I

Mill J

The Agency’s analysis fails to include any discussion of the costs associated
with the cleaning of a mill with “toothbrushes, pipe cleaners and cotton
swabs.” Consider that the volume of an average mill will easily exceed 1.0
million cubic feet. For perspective, that is equivalent to a box 205 feet long,
70 feet high and 70 feet wide, filled with seven floors of machinery, spouting,
motors, electrical conduit, wheat bins, flour bins, etc. Suffice it to say we
believe the Agency’s concept of micro-sanitation to be so lacking in real-
world practicality that it is rendered meaningless.


The Flour Mill Budget
It was assumed that a facility fumigates three times per year at a cost of
$19,211 per fumigation (Adams et al,. 2010).

NAMA response: We believe this cost estimate to be very low. Fumigating a
mid-sized mill will typically cost in the $35,000-40,000 range.

“This analysis uses the mid-range estimate that was based on average
chemical cost and average dose (40g/m3).”

NAMA response: We believe this dose to be low. At the beginning of
fumigation the initial concentration is usually 50g/m3 or greater, depending
on temperature and expected half-loss time.

Cost Categories and Estimates – Labor to move product
(Page 11) One estimate of labor needed to perform this task indicated that
10-15 people could remove the product for a mill with a capacity of 8,000
cwts per day in approximately 24 hours; five people could then replace it
after the fumigation and get the mill running again in about 12 hours (Ron
Galle, personal communication, July 2010)

NAMA response: We have confirmed with Mr. Galle that his statement refers
only to the labor necessary to remove product from warehouses and bulk
bins. It would require the same amount of labor and time to move products
back into storage. Further, his estimate does NOT include the people or time
to conduct the micro-sanitation referred to elsewhere in the Agency

The Agency document also fails to include consideration that it is not possible
to have a zero inventory at the time of fumigation. Mills must have product
on hand immediately after fumigation in order to satisfy customers. The only
alternative would be to have product shipped from other locations during the
fumigation and immediately afterward, but the Agency does not include this
cost in its analysis.

Expected Transition Time for Scenario #1
(Page 13) To transition from chemical fumigation with food contact to
fumigation without food contact, a flour mill will need time to plan and
practice making the necessary personnel and inventory adjustments.

NAMA response: The Agency acknowledges in the Propose Order that the
current SF label does not allow for SF fumigation under this usage scenario.
Further, even if the registrant revises the label to allow for it, we have no
way of knowing if it would revise the label and therefore cannot plan on it.

However, even if the label is revised we believe few, if any, mills will adopt
this strategy because of the potential liability associated with potential
residues, however small. Millers have spent many years developing
allegiance among their customers and they will not risk that valuable
relationship, ever. There would be enormous indirect costs associated with
lost markets and brand damage. Additionally, there are substantial direct
costs associated with executing a recall.

Heat disinfestation of a flour mill or food processing facility entails heating
the structure to approximately 55°C (130°F) over a 6-8 hour period of time,
holding that temperature for 24 hours, and then allowing it to cool over a
period of 12 hours.

NAMA response: This statement vastly understates the time required to
conduct a safe and effective high heat treatment. If the heat treatment
results in a temperature change (!t)) that is too fast on the heat up, or too
fast on the cool down, the structural integrity of the building can be
dangerously compromised. Our experience is that it takes up to 24 hours to
bring the temperature up to the desired level. For example, if the target
temperature is 130°F, millers will commonly heat the structure up to 110°F
fairly quickly, but after that will slow the heat-up to 2-3 degrees per hour in
order to minimize physical damage to the structure as well as drive heat into
the walls, floors, ceilings and equipment for improved effectiveness. Thus,
the entire heat-up can take up to 24 hours.

As described earlier, mill downtime results in lost revenues that are very
expensive. Thus, the Agency’s economic analysis greatly understates the true
cost of a high heat treatment.

On page 23 of the BEAD document, the authors write “If there is an
immediate revocation of sulfuryl fluoride tolerances, there is a possibility that
contaminated food could enter the food chain and result in human illness.”

NAMA response: Millers take very seriously their obligation and commitment
to produce to continue producing safe and wholesome foods. That will never
change, regardless of the tools and techniques the government allows us to
use. It is true that the cost of ensuring food safety and wholesomeness
would increase during a time of economic stress for millers as well as

The North American Millers’ Association believes a thorough analysis and
consideration of these comments will lead the Agency to conclude a
revocation of SF residue tolerances is not warranted, nor would a revocation
confer health benefits for the nation’s consumers. We appreciate the
opportunity to offer these comments and any consideration you may afford


James A. Bair
Vice President

Print Friendly, PDF & Email

1400 Crystal Drive, Suite 650      Arlington, VA 22202      TEL: 202.484.2200      FAX: 202.488.7416
Non-Discrimination Statement | Privacy Policy | Terms of Use | Login