Potential Small Entity Representative (SER)

Response to Initial Outreach Document:

Exhaust and Evaporative Emissions Control from Small SI Engines and
Equipment and Marine SI Engines and Vessels

Geoffrey T. Ward, Ph.D.

Engineering Manager

Agri-Industrial Plastics Company

Fairfield, Iowa

July 25, 2006

Executive Summary

Agri-Industrial Plastics Company (AIP) is a small custom blow molder,
who supplies fuel tanks for both the Small SI Engine and Marine SI
Engine markets.  Thus, the areas of concern for AIP are based on the
proposed standard for permeation from fuel tanks.  In response to the
EPA permeation standard for Recreation Vehicles, AIP purchased a
multi-layer fuel tank machine, and as such has good experience regarding
cost and lead time requirements for meeting such a standard.

The primary issue that will affect AIP is the proposed implementation
date of MY2009 for both engine categories.  While the technology exists
to meet this implementation date, the capacity does not.  We would
propose that a date of MY2011 would be a more realistic implementation
date, given an estimated publication date for the finalized rule of the
end of CY2007.

The proposed gap of two years between implementation of general fuel
tanks and rotational molded fuel tanks is considered to be unacceptable
by AIP.  This presents a potential loss of business issue, in that
existing customers using blow molded fuel tanks will most certainly
consider redesigning for rotational molding, simply because of the
benefit of delaying compliance for two years.

The estimated cost of implementation as presented is considered to be
extremely low, even by conservative standards.  Our experience to date
suggests that a typical monolayer fuel tank in these markets, currently
in the $10-$20 price range, will likely increase by about 40-50% when
moving to a multi-layer construction.

Additional details regarding these issues will be provided as responses
to the detailed questions presented in the Outreach Document.

Detailed Responses to Questions for Potential SERs

1A – How useful would each of the small flexibility ideas discussed be
for your business in meeting the applicable standards?

Broadening the definition of engine families for certification

This would be very useful in terms of reducing the number of different
sample fuel tanks required to be provided for certification testing.  We
would propose that for evaporative emission, families could be defined
by the technology used for certification.  For example, all multi-layer
fuel tanks manufactured at a specific supplier for a given OEM would be
considered a family.  Thus, a single tank could be tested, with the
emission level applied across the entire family.  This would minimize
the amount of available production capacity that would be used up to
produce sample tanks for testing.

Minimizing compliance testing requirements

Again, this would be very useful, based on the capacity reasons noted
above.

Design-based certification

This would also be very useful.  While the current Recreational Vehicle
standard allows for design-based certification of steel fuel tanks, the
CARB standard extends this to co-extruded multi-layer plastic fuel
tanks.  Given the extensive use of this proven technology in automobiles
at a much lower evaporative emissions limit, we would suggest that
substantial data exists to justify a design-based standard for this
technology.  This would further decrease the demand against the limited
production capacity that is available.  As an option, an OEM could still
choose to test a fuel tank in the multi-layer family for the potential
purpose of defining a Family Emission Limit to be used in an averaging
calculation with other technology families.

Use of emission credits

This would be extremely useful, in fact almost critical when considering
the options for providing service fuel tanks for obsolete products.  The
low number of service tank requirements makes it impractical to break
into normal production on a limited capacity multi-layer machine to
provide service parts.  The ability to apply credits from active
production years would allow for different options to provide these
service parts at a higher emission limit.

Delay of the implementation date of the standards

This is not only useful, it will be necessary if AIP is to be able to
continue to provide fuel tanks to these markets.  This also extends
beyond AIP, as it is questionable as to whether the necessary capacity
exists anywhere in the industry.  This point will be expanded in the
next question.

Hardship provisions

This is not likely to affect AIP, as this appears to be aimed primarily
to target small OEMs.  We do have some small customers with low EAU
products that we cannot serve with the multi-layer machine.  If these
companies were granted hardship provisions that would enable them to
continue to take monolayer product, this may prevent or delay a loss of
business with these customers.

Limited temporary exemptions for small boat builders

This would not affect AIP.

2A – Can you comment on the length of a potential delay option?

Based on our experience in working to comply with the Recreational
Vehicle standard, it was necessary to order a multi-layer machine in
May, 2004 in order to meet a MY2008 implementation date.  Based on
current orders to meet this standard, as well as the CARB Small SI
Engine standard, the capacity on this machine is full.  There is
currently not enough of this type of machine to meet the full capacity
requirements for these existing standards.  In addition, many of the
machines that are available are too large to produce the smaller fuel
tanks in the Small SI Engine market.

The current lead time for a new machine is at least 12 months for
delivery.  With machine assembly and break-in period, this lead time
increases to nearly 18 months, after which time sample tanks for
certification and qualification will need to be produced.  Given the
testing time, we estimate that a new machine would be ready for
production approximately 2 years after placing the order.

It would not be prudent to order a machine to meet a standard until that
standard is finalized.  With an estimated publication date for the final
rule approaching the end of CY2007, this leads to a production start-up
date at the end of CY2009.  For some models, this would be sufficient to
meet MY2010 production.  However, some OEMs begin their model year much
earlier, thus a practical implementation date would be MY2011 for Small
SI Engine fuel tanks.

Regarding Marine SI fuel tanks, the numbers published in the
supplementary reports show a range of approximately 80,000-90,000 PWC
vehicles sold domestically per year.  Each of the fuel tanks in this
market is currently blow molded.  It would take this entire market worth
of fuel tanks to come close to filling a multi-layer machine, although
there would still be some remaining capacity.  It is not a practical
business case to purchase a dedicated machine of this type in
anticipation of serving an entire market segment.  The timeline given
above would suggest that capacity could be available by MY2011, however
the lack of a business case suggests that it would be risky to assume
that this capacity will be purchased solely for that purpose.  Thus,
machine capacity currently earmarked for other markets will need to be
utilized, or different technologies will likely need to be employed.

Other technologies, such as thermoforming, may be able to assist in the
capacity, however not many of these machines exist, either.  In
addition, this would result in a loss of business for AIP if our current
customers were forced to find an alternate technology.

3A – Do your answers to these questions differ based on the level of
the standard and/or timing of the program implementation date?

They may.  If the standard was relaxed initially with time for
multi-layer capacity to become available, it may be possible to
introduce a different technology, such as fluorination with some level
of confidence.  Even so, the shear numbers of fuel tanks in the Small SI
Engine market may overwhelm the ability of fluorination capacity to
become available by MY2009 production.  Even this may require a delay to
at least MY2010.

4A – Do you have suggestions for other potential small entity
flexibility options, or any additional suggestions on the flexibility
options discussed?

We would strongly endorse an averaging program to allow OEMs to take
advantage of low emissions levels from a strong technology such as
multi-layer construction, and average these results with other tanks
made with a barrier that may not meet the standard.  This would benefit
AIP from the sense that we could continue to provide low EAU tanks for
our existing customers without using up valuable multi-layer machine
capacity that is required to meet their higher volume orders. 
Understanding that this averaging approach is included in the
Recreational Vehicle standard, this may already be under consideration
for the proposed standard.

1B – 3B:  Level of Exhaust Standards questions -  Not applicable to
AIP

1C – Can you comment on the evaporative emission standards under
consideration for Small SI engines and equipment or Marine SI engines
and vessels?

From a technology standpoint, the technology certainly exists to meet
this level of standard for currently blow molded tanks.  However, the
primary concern is that the capacity does not exist to fulfill the
entire market, nor will it exist in time to meet an implementation date
of MY2009, as noted above.  While the CARB standard for the Small SI
engines begins in MY2008, many of our customers are not planning on
transitioning their entire fleet of engines to comply with this standard
at that time.  Thus, it is not a matter of recertifying existing barrier
fuel tanks to a different standard.

2C – Specifically, how would this affect your businesses?

Based on our research in response to the Recreational Vehicle standard,
the most reliable and cost effective way to meet this proposed
evaporative standard is to begin supplying multi-layer co-extruded fuel
tanks.  This would require at least one additional multi-layer blow
molding machine at a large capital expense, with a long lead time before
a payback period would begin.  This machine could likely be justified
for the Small SI Engine market, however as noted above, the business
case is much more difficult to justify for the Marine SI market.

For the Marine SI market, the basic options for AIP would be:  a)
Purchase yet another machine, anticipating that enough additional
business from the Recreational Vehicle and/or Small SI Engine markets
would fill the excess capacity, b) Attempt to fit as much of the
existing Marine SI business into the current machine and the potentially
purchased Small SI machine, or c) Inform our existing Marine SI
customers that we will not be able to supply multi-layer tanks for this
market.  The first option comes with a very high financial risk for a
small company.  The second option comes with a risk of overselling the
available capacity or a potential loss of business for the products that
would not fit in the existing capacity.  The last option would result in
a potential loss of business situation, if we could not provide them
with monolayer tanks that were subsequently treated in some manner. 
Thus, all likely scenarios resulting from a Marine SI Engine standard
contain a potential level of risk that would normally be considered
unacceptable from a business standpoint.

3C – Would the new standards require changes to your engine,
equipment, vessel, or fuel system components?

As a fuel tank supplier, the direct answer from our perspective is no. 
However, it should be noted that for much of the Small SI Engine market,
and possibly all of the PWC market, current fuel tanks are supplied in a
natural resin color to allow for translucency.  Most, if not all, of the
viable barrier solutions for blow molded fuel tanks will result in
opaque fuel tanks, which in many cases will result in the need for some
type of fuel level gauge or sensor.  Many OEMs are either not aware of
this or are fighting it.

1D – What do you estimate would be the per unit variable and/or fixed
costs for your businesses to meet the standards EPA is considering?

Raw material cost increase from a monolayer HDPE resin to a multi-layer
recipe is currently estimated to be in the $0.25 - $0.40 range on a per
pound basis, depending on the HDPE resin price at the time and the
acceptable regrind contribution.  In addition to the base material
price, variable scrap costs may also incur based on the amount of excess
material required as a function of tank design.

Given the above material cost in addition to increased overhead rates
for multi-layer equipment, our experience in the Recreational Vehicle
market suggests a per unit sell price will increase in the range of
40-50% over the cost of an equivalent monolayer fuel tank.  Thus,
current fuel tanks selling in the $10-$20 range will increase to $15-$30
on the average.  This price will increase further if some type of fuel
gauge or sensor will need to be added to the tank design to move from a
natural color to a black multi-layer tank.

In addition to these direct costs, it is important to consider that with
a single machine available to meet all demands, typical lean
manufacturing principles designed to minimize run lengths and increase
inventory turns do not apply very well.  In fact, we anticipate the need
to potentially hold finished inventory for up to several months for low
EAU applications.

Our cost studies conducted when researching 5-15 gallon fuel tanks in
the Recreational Vehicle market showed that the markup for a multi-layer
fuel tank was at the low end when compared to other viable options.  As
some Small SI fuel tanks are smaller than this, other options may become
more cost competitive as the more costly monolayer materials make up
less of the total cost, and the batch process of fluorination can treat
more tanks at the same price.

The cost of a blow mold designed for a multi-layer fuel tank may add up
to a few thousand dollars to add certain recommended design features to
ensure integrity of the fuel tanks made with this process.  In addition,
a common approach for molding fuel tanks is to utilize an injection
molded fill neck and/or other components.  These components are commonly
insert-molded into the blow mold and captured during the molding
process.  Due to differences with the multi-layer extrusion process, for
certain applications it may be necessary to add a secondary welding
operation to the overall process.  The tooling cost for welding these
features can easily reach up to tens of thousands of dollars, depending
on the application and welding approach employed.

In the case of fluorinated fuel tanks, the logistical costs associated
with additional shipping, and multiple handling of the tanks, as well as
the transfer of assembly operations that may otherwise be completed
in-line during the molding cycle have not yet been estimated by AIP. 
These costs have been cited by many customers when concluding that the
overall cost of fluorination is not competitive with the multi-layer
option, despite a lower price estimate when only considering the direct
cost of fluorination added to the price of a monolayer tank.

Much of the R&D costs at AIP have already been absorbed during the
process of evaluating different options to meet the Recreational Vehicle
standard.  However, as recent advancements in monolayer barrier
solutions become increasingly viable, additional R&D expenses will be
required to evaluate these options in order to provide the necessary
flexibility to meet the needs of the entire industry.  The level of R&D
costs associated with this is difficult to predict at this time.  In
addition, with the current certification testing requirement in the
proposed standard, the cost to provide each set of sample fuel tanks for
testing can reach up to $5000.

Additional confidential details outlining the potential per unit price
increase are provided in the Appendix of this response report.

2D – If applicable, what types of hardware, capital equipment and/or
operational changes would you potentially require?

The capital cost for a new multi-layer machine of the size required for
much of the Small SI Engine fuel tanks currently produced at AIP is
likely to approach $3 million, including auxiliary equipment to support
the production line.  For a larger machine to support the PWC market,
this would increase to about $4 million.  This capital cost is magnified
when considering the cost of money required to purchase a machine that
would not produce production fuel tanks for up two years after placing
the order.  In addition, it may become necessary to purchase secondary
equipment to support any welding operations that are not currently
performed, as discussed above.

3D – How long would these operational changes take (i.e. from initial
planning to construction and start-up)?

As stated above, we anticipate a lead time to production of up to two
years.  Barring any reason to purchase additional equipment based on
existing demand, this timeline would not begin until the final rule is
published.

4D – How useful would delays for small businesses be and how long
would such delays need to be to be beneficial?

These delays would be very useful, as discussed earlier.

5D – Would a phase-in, perhaps based on percent of total per unit
sales, be helpful?

Yes, extremely helpful.

6D – Based on your experience with earlier EPA standards, does the
expectation of manufacturers to pass on the added costs to consumers
seem reasonable?

Whether it is reasonable or not, this will be a necessary condition for
us to be able to supply fuel tanks.  We simply cannot afford to absorb
these costs.

[The appendix was marked confidential and has been deleted from this
document.  This appendix was titled “Detailed Cost Assumptions.”]

