To:	Ken Feith

From:	Elliott H. Berger and Brian Myers

Date:	January 24, 2007

Re:	Observations regarding key issues reviewed at EPA’s informal HPD
meeting

The following captures our understanding of, and opinions regarding, the
key issues discussed at the recent informal HPD meeting held by EPA in
Ariel Rios North on January 24, 2007.

Although we don’t have a preference for Method A or B, we do strongly
feel that only one Method be selected and required by the EPA.  It will
be too time consuming and too costly to require that companies test by
both methods.  Likewise, we prefer that the choice be from one of the
ANSI S12.6 procedures, either A or B, with reasonable modifications from
EPA, and / or the upcoming efforts of S12/WG11 to revise S12.6.

With respect to the A vs. B decision, perhaps the most important
consideration is that the labeled value, or if two values or a range are
chosen, the lower of the labeled values,  can be applied to occupational
HCPs without the need for derating by OSHA or otherwise.

We support the concept that retesting for a product can occur in the
same or a different lab at the manufacturer’s option and that
compliance audit testing take place in the same lab with an EPA witness
if possible.  As for the frequency of retesting (in the absence of any
suspicion that the product has changed or has been mislabeled), we
suggest that 7 years is the appropriate time frame that provides the
best compromise between concern for product changes and recognition of
the cost and time required for retesting.

The primary label for the packaging should include in the upper panel
the words Noise Reduction Statistic (alternately the Noise Reduction
Range) and in the upper part of the top section contain a horizontal bar
that extends from the 80 percentile at the left end to the 20th
percentile at the right end.  Indicate that this is the NRS or NRR in
decibels and that the left-end lower number indicates the value
“possible for most users to achieve” (if Method-B data are used), or
the value “possible for most individually trained users to achieve”
(for Method-A data).  The right-end upper value should be labeled as
“possible for a few proficient users to achieve,” and these words
are suitable for either Method-A or B data.

The middle part of the panel should use the following wording from the
Berger/Gauger paper:

For existing products the range of Noise Reduction Statistics is about 0
- 30 for the Low Value, and 20 - 40 for the High Value.  Larger values,
and a smaller range between the Low and High Value, denote greater
effectiveness and reliability.  Contact the manufacturer or see the
Secondary Label for additional guidance.

The preceding applies to most HPDs.  However, for ANR devices
specifically targeted towards aviation environments a suitable graphic
should replace the bar and the user will be directed towards the
accompanying information for guidance.  Likewise an icon for impulsive
noise/gunfire should be used in place of the bar also with guidance to
look elsewhere for information for level-dependent devices.

The regulation should be more clear about the fact that for products
like the E(A(R Classic® that is sold in a pillow pack, but normally
dispensed from a large box of 200 or more pairs, that the primary label
need not appear on the individual package, since in fact the package is
smaller than the label.  This is inferred in the current document, but
not explicit.

With respect to the secondary and perhaps tertiary labels that will
include instructions, how to use the NRR, and other ancillary
information, these label(s) should NOT be required to be on the box or
even in the box.  We don’t have room on the box now because of the
many languages we have to incorporate, and putting it in the box is
costly and a waste of paper.  The regulation should require that the
information be readily available by calling an 800 number for a fax, or
visiting a web site, and the labeling may also include reference to the
EPA/NIOSH repository web site (see next item).

We support the idea of a NIOSH-developed NRR calculator into which
manufacturers can upload their data to create a Hearing Protector Data
Sheet (HPDS).  We would like to work with NIOSH to make sure that the
categories selected to distinguish types of hearing protectors are the
most useful to our customers.

EPA should build flexibility into the regulation so that it can easily
be updated to reflect any changes that occur in the future in the
relevant ANSI standards that are cited.

The test procedures for ANR muffs should provide the user only the total
attenuation.  It is not needed to provide the passive and active
components separately to the end user, though those are required by the
lab in the intermediate computations.  The procedure that should be
cited for the combination of MIRE and REAT test data is the footnote
that is included in the draft S12.68 standard that discusses this issue.

During the meeting, Berger proposed the following summary of the pros
and cons of Methods A and B, and it generally received a favorable
response.

Overview of preferred Method with respect to eleven different
considerations

Issue	Method A	Method B	Either

Within-lab variability is smaller	(



Between-lab variability is smaller

(

	Cost of testing is less	(



Speed with which tests can be conducted	(



Provides realistic estimate of magnitude of field attenuation

(

	Provides best estimate of rank ordering of field attenuation

(

	No derating needed

(

	“Justifiable” and/or understandable to typical person	(



Can be used with dual number-rating system

	(

Provides best estimate of device’s inherent performance	(



Compatible with international standards and regulations

	(



  FILENAME  \* Lower \p  \* MERGEFORMAT  c:\...\_word
files\regulatory\epa jan 07 workshop followup.doc 

Observations regarding informal HPD meeting		Page -   PAGE  2 

