[Federal Register Volume 62, Number 156 (Wednesday, August 13, 1997)]
[Notices]
[Pages 43416-43425]
From the Federal Register Online via the Government Publishing Office [www.gpo.gov]
[FR Doc No: 97-21331]


-----------------------------------------------------------------------

DEPARTMENT OF TRANSPORTATION

National Highway Traffic Safety Administration
[Docket No. 94-068; Notice 2]


Highway Safety Programs; Model Specifications for Calibrating 
Units for Breath Alcohol Testers; Conforming Products List of 
Calibrating Units

AGENCY: National Highway Traffic Safety Administration (NHTSA), DOT.

ACTION: Notice; request for comments.

-----------------------------------------------------------------------

SUMMARY: This notice amends the Model Specifications for Calibrating 
Units for Breath Alcohol Testers by incorporating an alternative 
testing procedure using National Institute for Standards and Technology 
(NIST) Reference Gas Mixtures (RGMs) for the evaluation of dry gaseous 
ethanol calibrating devices and making other changes that were 
previously proposed to simplify the Model Specifications and to make 
them easier to read. This notice also proposes and seeks comment on a 
new alternate procedure for evaluating the accuracy of both wet bath 
and dry gas breath alcohol calibrating units using infra-red 
spectroscopy. Published with this notice is an amended Conforming 
Products List (CPL) of calibrating units that meet the Model 
Specifications. This amended list includes five new listings, one wet 
bath unit and four dry gas units.

DATES: The amendments to the Model Specifications and the issuance of 
the Conforming Products List of calibrating units meeting the Model 
Specifications become effective on August 13, 1997. Comments on the 
alternate testing procedure using infra-red spectroscopy proposed as an 
amendment to the Model Specifications published herein must be received 
by October 14, 1997.

ADDRESSES: Comments regarding the alternate testing procedure should 
refer to the docket number and the number of this notice and be 
submitted (preferably in ten copies) to the NHTSA Docket Section, Rm. 
5109, 400 Seventh St., S.W. Washington, D.C. 20590 (Docket hours are 
from 9:30 a.m. to 4 p.m.).

FOR FURTHER INFORMATION CONTACT: Dr. James F. Frank, Impaired Driving 
Division, Office of Traffic Injury Control Programs (OTICP), NTS-11, 
400 Seventh St., SW, Washington, DC 20590. Telephone (202) 366-5593.

SUPPLEMENTAL INFORMATION: On August 18, 1975 (40 FR 36167), NHTSA 
published a standard for Calibrating Units for Breath Alcohol Testers. 
A Qualified Products List of calibrating units for breath alcohol 
testers, of devices which met the standard, was

[[Page 43417]]

first issued on November 30, 1976 (41 FR 53389).
    On December 14, 1984, NHTSA issued a notice to convert the 
mandatory standards for evidential breath testers and calibrating units 
for breath alcohol testers to Model Specifications for such devices (49 
FR 48855 and 49 FR 48865, respectively) and to establish a Conforming 
Products List (CPL) of evidential breath testers and calibrating units 
meeting the Model Specifications. Amendments to the CPL have been 
published in the Federal Register since that time. Evidential breath 
testers are instruments that measure the alcohol content of deep lung 
breath samples with sufficient accuracy for evidential purposes. 
Calibrating units provide known concentrations of ethanol vapor for the 
calibration or calibration checks of instruments which measure breath 
alcohol.
    NHTSA published a notice in the Federal Register (59 FR 67377) on 
December 29, 1994, amending the Model Specifications for calibrating 
units for breath alcohol testers and updating the CPL for calibrating 
units. The notice also proposed and sought comments about providing an 
alternate testing procedure for evaluating the accuracy and precision 
of dry-gas ethanol calibrating units.
    Officials who use breath alcohol testers must verify their accuracy 
at appropriate intervals during use. The traditional means for ensuring 
accuracy has been by checking the breath tester calibration by use of a 
``wet bath'' calibrator, a device which provides moist alcohol in air 
samples at accurately known concentrations. Dry gas calibrating units 
have become available as an alternate means for calibration checking.
    A dry gas calibrator produces alcohol-in-inert gas samples (e.g., 
nitrogen or argon) at accurately known concentrations from a compressed 
gas cylinder. Dry gas calibrators, like wet bath calibrators, can be 
used to calibrate certain types of breath testers, but an evaluation of 
their precision and accuracy requires alternate procedures. Today's 
notice amends the Model Specifications for Calibrating Units for Breath 
Alcohol Testers by incorporating an alternative testing procedure using 
National Institute for Standards and Technology Reference Gas Mixtures 
for the evaluation of dry gaseous ethanol calibrating devices and 
making other changes that were previously proposed to simplify the 
Model Specifications and to make them easier to read. Additional minor 
changes were made to ensure accuracy and improve clarity of the 
document. Also, the term BrAC has replaced the term BAC throughout the 
model specifications to ensure consistency with usage recommended in 
the Uniform Vehicle Code.
    Today's notice also proposes an additional new alternate procedure 
for evaluating wet bath and dry gas calibrating units using infra-red 
spectroscopy. The agency believes that use of infra-red spectroscopy 
will offer several important advantages in the evaluation of both wet 
bath and dry gas calibrating units. Comments are sought regarding the 
agency's proposal.

A. Comments Received

1. Overview

    The agency received two comments in response to the notice of 
December 29, 1994: one from Scott Specialty Gas Co. (Scott Gas), a 
manufacturer of a dry gas calibrating unit, and one from U.S. Alcohol 
Testing (USAT), a manufacturer of an evidential breath test device and 
a wet bath calibrating unit that is currently listed on the NHTSA CPL.
    Scott Gas was generally supportive of the proposed revisions to the 
Model Specifications. USAT stated that it would favor the use of dry 
gaseous ethanol calibrating devices when ``it has been adequately 
demonstrated that dry-EtOH [calibration units] give results comparable 
to those obtained with conventional wet bath simulator calibration 
units.''
    Neither of the respondents specifically commented on the proposed 
revisions to simplify the Model Specifications. As stated in the 
notice, these proposed revisions did ``not represent substantive 
alterations in the procedures followed or in the criteria used to 
determine whether devices meet these model specifications.'' The 
proposed revisions have been adopted without change.
    Both Scott Gas and USAT raised questions in their comments about 
those aspects of the Model Specifications relating to the proposed new 
alternate testing procedure for evaluating the accuracy and precision 
of dry gas calibrating units. The comments addressed a number of key 
issues, including the comparability of wet bath and dry gas calibrating 
units and certain specific conditions affecting dry gas calibrating 
units. The issues that were contained in the comments are summarized 
and discussed below.

2. Comparability Between Wet Bath and Dry Gas Calibrating Units

    USAT commented that ``[T]he use of a dry gas EtOH standard makes no 
physical sense until it can be demonstrated that the presence of water 
vapor in the breath samples analyzed has no effect on the analytical 
outcome on the ethanol concentration of the breath samples analyzed by 
the [evidential breath tester].''
    While it is true that dry gas and human breath differ in moisture 
content, NHTSA has found no reason to exclude the use of dry gas 
calibrating units solely on this basis. If a calibrating unit (either 
wet bath or dry gas) meets the precision and accuracy criteria of the 
Model Specifications, the calibrating unit should be considered 
acceptable for general use.
    Independent research has confirmed the comparability of dry gas and 
wet bath calibrating units and the accuracy of dry gas calibrating 
units. Kurt M. Dubowski and Natalie A. Essary studied the performance 
of dry gas calibrating units and concluded that ``dry gas vapor-alcohol 
control [VAC] samples conformed to established formal specifications 
and * * * compared favorably with simulator effluents for control tests 
of breath alcohol analyzers which are capable of adjusting VAC results 
for ambient atmospheric pressure.'' 1 Lance D. Silverman, et 
al. reported on the comparability of wet bath and dry gas calibrating 
units. These researchers determined that there was substantial 
equivalence between both types of calibrating units. Their data ``based 
on collection of ethanol in an impinger and titration using a modified 
California Department of Health method * * * confirm[ed] the alcohol 
content of EBS compressed gases standards by an absolute, wet chemical 
method.'' 2
---------------------------------------------------------------------------

    \1\ Dubowski, K. and N.A. Essary ``Vapor-Alcohol Control Tests 
with Compressed Ethanol-Gas Mixtures: Scientific Basis and Actual 
Performance.'' Journal of Analytical Toxicology (1996)20, 484.
    \2\ Silverman, L.D., Wong, K. and Miller, S. ``Confirmation of 
Ethanol Compressed-Gas Standard Concentrations by a NIST-traceable, 
absolute chemical method and comparison to wet breath alcohol 
simulators.'' Accepted for Publication in the Journal of Analytical 
Toxicology, 1997.
---------------------------------------------------------------------------

3. Should the Model Specifications Be Expanded To Address Unique 
Conditions Affecting Dry Gas Calibrating Units?

    USAT challenged the use of dry gas calibrating units based on the 
following factors: (a) condensation in the cylinder as a consequence of 
low temperatures during shipment; (b) the need to make corrections due 
to changes in atmospheric pressure; and (c) the performance of dry gas 
calibrators over a range of temperatures and concentrations.
    NHTSA has considered these comments carefully and has concluded 
that dry gas calibrating units are suitable for evaluation according to 
the Model

[[Page 43418]]

Specifications and believes that the Model Specifications are 
sufficient to ensure the accuracy and precision of dry gas calibrating 
units. However, in light of the concerns raised by USAT, the agency has 
amended the procedures for submitting a product for certification. When 
a manufacturer submits a product to the agency for testing, it now must 
submit also a set of the instructions that are provided to end users. 
The instructions must sufficiently describe the procedures to be 
followed to protect against condensation in dry gas cylinders that 
might occur as a result of freezing during shipment and to correct for 
atmospheric pressure.
(a) Condensation in Dry Gas Cylinders as the Result of Freezing
    USAT commented that dry gas calibrating units were previously shown 
to have a ``memory effect when transported or stored at temperatures 
somewhat below room temperature.'' NHTSA acknowledges that dry gas 
calibrators could freeze during shipment and this could affect test 
results. As a result of freezing, alcohol could condense in the inside 
surface of the cylinder. If this were to happen, re-equilibration of 
the alcohol with the nitrogen after warming to room temperature could 
take a long time. It is possible that the gas in such cylinders might 
be used before re-equilibration occurred with the result that samples 
would be obtained at incorrect concentrations.
    Manufacturers of dry gas calibrating units recommend that, after 
receiving the dry gas cylinders, users should warm the cylinders to 
room temperature, then lay them down on a flat surface and physically 
roll them back and forth for a period of ten minutes to ensure 
equilibration of the contents. To test whether this procedure would 
ensure that the dry gas calibrators remained accurate, several 
cylinders of Lion Laboratories AlcoCal dry gas calibrators were placed 
in the freezer compartment of a refrigerator overnight at a temperature 
of -15 deg.C, then taken out of the freezer, warmed to room temperature 
and rolled on a table top for ten minutes. Data was collected 
confirming that tanks that were rolled after freezing gave accurate 
results.3
---------------------------------------------------------------------------

    \3\ Flores, Arthur, ``Dry Gas Calibration Units Report'' U.S. 
Department of Transportation Volpe National Transportation Systems 
Center, Cambridge MA, September 1996.
---------------------------------------------------------------------------

    As described in the section on procedures for product submission 
included at the end of this notice, when manufacturers submit their 
instruments for testing, they are required to submit copies of the 
instructions they provide to end users. NHTSA will examine these 
instructions to ensure that they provide sufficient information about 
this procedure. Products submitted without this information will not be 
tested.
(b) The Effect of Variable Atmospheric Pressure on Dry Gas Calibrators
    USAT commented that dry gas calibrating units may exhibit a 
pressure-dependent concentration effect that wet bath calibrating units 
do not. The packaging of a dry gas calibrator compresses a large volume 
of an alcohol-in-inert gas mixture into a metal cylinder of only about 
one (1) liter. The concentration of the alcohol in the gas is given by 
the Ideal Gas Law 4: PV = nRT, where P is the pressure of 
the gas, V is the volume, n is the number of moles of gas, R is the gas 
constant, and T is the temperature of the gas. The concentration of the 
gas is obtained as a function of pressure and temperature: 
Concentration = n/V = P/RT.
---------------------------------------------------------------------------

    \4\  Farrington Daniels & Robert Alberty, ``Physical Chemistry'' 
3rd Ed., John Wiley & Sons, New York, 1966.
---------------------------------------------------------------------------

    When a calibration check is performed, some of the gas in the 
cylinder is released by operating the release valve. The volume of the 
released gas will expand and its pressure will drop until prevailing 
atmospheric pressure is reached. The gas is prepared so that the 
desired concentration is obtained at normal atmospheric pressure, 760 
millimeters of mercury. However, atmospheric pressure varies slightly 
from day to day and can change suddenly at times. The most significant 
effect comes from high elevations, where prevailing atmospheric 
pressure is significantly lower than 760. Atmospheric pressure 
corrections are made using an equation derived from the Ideal Gas Law: 
C = C760 X P/760, where C is concentration and P is the 
prevailing atmospheric pressure.
    In order for any calibrating unit to operate properly under such 
atmospheric pressures, accurate pressure correction must be made. The 
agency has tested the dry gas calibrating units placed on the CPL in 
this publication using this pressure correction procedure and has 
determined that these devices meet the Model Specifications. The agency 
concludes that the pressure dependent concentration effect is 
consistent and well established and that pressure correction procedures 
suggested by manufacturers are effective and produce accurate results.
    As described in the section on procedures for product submission 
included at the end of this notice, when manufacturers submit their 
instruments for testing, they are required to submit copies of the 
instructions they provide to end users. While manufacturers already 
provide information on pressure corrections in their instructions to 
end users, these Model Specifications have been amended to require that 
the instructions include information about how atmospheric pressure 
corrections should be made. NHTSA will examine manufacturers' 
instructions to ensure that they provide sufficient information about 
these pressure correction procedures. Products submitted without this 
information will not be tested. NHTSA believes that these procedures 
will be effective when used by properly qualified breath alcohol 
technicians.
(c) The Performance of Dry Gas Calibrators Over Range of Temperatures 
and Concentrations
    Throughout its written comments, USAT argues that dry gas standards 
should not be accepted because they have not been shown to be 
comparable to wet bath standards. USAT argues:

    Further substantial equivalence of the dry-EtOH and wet 
simulators must be shown over the range of environmental 
temperatures and pressures likely to be encountered during normal 
field usage of any of the devices appearing on the CPL * * * [and] 
over the range of NHTSA tested concentrations * * * throughout the 
operating lifetime of the dry gas [calibrating units] * * *
    Results of comparative performance of dry-ETOH [calibrating 
units] versus wet simulator [calibrating units] need to be publicly 
presented in scientific forums and published in the technical 
literature to establish a level of confidence that dry gas 
[calibrating units] yield substantially equivalent results to those 
obtained for decades from conventional wet simulator [calibrating 
units].

USAT commented that ``Dry gas EtOH [calibrating units] must be required 
to show equivalent performance over the entire range of environmental 
conditions used to test wet bath simulator [calibrating units].'' The 
agency tests both wet bath and dry gas calibrating units according to 
the Model Specifications. The agency believes that the Model 
Specifications require testing over an appropriate range of 
temperatures and concentrations. Dry gas calibrating units are required 
to show equivalent performance over the entire range of environmental 
conditions used to test wet bath calibrating units.

[[Page 43419]]

4. Are Dry Gas Calibrating Units Sufficiently Accurate?

    USAT states that it would favor use of dry gas calibrating units 
when ``it has been adequately demonstrated that dry EtOH [calibrating 
units] give results comparable to those obtained with conventional `wet 
bath simulator calibration units'.''
    The same Model Specifications used to test the accuracy and 
precision of wet bath calibrating units are used to ensure the quality 
and performance of dry gas calibrating units. All units are tested over 
the same range of temperatures and concentrations. All dry gas 
calibrating units placed on the CPL in this publication conform to the 
Model Specifications. Any unit that fails to meet the requirements of 
the Model Specifications would not be included on the agency's list of 
conforming products.

5. Miscellaneous Issues

(a) Quality Assurance Plan
    Scott Gas recommended that the agency require Quality Assurance 
Plans (QAPs) for calibrating units. QAPs are used to provide 
information on the correct use, proper maintenance procedures and other 
specific requirements of a calibration device. Scott Gas recommended 
that the QAP address issues such as NIST traceability, mechanisms for 
product coding and traceability, list of proper delivery equipment, 
specifications on the containers being submitted for approval, shipping 
and storage information, written laboratory certification and 
manufacturing procedures, DOT specification documentation on 
containers, a specified uncertainty at the 95% confidence level and 
shelf life results.
    NHTSA strongly endorses the need for quality control in 
manufacturing, but believes that this is addressed appropriately by the 
manufacturers of these instruments. When calibrating units are used by 
law enforcement officials, quality control measures are also taken 
under the programs of each state. In transportation workplace testing, 
quality control is ultimately handled by the existing requirement for 
QAPs for evidential breath testers and alcohol screening devices 
(Screeners) which address calibration accuracy. The evidential breath 
tester QAPs call for calibration checks using an approved calibrating 
unit. If an evidential breath tester or a Screener gives an incorrect 
reading when a calibration check or a calibration is conducted, it 
suggests that there is an error in the system consisting of the 
evidential breath tester (or Screener), the breath alcohol technician, 
or the calibrating unit. NHTSA believes that the safeguards already in 
place in the QAPs for evidential breath testers and Screeners make it 
unnecessary to require an additional QAP specific to the calibrating 
unit.
(b) Stability of Dry Gas Calibrators Over Their Operating Life
    USAT commented that ``Further substantial equivalence of the dry-
EtOH and wet simulators must be shown over the range of NHTSA tested 
concentrations * * * throughout the operating lifetime of the dry gas 
[calibrating units] * * *'' Scott Gas also commented that 
``presentation of gas manufacturer stability documentation to NHTSA, 
before inclusion on the CPL, plus NHTSA evaluation of aged product 
should be done in order to assess the ``real life'' performance of the 
product.''
    The agency's experience indicates that dry gas calibrating units 
are normally stable even after years of storage. In addition, NHTSA has 
verified that National Institute of Standards and Technology Reference 
Gas Mixtures used to evaluate dry gas cylinders remained stable to 
within plus-minus0.001 BrAC for a one year period. The 
agency has concluded that manufacturers will not be required to provide 
stability documentation.
    NHTSA shall certify that the CPL does, in fact, reflect calibrating 
units which meet the performance criteria set forth in the Model 
Specifications. NHTSA reserves the right to test any unit on the CPL 
throughout its useful life to ensure that the unit is performing in 
accordance with the Model Specifications. In addition, in the section 
on procedures for a product submission, included at the end of this 
notice, NHTSA requests that users of calibrating units provide both 
acceptance and field performance data to NHTSA's Office of Traffic 
Injury Control Programs. NHTSA will conduct a special investigation if 
information gathered from the field indicates that a device on the CPL 
is not performing in accordance with the Model Specifications.
    After the recent expansion of the use of dry gas calibrators, one 
manufacturer found that the concentration of some dry gas calibrators 
had changed from the stated concentrations after weeks or months of 
storage. A recall of all cylinders in use was ordered. The problem was 
investigated and, after extensive testing it was traced to defects in 
certain cylinders and was corrected.
(c) National Institute of Standards and Technology Reference Gas 
Mixtures
    In the Notice published on December 24, 1994, NHTSA proposed to 
revise the Model Specifications to permit use of National Institute of 
Standards and Technology Reference Gas Mixtures (NISTRGMs) as reference 
samples to evaluate the accuracy of dry gas calibrating units by gas 
chromatography.
    Use of these dry gas standards allows reliable evaluation of dry 
gas calibrators by the gas chromatograph technique. USAT commented 
that:

    It is rumored that NISTRGMs are manufactured by Scott Specialty 
Gases/Scott Medical Products Inc. If true, the NHTSA-proposed 
substitution of NISTRGMs to replace wet bath simulator standards for 
the testing of any Scott Gas gaseous standards amounts to one 
manufacturer certifying itself and claiming the blessing of both 
NIST and NHTSA.

The NISTRGMs obtained by the Volpe center were manufactured by Scott 
Specialty Gases, but were obtained from and analyzed independently by 
the Department of Commerce National Institute for Standards and 
Technology (NIST). NIST attested in writing to the accuracy of each 
individual cylinder of gas which was obtained by the Volpe Center.
(d) The Comparability of Dry Gas Calibrating Units When Used With a 
Variety of Evidential Breath Testing Devices
    USAT commented that ``dry gas standards are likely to give 
different results when used on [evidential breath testers] based on 
different technologies.'' According to USAT, there have been reports 
that dry gas calibrating units do not yield the same results for 
certain breath testers as wet bath calibrating units. USAT asserts that 
a small ``offset'' in test result reportedly occurs when dry gas 
calibrators are used for these breath testers compared with wet 
calibrators at the same concentration. The offset for fuel cell breath 
testers is reported to be -0.002 BrAC when dry calibrators are used to 
check calibration of fuel cell evidential breath testers.
    Performance requirements contained in NHTSA's Model Specifications 
for evidential breath testers require that these instruments be 
accurate to 0.005 or 5% of test BrAC, whichever is greater, 
with a standard deviation not greater than 0.004. The performance 
requirements for calibrating units require the devices to be accurate 
to within 0.002 BrAC of the test BrAC with relative standard deviation 
of 2%. Any offset associated with a particular calibrator is not 
considered.

[[Page 43420]]

    Agency testing indicates that dry gas calibrating units can be used 
with infra-red and fuel cell breath testers.5 The agency 
tested four fuel cell testers, one fuel cell/infra-red combination 
tester with readout from the fuel cell sensor, and one infra-red tester 
to obtain wet dry comparison data. The instruments tested were:

    \5\ Flores, Arthur, ``Dry Gas Calibrating Units Report'', U.S. 
Department of Transportation, Volpe National Transportation Systems 
Center, Cambridge, MA, September, 1996.
---------------------------------------------------------------------------

Life Loc, Inc. PBA 3000 and PBA 3000X (FC)
CMI, Inc. Intoxilyzer 400 (FC)
Intoximeters, Inc. AlcoSensor IV (FC)
National Draeger, Inc. Breathalyzer 7410-II (FC)
Intoximeters, Inc. EC-IR (FC/IR)
CMI, Inc. Intoxilyzer 5000 (IR)

Measurements were made alternately using first a sample from a wet bath 
calibrator, then a sample from a dry gas calibrator. Five measurements 
of each type of sample were made on each of the testers. The wet 
calibrator solutions were prepared to produce the same concentrations 
as the corresponding dry gas. Wet samples were produced using RepCo 
Marketing simulators (wet bath calibrating units). Dry samples were 
obtained from dry gas calibrating units from Scott Specialty Gases, 
Inc. (0.04 BrAC) and Gateway Airgas, Inc. (0.04 and 0.088 BrAC). The 
concentration of the Scott gas was verified by Intoximeters, Inc. and 
the concentration of the gas from Gateway Airgas was verified by infra-
red spectroscopy at the Volpe center. The factory calibrations of the 
breath testers were not adjusted. The reliability of the ``true value'' 
of the wet or dry standards can be taken as known values to within 
about 0.001 BrAC. Thus, the true value of a wet sample or a 
dry sample at 0.040 BrAC can be expected to be correct to within about 
0.001 BrAC.
    The differences between wet bath and dry gas calibrating units were 
negligible when the comparisons were made using infra-red breath 
testers. These differences were around 0.002 BrAC and are not 
noticeable unless comparisons are made carefully, because this value is 
near the accuracy limit of the calibrating units.
    The differences observed when comparisons were made using fuel cell 
type breath testers, the next most widely used type of breath tester, 
were more noticeable, especially at high alcohol levels. The offset for 
fuel cell breath testers averaged somewhat less than 4% of the nominal 
BrAC when dry gas calibrators were used to check calibration of fuel 
cell evidential breath testers. The offsets found for the breath 
testers ranged from -0.0014 BrAC to 0.0026 BrAC when compared at the 
0.04 BrAC level, and from -0.0020 to 0.0052 when compared at the 0.088 
level. The standard deviations for the wet and dry data were in the 
fourth decimal place except in one instance when a value of 0.002 was 
obtained, which was still acceptable. These results indicate that the 
offsets are small and reproducible enough that reliable corrections can 
be applied to ensure accurate test results. The offsets observed cannot 
be assumed to arise only from the inherent differences in measurement 
of moist samples compared to the measurement of dry samples since there 
are also uncertainties of 0.001 in the true concentration 
of wet bath or dry gas calibration unit vapors.
    Offsets must be indicated by manufacturers in their instructions to 
end users. Manufacturers are required to include their instructions in 
a submission of a calibrating unit for testing. The agency will examine 
the instructions to ensure that they provide sufficient information on 
offsets necessary for certain breath testers. Products submitted 
without this information will not be tested.
    Gas Chromatograph breath testers depend on extensive surface 
interaction with the sample being analyzed, and the greatest 
differences between dry and wet standards are seen with this type of 
breath tester. In its laboratory, NHTSA has found that the effects are 
not stable. They vary with type and condition of resolving column used. 
Accordingly, NHTSA believes that dry gas calibrating units should not 
be used with gas chromatograph breath testers because the results are 
too variable. The agency will include a footnote on the CPL concerning 
the use of dry gas standards with gas chromatograph evidential breath 
testers, indicating that the agency does not recommend the use of dry 
gas calibrating units with gas chromatograph evidential breath testers.

B. Procedures for a Product Submission

    Testing of calibrating units submitted by manufacturers to these 
Model Specifications will continue to be conducted by the DOT Volpe 
National Transportation Systems Center (VNTSC). Tests will continue to 
be conducted semi-annually or as necessary. Manufacturers wishing to 
submit calibrating units for testing must apply to NHTSA for a test 
date (Office of Traffic Injury Control Programs, NTS-11, NHTSA, 400 
Seventh Street, S.W., Washington, D.C. 20590). Normally, at least 30 
days will be required from the date of notification until the test can 
be scheduled. One week prior to the scheduled initiation of the test 
program, the manufacturer will deliver at least one unit of the device 
to be tested to: VNTSC, DTS-75, 55 Broadway, Kendall Square, Cambridge 
MA 02142. The manufacturer shall be responsible for ensuring that the 
unit is operating properly. If the manufacturer wishes to submit a 
duplicate, backup unit, it may do so.
    When a manufacturer delivers a device to be tested, it shall also 
deliver to VNTSC specifications and drawings that fully describe the 
unit and the Operator's Manual and Maintenance Manual normally supplied 
with purchase of the equipment. Proprietary information will be 
respected. (See 49 CFR Part 512, regarding the procedures by which 
NHTSA will consider claims of confidentiality.)
    The manufacturer shall also deliver the instructions that will 
accompany the device when it is sold. The instructions shall include 
information about the procedures to be followed to protect against 
possible condensation that might occur as a result of freezing during 
shipment and to correct for atmospheric pressure. The instructions 
shall also include information about any offsets that may apply to the 
use of a particular type of breath tester. NHTSA will examine these 
instructions to ensure that they provide sufficient information about 
these matters. Products submitted without this information will not be 
tested.
    The manufacturer will have the right to check the calibrating unit 
between arrival in Cambridge and the start of the test, and to ensure 
that the calibrating unit is in proper working condition but will have 
no access to it during the tests. Any malfunction of the calibrating 
unit which results in failure to complete any of the tests 
satisfactorily will result in a finding that it does not conform to the 
Model Specifications. If a unit fails to conform, it may be resubmitted 
for testing after appropriate corrective action has been taken.
    On the basis of these results, NHTSA will publish a Conforming 
Products List (CPL) identifying the calibrating units that conform to 
the Model Specifications.
    Retesting of units will be conducted when necessary. NHTSA intends 
to modify and improve these Model Specifications as new data and 
improved test procedures become available. (The test procedures may be 
altered in specific instances, if necessary, to meet the unique design 
features of a calibrating unit). If these Model Specifications are 
modified, notification will be provided in the Federal Register. If 
NHTSA determines that retesting to the modified

[[Page 43421]]

specifications is necessary, a manufacturer whose equipment is listed 
on the CPL will be notified to resubmit the equipment for testing to 
the modified specification only.
    NHTSA will certify that the CPL does, in fact, reflect calibrating 
units which meet the performance criteria set forth in the Model 
Specifications. NHTSA reserves the right to test any unit on the CPL 
throughout its useful life to ensure that the unit is performing in 
accordance with the Model Specifications.
    If at any time a manufacturer plans to change the design of a 
calibrating unit currently on the CPL, the manufacturer shall submit 
the proposed changes to the Office of Traffic Injury Control Programs 
for review. Based on this review, NHTSA will decide whether the change 
will require retesting of the unit. Normally, such retesting will be 
accomplished the next time testing is performed. Guidance to 
manufacturers on considerations governing this decision are available 
from NHTSA's OTICP, upon request.
    OTICP will be the point of contact for information about acceptance 
testing and field performance of equipment already on the list. When it 
is available, NHTSA requests that users of calibrating units provide 
both acceptance and field performance data to OTICP. Information from 
users will be used to: (1) help NHTSA determine whether units continue 
to perform according to the NHTSA Model Specifications and (2) ensure 
that field use does not indicate excessive breakdown or maintenance 
problems.
    If information gathered indicates that a device on the CPL is not 
performing in accordance with the Model Specifications or demonstrates 
problems involving the device, NHTSA will direct VNTSC to conduct a 
special investigation. This investigation may include visits to users 
and additional tests of the unit obtained from the open market. If the 
investigation indicates that the units actually sold on the market are 
not meeting the Model Specifications, then the manufacturer will be 
notified that the unit may be removed from the list. In this event the 
manufacturer shall have 30 days from the date of notification to reply.
    Based on the VNTSC investigation and any data provided by the 
manufacturer, NHTSA will decide whether the unit should remain on the 
list. Upon resubmission, the manufacturer must submit a statement 
describing what has been done to overcome the problems that led to the 
dropping of the unit in question from the list.

C. Infra-red Spectroscopy

    In this notice, NHTSA is proposing an alternate procedure which 
uses infra-red spectroscopy for the evaluation of dry gas units (see 
Appendix A). It is proposed as an amendment to the Model Specifications 
for Calibrating Units published in this notice. In infra-red 
spectroscopy, the wet bath or dry gas sample to be analyzed is passed 
into a chamber through which infra-red radiation is transmitted. The 
wavelength of the transmitted radiation is chosen so that some of it is 
absorbed by alcohol. According to the Beer-Lambert Law of absorption of 
radiation,6 the amount of energy absorbed by the sample in 
the chamber is proportional to the concentration of the alcohol in the 
sample. By measuring the amount of radiation transmitted when the 
sample chamber is empty and the amount transmitted when the sample is 
present, the concentration of the alcohol in the sample can be 
determined.
---------------------------------------------------------------------------

    \6\ Farrington Daniels & Robert Alberty, ``Physical Chemistry'' 
3d Ed. John Wiley & Sons, New York, 1966.
---------------------------------------------------------------------------

    The agency believes that use of infra-red spectroscopy will offer 
several important advantages. First, the technique can be used to 
evaluate both wet bath calibrating units and dry gas calibrating units 
because surface interactions do not effect the analysis. Second, 
standards used in the evaluations can be prepared at the Volpe Center, 
eliminating the necessity of obtaining standards from an outside 
source.

D. Comments

    Interested persons are invited to comment on the proposed alternate 
procedure described in this notice. It is requested, but not required 
that 10 copies be submitted. Comments must not exceed 15 pages in 
length (49 CFR 553.221). Necessary attachments may be appended to those 
submissions without regard to the 15 page limit. This limitation is 
intended to encourage commentors to detail their primary arguments in a 
concise fashion.
    All comments received before the close of business on the comment 
closing date indicated above will be considered, and will be available 
for examination in the docket at the above address, both before and 
after that date. To the extent possible, comments filed after the 
closing date will also be considered. However, the amendments to the 
Model Specifications may be published at any time after that date, and 
any comments received after the closing date and too late for 
consideration with regard to the action will be treated as suggestions 
for future revisions to the Specifications. NHTSA will continue to file 
relevant material in the docket after the closing date as it becomes 
available. It is recommended that interested persons continue to 
examine the docket for new material.
    Those persons who desire to be notified upon receipt of their 
comments in the docket should enclose a self-addressed stamped postcard 
in the envelope with their comments. Upon receiving the comments, the 
docket supervisor will return the postcard by mail.

E. Conforming Products List

    The Conforming Products List (CPL), which appears as Appendix B to 
this notice, lists the calibrating units that have been retested to 
date at the lower BACs (i.e., at 0.020, 0.040, 0.080, and 0.160) and 
found to conform to the Model Specifications reprinted herein. The CPL 
also lists devices that have not been tested at these lower BAC levels, 
but which were listed on a previous CPL for calibrating units (58 FR 
26030) on the basis that they were tested and found to conform to the 
earlier model specifications when tested at BAC levels 0.050, 0.100 and 
0.150. These devices have been identified with an asterisk.
    This CPL also includes five new listings: four dry-gas calibrating 
units and one wet-bath calibrating unit. The dry gas units include: 
Model EBSTM'' Gaseous Ethanol Breath Standard submitted by 
Scott Specialty Gases, Inc. of Plumsteadville, PA; the Ethanol Breath 
Alcohol Standard submitted by Gateway Airgas (previously known as A.G. 
Specialty Gas Company, or Acetylene Gas Company) of St. Louis, MO; the 
AlcoCal Breath Alcohol Standard submitted by Lion Laboratories, plc of 
Cardiff, Wales, UK; and Compressed ethanol-in-nitrogen submitted by 
Liquid Technology Corporation of Orlando, FL. All of the dry-gas 
calibrating units were tested using the alternate procedure that uses 
the NISTRGM. The new wet-bath unit is Model 3402C submitted by RepCo 
Marketing, Inc., of Raleigh, NC.
    In consideration of the foregoing, NHTSA amends the Model 
Specifications for Calibrating Units, as last published in the Federal 
Register on December 29, 1994 (59 FR 67377), as set forth below. NHTSA 
proposes to further amend these Model Specifications, as set forth in 
Appendix A.

[[Page 43422]]

Model Specifications for Calibrating Units for Breath Alcohol Testers

1.0  Purpose and Scope
    These specifications establish performance criteria and methods for 
testing of calibrating units which provide known concentrations of 
ethanol vapor for the calibration or calibration checks of breath 
alcohol testers. The results of this testing are intended for use in 
the conformance testing for the maintenance of a Conforming Products 
List for calibrating units.
2.0  Definitions
    2.1  Conformance testing. Testing to check the conformance of a 
product with these model specifications in advance of and independent 
of any specific procurement action.
    2.2  Concentration units. Blood alcohol concentration: grams 
alcohol per 100 milliliters blood or grams alcohol per 210 liters of 
breath in accordance with the Uniform Vehicle Code, Section 11-
903(a)(5).7 BrAC is often used to indicate that the 
measurement is a breath measurement, i.e. gram alcohol per 210 liters 
of breath.
---------------------------------------------------------------------------

    \7\ Available from National Committee on Uniform Traffic Laws 
and Ordinances, 405 Church Street, Evanston, IL 60201.
---------------------------------------------------------------------------

    2.3  Relative Standard Deviation (RSD). The ratio of the standard 
deviation (SD) of a series of measurements to the mean of the series 
expressed as a percentage:

RSD=(SD/Mean) x 100 percent

    2.4  Standard Deviation (SD). A common indication of precision in 
the measurement of the concentration of a succession of N vapor 
samples.

SD={Sum (Xi-Xm)\2\/(N-1)}\1/2\
where Xi=a single measurement result;
Xm=the average of the measurements;
N=the number of measurements made in the test.

    2.5  Systematic Error (SE). An indication of the accuracy of the 
measurement of the concentration of a succession of vapor samples.

SE=Xm-test BrAC
    2.6  Least Squares Fit Calibration Curve. A line fitted to a number 
of measurement pairs, one the independent value (X) and the other the 
dependent value (Y), over a measurement range.
    The fitted line is of the form: Y=a+bX, where intercept, 
a=Ym-bXm, and slope, 
b=(SumXiYi-NXmYm)/
(SumXi2-nXm2).
3.0  Tests and Requirements
    If the BrAC of the CU is fixed, perform the tests at the fixed 
BrAC; otherwise, prepare the CU for testing at 0.08 BrAC except as 
otherwise required in Test 1 below. Each of the tests require 10 
measurements to three decimal places using the test procedure specified 
in 3.1. The CU will be operated according to the manufacturer's 
instructions. Unless otherwise specified, the tests will be performed 
in the absence of drafts and at prevailing normal laboratory 
temperature, humidity, and barometric pressure. Performance 
requirements are:

-0.002 BrAC  SE  + 0.002 BrAC; RSD  2%

    Test 1. Precision and Accuracy. Test at each specified BrAC.
Test 1.1: 0.020 BrAC
Test 1.2: 0.040 BrAC
Test 1.3: 0.080 BrAC
Test 1.4: 0.160 BrAC
    Test 2. Ambient Temperature. Use a temperature chamber controllable 
to 2 deg.C. Soak the CU at the specified temperature for 1 
hour, being careful to prevent drafts on the device, then test at that 
temperature.

Test 2.1: 10  deg.C
Test 2.2: 30  deg.C.

    Test 3. Input Power. If the CU is powered by nominal voltages of 
120 volts AC or 12 volts DC, condition the device for one half hour at 
the appropriate input voltage specified below, then test at that 
voltage. Monitor the input power with a voltmeter accurate to 
2% full scale in the range used and re-adjust the voltage, 
if necessary. If the voltage is AC, conduct tests 3.1 and 3.2. If the 
voltage is DC, conduct tests 3.3 and 3.4.

Test 3.1: 108 Volts/AC
Test 3.2: 123 Volts/AC

Test 3.3: 11 Volts/DC
Test 3.4: 15 Volts/DC

    Test 4. Electrical Safety Inspection. Examine the CU for protection 
of the operator from electrical shock. Examine for proper use of input 
power fuses, and verify that there are no exposed male connectors at 
high potential. Determine that overheating does not occur during 
operation and that undue fire hazards do not exist.
3.1  Test Procedure (Original, Wet-bath)
    Equipment and Supplies: Gas Chromatograph capable of complete 
resolution of ethanol in test samples, with heated gas sampling valve. 
Water bath thermostated at 34 deg.C 0.1 deg.C. Glass 
Reference Sample Bottles (300 ml capacity or greater) with Stopper and 
Inlet and Outlet Air Hoses (see Figure 1). Hoses should be about 1/8'' 
OD Teflon tubing. Reference Ethanol Solutions prepared using class A 
glassware and American Chemical Society reagent grade ethanol or USP 
grade ethanol. The purity of the ethanol used shall be compared with 
the National Institute of Standards and Technology (NIST) Standard 
Reference Material for ethanol. Use the value of Harger, et al., for 
the partition ratio for concentration of ethanol in head space to 
concentration in solution at 34 deg.C, Ka/w = 0.000393 8 to 
prepare two solutions which, when thermostated at 34 deg.C, produce 
head space ethanol vapor concentrations that bracket the test BrAC by 
no more than 20%. Small Air Pump for bubbling air through 
reference solutions (see Figure 1).
---------------------------------------------------------------------------

    \8\ RN Harger, BB Raney, EG Bridwell, MF Kitchel, J. Biol. Chem. 
183, 197-213 (1950). Additional data from Harger in a private 
communication (see 49 FR 48869).
---------------------------------------------------------------------------

    Step 1. Prepare the Gas Chromatograph for measurement of vapor 
samples. Adjust instrument temperatures, gas flows, detector, and 
recording device for optimum response for ethanol. Prepare the CU for 
use according to manufacturer's instructions.
    Step 2. Fill two reference solution bottles to \3/4\ full with 
above reference solutions. Insert stopper assemblies with bubble line 
and alcohol vapor line in place and put bottles in the water bath with 
water level up to the stopper. Connect air pump to bubble line. Connect 
alcohol vapor line to gas chromatograph sampling valve inlet fitting. 
Allow 1 hour for temperature equilibrium to be achieved.
    Step 3. Turn on air pump which has been pre-set to pump air through 
the reference solution bottle-gas chromatograph sampling assembly at a 
rate just sufficient to thoroughly flush the system in 10 seconds. 
After flushing is complete, allow the sample to relax to atmospheric 
pressure, then inject the reference sample onto the gas chromatograph 
column. In this way, obtain 5 chromatograms of one of the reference 
solution head space ethanol vapors.
    Step 4. Thoroughly flush the sample loop with vapors from the CU 
device, while avoiding over-pressurizing of the sampling system. To 
prevent condensation of alcohol, warm the transfer line if necessary. 
Allow the sample to relax to atmospheric pressure, then inject the 
sample onto the column. In this way, obtain 10 ethanol chromatograms 
using the CU device.
    Step 5. Repeat step 3 using the second reference solution.
    Step 6. Calculations. Peak height to BrAC conversion factor. For 
each ethanol peak obtained in step 2 and step

[[Page 43423]]

5, calculate a conversion factor for ethanol concentration by dividing 
the equivalent BrAC of the vapor sample by the peak height obtained for 
that sample. From the ten samples, obtain the mean and the RSD of the 
conversion factors. If the RSD obtained fails to meet the criteria for 
RSD in 3.0, perform necessary troubleshooting and repeat the procedure 
from Step 1. Use the mean of the conversion factors to calculate the 
BrAC for each of the 10 ethanol peaks obtained in step 4. Calculate the 
mean, the RSD, and the systematic error of the experimental BrACs.

BILLING CODE 4910-59-P

Figure 1. Wet Bath Reference Sample Set-up. Sample lines \1/8\'' 
Teflon. The bubble line should extend at least 4 inches below 
surface of the solution. The length of the alcohol vapor line from 
the headspace to the gas chromatograph should be minimized.
[GRAPHIC] [TIFF OMITTED] TN13AU97.000


BILLING CODE 4910-59-C
3.2  Test Procedures (for dry gas Calibrating Units): Alternate Test 
Method Using National Institute of Standards and Technology Reference 
Gas Mixtures (NISTRGMs) in Place of Wet Bath Reference Samples
    The following alternate method for the evaluation of dry gaseous 
ethanol calibration devices is presented.
    Additional required material: For the alternate method for 
evaluation of dry gaseous ethanol calibration devices, the following 
will be required: Four cylinders of National Institute of Standards and 
Technology ethanol-in-inert gas Technical Reference Gas Mixtures 
(NISTRGMs) which span the BrAC range 0.01 to 0.16.
    Alternate Procedure for evaluation of dry gaseous ethanol 
calibration devices. This procedure substitutes the use of NISTRGMs in 
place of the wet bath reference samples when evaluating dry gas CUs.
    Step A1. Connect one of the NISTRGM cylinders to the inlet of the 
gas chromatograph sampling valve and pass reference gas through the 
sampling system at a rate just sufficient to thoroughly flush the 
system in about 10 seconds. Allow the sample to relax to atmospheric 
pressure, then inject the sample onto the column. In this way, obtain 5 
chromatograms of the reference gas.
    Step A2. Repeat Step A1 for each of the four NISTRGM reference gas 
mixtures.
    Step A3. Calculate the RSD of the concentration divided by peak 
height data obtained in Step A1 and Step A2. If the calculated RSD 
meets the criteria of 3.0, calculate the slope and intercept of the 
least squares fit calibration line for conversion of peak height to 
BrAC. Using the average peak height of each NISTRGM and the slope and 
intercept data, calculate the concentration of each NISTRGM. If the 
resulting concentrations are within the stated accuracy of the NISTRGM, 
proceed to Step A4.
    Step A4. Connect the calibrating device to the inlet of the gas 
chromatograph sampling system and allow the calibrating device gas to 
flow at a rate just sufficient to thoroughly flush the sampling system 
in about 10 seconds. Allow the sample to relax to atmospheric pressure, 
then inject the sample onto the column. In this way, obtain 10 
chromatograms of the calibrating device gas.
    Step A5. Calculations. Using the peak height data obtained in Step 
A4 and intercept and slope data obtained in Step A3, calculate the BrAC 
for each of the 10 peak heights. Calculate the mean, RSD, and 
systematic error of the calculated BrACs.

    Authority: 23 U.S.C. 402; delations of authority at 49 CFR 1.50 
and 501.

    Issued: August 7, 1997.
James Hedlund,
Associate Administrator for Traffic Safety Programs.

Appendix A--Proposed Alternate Procedure Using Infra-Red Spectroscopy

    This appendix presents an alternate procedure using infra-red 
spectroscopy that is suitable for evaluating vapor samples from either 
wet-bath CUs, or from dry-gas CUs.

[[Page 43424]]

    3.3  Proposed Test Procedures (for dry gas or wet bath calibrating 
units).
    3.3.1  General. General. The method uses the Beer-Lambert Law of 
absorption of radiant energy by fluids

I = Io X e-abc
Where:
Io is the energy entering the sample chamber of a 
spectrophotometer containing the sample to be analyzed.
I is the energy transmitted from the sample chamber.
a is the absorptivity of the sample.
b is the radiation path length of the sample chamber.
c is the concentration of the sample in the sample chamber.

    A convenient form of the Beer-Lambert law is

Ln(Io/I) = abc

where the term (Ln(Io/I), the logarithm of the ratio of 
incident to transmitted energy, is called the absorbance of the sample. 
In the procedure described below, the terms a and b are treated as a 
single quantity, ab, and the term c is BrAC.
    3.3.2  Test Procedure.
    Equipment and Supplies. Infra-red Spectrophotometer with sample 
chamber that can be heated to above 40 deg. C. A non-dispersive 
instrument with appropriate band pass filters and configured to measure 
breath alcohol samples, such as an infra-red evidential breath tester 
listed on the NHTSA Comforting Products List for evidential breath 
testers may be used. The detector voltage of the instrument must be 
accessible for measurement. The sampling hoses of the device may be 
altered for more convenient processing of test samples. Water bath 
thernostated at 34 deg.C 0.1 deg.C. Glass Reference Sample 
Bottles (300 ml capacity or greater) and Stoppers with Bubble and 
Alcohol Vapor lines (see Figure 2). Reference Ethanol Solutions 
prepared using Class A glassware and American Chemical Society reagent 
grade ethanol or USP grade ethanol. The purity of the ethanol used 
shall be compared with the National Institute of Standards and 
Technology (NIST) Standard Reference Material for ethanol. Use the 
value of Harger, et al., for the partition ratio for concentration of 
ethanol in head space to concentration in solution at 34 deg. C, 
Ka/w = 0.0003932 to prepare two aqueous alcohol 
solutions which bracket the test BrAC by no more than 20%. 
A cylinder of inert Flushing Gas, which is optically clear in the 
absorption region used for measurement. This gas will be used to flush 
the sample chamber of the spectrophotometer and to deliver reference 
headspace vapors and wet bath sample vapors into the sample chamber. 
Pressure regulating valve with teflon delivery hose for controlling 
flow and delivery of flushing gas.
    Step B1. Prepare the spectrometer for measurement of vapor samples. 
Prepare the CU for use according to manufacturer's instructions.
    Step B2. Fill a reference sample bottle to \3/4\ full with water 
and two reference sample bottles to \3/4\ full with the above reference 
solutions. Insert stopper assemblies ensuring that the end of the 
bubble line reaches to at least 4 inches below the surface of the 
solution, then place the bottles in the water bath with water level up 
to the stopper. Allow 1 hour for temperature equilibrium to be 
achieved.
    Step B3. Connect the bubble line of the sample bottle containing 
water only to the flushing gas valve and the vapor line to the 
spectrophotometer inlet and flush the sample chamber with water vapor 
and obtain the detector voltage reading. Then flush the detector 
chamber with flushing gas only and obtain the detector reading. Repeat 
2 times to obtain 3 sets of readings. If the CU being evaluated is a 
wet bath device, skip this step and proceed to Step 4.
    Step B4. In the manner of Step 3, obtain 5 sets of detector 
readings using one of the reference alcohol solution bottles.
    Step B5. In the manner of Step 3, obtain 10 sets of detector 
readings from the CU being evaluated. If the CU is a wet bath device, 
use the flushing gas fill the sample chamber, operating the device 
according to manufacturer's instructions. If the CU device is a dry gas 
device, fill the sample chamber according to manufacturer's 
instructions.
    Step B6. Repeat Step 5 using the other reference alcohol solution 
bottle.
    Step B7. Repeat Step 3.
    Step B8. Calculations. For each measurement pair, I0 is 
the detector voltage obtained for the flushing gas alone in the sample 
chamber and I is the voltage obtained for the flushing gas with 
reference sample or test sample in the sample chamber corrected for 
water vapor absorption, i.e.; the detector voltage obtained for 
headspace reference samples at 0.000 BrAC. Use the average of 6 voltage 
readings obtained for the water samples for the correction for water 
vapor absorption (I=Isample-Iwater). In the case 
of wet bath device samples, there is no correction for water vapor 
absorption. If the detector is biased, I will be the difference between 
the bias voltage and the above voltage.
    Calulate the absorbance of each of the 10 reference samples. Divide 
each absorbance by the corresponding BrAC of the sample. Obtain the 
mean (which is the factor ab), SD, and RSD for the 10 ratios. If the 
RSD is more than 2%, trouble shoot the procedure and repeat.
    Calculate the absorbance for each of the 10 CU test samples. Divide 
each by the ab factor to obtain the BrAC for each of the 10 CU samples. 
Obtain the mean, SD, RSD, and SE.

BILLING CODE: 4910-59-P


[[Page 43425]]



Figure 2. Equipment set-up. Bubble and sample lines \1/8\'' teflon, 
minimized length. Depth of bubble line into reference solution at 
least 4''. The alcohol vapor line from the headspace to the IR 
specrophotometer should be minimized.
[GRAPHIC] [TIFF OMITTED] TN13AU97.001


BILLING CODE: 4910-59-C

Appendix B--Conforming Products List of Calibrating Units for Breath 
Alcohol Testers [Manufacturer and Calibrating Unit].\1\
---------------------------------------------------------------------------

    \1\ Infra-red (IR) and fuel cell breath testers may be 
calibrated with either wet-bath or dry-gas CUs. However, it is 
inadvisable to use dry gas CUs when calibrating gas chromatograph 
EBTs.
---------------------------------------------------------------------------

1. CMI, Inc., Owensboro, KY:
     Toxitest II
2. Federal Signal Corporation, CMI, Inc., Minturn, CO:
     Toxitest Model ABS120*
3. Gateway Airgas, Inc. (Formerly known as AG Specialty Gas, and 
Acetylene Gas Company), St. Louis, MO.
     Ethanol Breath Alcohol Standard (a dry gas standard).
4. Guth Laboratories, Inc., Harrisburg, PA:
     Model 34C Simulator \2\
---------------------------------------------------------------------------

    \2\ Several variations of the Model 34C Simulator have also been 
submitted to NHTSA for evaluation and meet these Model 
Specifications. They are: Model 34C Cal DOJ; Model 34-C-FM; and 34C-
NPAS.
---------------------------------------------------------------------------

     Model 3412
     Model 10-4
     Model 1214
5. Intoximeters, Inc., St. Louis, MO:
     Alco Breath Alcohol Standard* (a dry gas standard)
6. Lion Laboratories, plc, Cardiff, Wales, UK (a subsidiary of CMI, 
Inc.)
     AlcoCal Gas Standard (a dry gas standard).
7. Liquid Technology Corporation, Orlando, FL
     Alcohol-in-Nitrogen Calibrating Unit (a dry-gas 
standard).
8. Luckey Laboratories, Inc., San Bernadino, CA:
     Simulator*
9. National Draeger, Inc., Durango, CO.
     Mark II-A
10. PLD of Florida, Inc., Rockledge, FL:
     BA 500
11. Protection Devices, Inc., U.S. Alcohol Testing, Inc., Rancho 
Cucamonga, CA:
     LS34 Model 6100*
12. Repco Marketing, Inc., Raliegh, NC:
     AS-1
     Model 3402C
13. Scott Specialty Gases, Inc., Plumsteadville, PA
     Model EBS TM Gaseous Ethanol Breath Standard (a 
dry-gas standard).
14. Smith & Wesson Electronic Co., Springfield, MA:
     Mark II-A Simulator*
15. Systems Innovation, Inc., Hallsteaed, PA
     True-Test MD 901*
16. U.S. Alcohol Testing, Rancho Cucamonga, CA:
     Alco-Simulator 2000*
     Alco--Simulator 61000
    * Instruments marked with an asterisk (*) meet the Model 
Specifications in 49 FR 48864 (December 14, 1984), i.e. instruments 
tested at 0.050, 0.100, and 0.150). Instruments not marked with an 
asterisk meet the model specifications detailed in this notice, and 
were tested at 0.020, 0.040, 0.080, and 0.160 BrAC.

[FR Doc. 97-21331 Filed 8-12-97; 8:45 am]
BILLING CODE 4910-59-P