Highway Safety Programs; Model Specifications for Calibrating Units for Breath Alcohol Testers

Federal Aviation Administration

National Highway Traffic Safety Administration

[Docket No. 94-068; Notice 1]

Highway Safety Programs; Model Specifications for Calibrating

Units for Breath Alcohol Testers

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 in order to align test levels with

those in Model Specifications for Devices to Measure Breath Alcohol.

Under the earlier calibrating unit specifications, devices were

evaluated at 0.050, 0.100, and 0.150 BAC. To achieve alignment with the

amended breath tester specifications, this notice amends the

calibrating unit specifications to perform evaluations at 0.020, 0.040,

0.080, and 0.160 BAC. It also sets the acceptable Systematic Error at

not greater than 0.002 BAC at all BAC levels tested, rather

than the previous 2%. Finally, it makes a number of changes

to automate and make more efficient the procedures used for the

prescribed gas chromatography (GC) testing, as compared with the manual

GC procedures previously used and described in the 1984 Model

Specifications. Published with this notice is an amended Conforming

Products List (CPL) of calibrating units that meet the amended Model

Specifications for Calibrating Units. This notice also proposes and

seeks comments concerning a new, alternate testing procedure for the

evaluation of dry gaseous ethanol calibration devices, and other

changes designed to simplify the Model Specifications and make them

easier to read.

DATES: The amendments to the model specifications that change the BAC

levels at which evaluations are performed, adjust the Systematic Error

level, automate GC procedures, and the issuance of the CPL become

effective on December 29, 1994. Comments on the alternate testing

method, and the simplified text of the Model Specifications presented

in this Notice must be received no later than January 30, 1995.

ADDRESSES: Comments regarding the alternate testing procedure and the

simplified text of the Model Specifications 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., SW,

Washington, DC 20590 (Docket hours are from 9:30 a.m. to 4 p.m.).

FOR FURTHER INFORMATION CONTACT: Dr. James F. Frank, Office of Alcohol

and State Programs (OASP), NTS-21, National Highway Safety

Administration, 400 Seventh Street, S.W., Washington, D.C. 20590.

Telephone (202) 366-5593.

SUPPLEMENTARY INFORMATION: On December 14, 1984, the National Highway

Traffic Safety Administration (NHTSA) issued a notice to convert

mandatory standards for evidential breath testers (EBTs) and

calibrating units for breath alcohol testers (CUs) to Model

Specifications for such devices (49 FR 48855 and 49 FR 48865

respectively). The Notices indicated that the Agency would continue to

test EBTs and CUs and would release its findings to provide States

which choose not to conduct their own testing with adequate information

upon which to base their purchasing decisions.

Since publication of the Model Specifications in 1984, States have

been moving toward a lowering of alcohol levels which indicate illegal

alcohol impaired driving and enacting new laws targeting youthful

offenders (i.e., ``zero tolerance'' laws). In addition, the Department

of Transportation (DOT) has been engaged in rulemaking to require

alcohol testing of transportation workers in safety sensitive positions

at low alcohol levels. For these reasons, on September 17, 1993 (58 FR

48705), NHTSA amended the Model Specifications for Devices to measure

breath alcohol by lowering the BAC level at which devices are

evaluated. Under the earlier specifications, evidential breath testers

(EBTs) were evaluated at 0.000, 0.050, 0.101, and 0.151 BAC. The

September 1994 notice amended the Specifications for EBTs to require

testing at 0.000, 0.020, 0.040, 0.080, 0.160 BAC. The Model

Specifications for calibrating units similarly called for evaluations

at 0.050, 0.100, and 0.150 BAC. Today's notice amends the Model

Specifications to perform evaluations of these devices also at 0.020,

0.040, 0.080, and 0.160 BAC. It also changes the acceptable Systematic

Error to not be greater than 0.002 BAC at each alcohol

concentration (rather than 2%), and makes a number of

changes to implement automated procedures for the use of gas

chromatography to assess the precision and accuracy of calibrating

units, rather than the manual procedures previously used and described

in the 1984 Model Specifications.

In addition, today's notice proposes to simplify the Model

Specifications, such as by omitting descriptive details regarding

laboratory procedures, thereby making the Model Specifications shorter

and easier to read. However, the proposed changes do not represent

substantive alterations in the procedures followed or in the criteria

used to determine whether devices meet these Model Specifications.

Today's notice also proposes to include in the Model Specifications an

alternate procedure for evaluating dry gaseous ethanol calibrating

devices. This procedure substitutes National Institute of Standards and

Technology Reference Gas Mixtures (NISTRGMs) in place of wet reference

samples previously used. The proposed revised Model Specifications,

incorporating both the alternate procedures and the simplifications,

are contained in Appendix A to this notice. Comments are sought

regarding the agency's proposal.

The model specifications are not intended to replace the current

qualification programs required in certain States for this equipment or

to directly regulate the manufacturers of CUs. Organizations and

agencies may adopt these model specifications and rely on NHTSA's test

results, or may conduct their own tests according to their own

procedures and specifications. They may wish to make use of this

program in addition to setting their own requirements. While the agency

is not imposing its findings on State and local governments, NHTSA

encourages each State to consider adopting the NHTSA model

specifications as their own. It should be noted that transportation

employers covered by 49 CFR Part 40, Procedures for Transportation

Workplace Drug and Alcohol Testing Programs, are required to use only

CUs that meet the criteria established by that regulation.

Procedures

Testing of CUs submitted by manufacturers to these Model

Specifications will continue to be conducted by the DOT Volpe National

Transportation Systems Center (VNTSC). Procedures for submitting

devices for evaluation have not changed. Tests will continue to be

conducted semi-annually or as necessary. Manufacturers wishing to

submit CUs for testing must apply to NHTSA for a test date (Office of

Alcohol and State Programs, NTS-21, 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, Kendall Square, Cambridge, MA 02142. The manufacturer shall be

responsible for ensuring that the unit is operating properly and is in

proper calibration. If the manufacturer wishes to submit a duplicate,

backup unit, it may do so. In addition to the Operator's Manual and the

Maintenance Manual normally supplied with the purchase of this

equipment, the manufacturer shall deliver to VNTSC specifications and

drawings which fully describe its units. Proprietary information will

be respected. (See 49 CFR Part 512, regarding the procedures by which

NHTSA will consider claims of confidentiality.)

The manufacturer will have the right to check the CU between

arrival in Cambridge and the start of the test, and to ensure that the

CU is in proper working condition but will have no access to it during

the tests. Any malfunction of the CU 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.

On the basis of these results, NHTSA will publish a Conforming

Products List (CPL) identifying the CUs 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 unique design

features of a CU). If these Model Specifications are modified,

notification will be provided in the Federal Register. If NHTSA

determines that re-testing to the modified 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 shall certify that the CPL does, in fact, reflect CUs which

meet the performance criteria set forth in the model specifications.

If at any time a manufacturer changes the design of a CU currently

on the CPL, the manufacturer shall submit the proposed changes to the

Office of Alcohol and State Programs (OASP) for review. Based on this

review, NHTSA will decide whether the change will require retesting of

the unit. Normally, such retesting will be accomplished at the next

testing period. Guidance to manufacturers on considerations governing

this decision are available from NHTSA's OASP upon request.

OASP 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 CUs provide both acceptance

and field performance data to OASP. 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 an instrument on the CPL is

not performing in accordance with the Model Specifications, NHTSA will

direct VNTSC to conduct a special investigation. This study 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 dropped 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 which led to the

dropping of the unit in question from the list.

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 and found to conform to the amended model

specifications herein. The CPL also lists devices that have not yet

been tested at the new (and lower) BAC levels (at 0.020, 0.040, 0.080,

and 0.160), but which were listed on our last CPL for calibrating units

(58 FR 26030) on the basis that they were tested and found to conform

to the previous Model Specifications (at 0.050, 0.100, and 0. 150).

These devices have been identified with an asterisk.

The fact that a unit does not appear on the list, or is not yet

listed as having met the new Model Specifications, does not necessarily

indicate that it failed to meet the Specifications. The unit may not

have been included because it has not yet been tested. As additional

testing is completed, any additional devices that meet the Model

Specifications, as amended, will be included in our next publication of

the CPL for calibrating units.

This publication of the CPL includes a number of changes that

should be noted. (1) The Breath Alcohol Simulator BAS311 manufactured

by Century Systems, Inc., Arkansas City, KS has been removed from the

list. The company has been out of business since the late 1970s and the

devices have not been sold since then. No devices are known to be in

use today. (2) National Draeger, Inc., Pittsburgh, PA now manufactures

the Mark II-A simulator previously made by Smith & Wesson Electronic

Co., Springfield, MA. NHTSA will now list National Draeger as the

manufacturer of the Mark IIA. NHTSA will also continue to list the

Smith & Wesson Mark IIA, as units are still in use in the field.

It should also be noted that Scott Specialty Gases/Scott Medical

Products, Inc., Plumsteadville, PA has submitted for testing Model

EBSTM Gaseous Ethanol Breath Standard, a dry gas standard. This

device has been tested against the proposed Model Specifications, using

the alternate testing procedure proposed in this publication. It has

been found to meet the proposed Specifications. If the alternate

procedure is incorporated into the Model Specifications, as proposed,

this device will be added to the CPL. In addition, A.G. Specialty Gas

(also known as Acetylene Gas Company), St. Louis, MO has submitted for

testing their Ethanol Breath Alcohol Standard. This dry gas device has

also been tested against the proposed Model Specifications, using the

alternate testing procedures described in this publication. It has also

been found to meet the proposed specifications. If the alternate

procedure is incorporated into the Model Specifications, as proposed,

this dry gas device will also be added to the CPL.

Comments

Interested persons are invited to comment on the proposed alternate

testing procedure described in this notice, as well as the simplified

and streamlined text of the proposed Model Specifications. 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.

This action has been analyzed in accordance with the principles and

criteria contained in Executive Order 12612 and it has been determined

that it has no federalism implication that warrants the preparation of

a federalism assessment.

In consideration of the foregoing, NHTSA amends the Model

Specifications for Calibrating Units, as last published in the Federal

Register on December 14, 1984 (49 FR 48865), as set forth below. NHTSA

proposes to further amend these Model Specifications, as set forth in

Appendix A.

Model Specifications for Calibrating Units for Breath Alcohol

Testers

1. Sections 4.2, 4.4 and 4.5.1 are amended by replacing the phrase

``2 percent'' each time it appears after the words ``plus or minus''

with the phrase ``0.002 BAC''.

2. Section 5.1.3.1 is amended by replacing the last sentence with:

``Otherwise, test the calibrating units at concentrations of 0.02,

0.04, 0.08, and 0.16 BAC.''

3. Section 5.1.3.4 is revised to read as follows:

5.1.3.4 Fit the above sealed reference solution bottle with the

transfer line which connects the headspace vapors of the reference

solutions to the GC sampling valve. Fit another hose which allows an

airpump to bubble air at a controlled rate to the bottom of the

reference solution (see Figure 1). Immerse the bottles to the level of

the rubber septa in a circulating water bath maintained at an

appropriate temperature between 25 deg. and 35 deg. C., constant to

within 0.1 deg. C.

4. Section 5.1.3.6 and 5.1.3.7 are deleted and the section numbers

reserved.

5. Sections 5.1.3.8 and 5.1.3.9 are revised to read as follows:

5.1.3.8 By control of the air pump, flush the headspace vapors of

the reference solution through the GC sampling system at a rate just

sufficient to completely flush the sampling system in 10 seconds.

Inject the sample onto the GC column. After injection of the vapor,

obtain the ethanol chromatogram and measure the area of the ethanol

peak.

5.1.3.9 Repeat 5.1.3.8 four more times so as to obtain a total of

5 ethanol chromatograms and peak areas from the first reference sample.

6. Section 5.1.3.10 is amended by deleting the words ``as shown in

Figure 1'' and by replacing the last two sentences with the following

two sentences, ``If the recommended flow rate from the calibration unit

exceeds 200 ml/min., a sample splitting device may be used to reduce

the flow rate into the gas-sampling system. Obtain 10 ethanol

chromatograms and measure each ethanol peak area.''

7. Sections 5.1.3.11 and 5.1.3.18 are revised to read as follows:

5.1.3.11 Using the second reference sample, repeat 5.1.3.8 five

times to obtain 5 ethanol chromatograms and peak areas from the second

reference sample.

5.1.3.18 Repeat 5.1.3.8 to 5.1.3.17 at other concentrations if

necessary to satisfy 5.1.3.1.

8. Section 5.3 is revised by replacing the heading ``AC-Line

voltage test.'' with the heading ``Input Power Tests.''; by renumbering

Sections 5.3.1, 5.3.2, 5.3.3 and 5.3.4 as 5.3.1.1, 5.3.1.2, 5.3.1.3 and

5.3.1.4; by inserting the heading ``AC Input Power.'' in 5.3.1 and by

adding a new 5.3.2 to read as follows:

5.3.2 DC Input Power

5.3.2.1 Apply DC power to the unit under test through a DC power

supply having an output adjustable from 0 to 20 volts and a current

rating as required by the instrument under test.

5.3.2.2 Monitor the DC power supply output voltage on a DC

voltmeter having an accuracy of 2% in the range 11-15

volts.

5.3.2.3 Adjust the voltage at the breath tester calibrating unit

to 11 volts. After at least one-half hour, check the voltage and

readjust to 11 volts if necessary. Then immediately repeat the

precision and accuracy test (5.1 as amended), at a concentration

between 0.02 and 0.16 BAC to obtain the relative standard deviation and

the systematic error.

5.3.2.4 Increase the voltage to 15 volts. After at least one-half

hour, readjust the voltage if necessary and again repeat the precision

and accuracy test (5.1, as amended) at a concentration between 0.02 and

0.16 BAC to obtain the relative standard deviation and the systematic

error.

Authority: 23 U.S.C. 402; delegations of authority at 49 CFR

1.50 and 501.

Issued on: December 23, 1994.

Michael B. Brownlee,

Associate Administrator for Traffic Safety Programs.

Appendix A--Model Specifications for Calibrating Units for Breath

Alcohol Testers

1. 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. Definitions.

2.2 Conformance testing. Testing to check the conformance of a

product with these model specifications in advance of and independent

of any specific procurement action.

2.3 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).1 BrAC is often used to indicate that the measurement is

a breath measurement. In these Model Specifications, concentration

units of test samples are referred to as BAC units and are grams

alcohol per 210 liters of air. See section 3.1.

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

\1\Available from National Committee on Uniform Traffic Laws and

Ordinances, 405 Church Street, Evanston, IL 60201.

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

2.4 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.

Standard deviation=

{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. An indication of the accuracy of the

measurement of the concentration of a succession of vapor samples.

Systematic error={(Xm-test BAC)/test BAC}100 percent

2.6 Least Squares Fit Calibration Curve. A line fitted to a number

of measurement pairs, one the dependent value (in section 3.1 stated

BAC, Y) and the other the independent value (peak height, X), over a

measurement range.

The fitted line is of the form

Y=a+bX,

where

intercept, a=Y-bXm

and

slope, b=(SumXiYi-nXmYm) /

(SumXi\2\-nXm\2\)

3.0 Tests and Requirements

If the BAC of the CU is fixed, perform the tests at the fixed BAC;

otherwise, prepare the CU for testing at 0.08 BAC except as otherwise

required in Test 1. 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 BACSE+0.002 BAC;

RSD2%

Test 1. Precision and Accuracy. Test at each specified BAC.

Test 1.1: 0.020 BAC

Test 1.2. 0.040 BAC

Test 1.3: 0.080 BAC

Test 1.4: 0.160 BAC

Test 2. Ambient Temperature. Use a temperature chamber controllable

to 2oC. 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 readjust the voltage,

if necessary.

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

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 (100 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 2 to prepare

two solutions which, when thermostated at 34 deg.C, produce head space

ethanol vapor concentrations that bracket the test BAC by no more than

20%. Small Air Pump for bubbling air through reference

solutions (see Figure 1).

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

\2\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 and place bottles

in the water bath with water level up to the stopper. Connect air pump

to inlet air hose of reference solution bottle. Connect outlet hose 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 chromatogram 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 chromatogram

using the CU device.

Step 5. Repeat step 3 using the second reference solution.

Step 6. Calculations. Peak height to BAC conversion factor. For

each ethanol peak obtained in step 2 and step 5, calculate a conversion

factor for ethanol concentration by dividing the equivalent BAC 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 BAC for each of the 10

ethanol peaks obtained in step 4. Calculate the mean, the RSD, and the

systematic error of the experimental BACs.

TN29DE94.034

The following alternate method for the evaluation of dry gaseous

ethanol calibration devices is presented.

Additional required material: For the proposed 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 BAC range 0.01 to 0.16.

Alternate Procedure for evaluation of dry gaseous ethanol

calibration devices. If it is evident that the gas chromatograph column

being used according to the model specifications described above is

causing an excessive systematic error in comparing samples from dry

gaseous ethanol calibrating devices with the moist head space vapor

samples from the thermostated reference solutions, the following

alternate procedure may be used. This procedure substitutes National

Institute for Standards and Technology Reference Gas Mixtures (NISTRGM)

in place of the moist reference samples.

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

chromatogram 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

BAC. 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

chromatogram 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 BAC

for each of the 10 peak heights. Calculate the mean, RSD, and

systematic error of the calculated BACs.

Appendix B--Conforming Products List of Calibrating Units for

Breath Alcohol Testers [Manufacturer and Calibrating Unit].

1. CMI, Inc., Owensboro, KY:

Toxitest II

2. Federal Signal Corporation, CMI, Inc., Minturn, CO:

Toxitest Model ABS120*

3. Guth Laboratories, Inc., Harrisburg, PA:

Model 34C Simulator3

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

\3\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 34C-FM; and 34C-

NPAS.

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

Model 3412

Model 10-4

Model 1214

4. Intoximeters, Inc., St. Louis, MO:

Alco Breath Alcohol Standard*

5. Luckey Laboratories, Inc., San Bernardino, CA:

Simulator*

6. National Draeger, Inc., Pittsburgh, PA:

Mark II-A

7. PLD of Florida, Inc., Rockledge, FL:

BA 500

8. Protection Devices, Inc./U.S. Alcohol Testing, Inc., Rancho

Cucamonga, CA:

LS34 Model 6100*

9. Repco Marketing, Inc., Raleigh, NC:

AS-1

10. Smith & Wesson Electronic Co., Springfield, MA:

Mark II-A Simulator*

11. Systems Innovation, Inc., Hallsteaed, PA:

True-Test MD 901*

12. 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 amended model specifications detailed in this

notice.

[FR Doc. 94-32091 Filed 12-23-94; 12:34 pm]

BILLING CODE 4910-59-P