Temperature-Indicating Devices; Thermally Processed Low-Acid Foods Packaged in Hermetically Sealed Containers, 11892-11924 [2011-4475]

Agencies

• Food and Drug Administration
• DEPARTMENT OF HEALTH AND HUMAN SERVICES

[Federal Register Volume 76, Number 42 (Thursday, March 3, 2011)]
[Rules and Regulations]
[Pages 11892-11924]
From the Federal Register Online via the Government Printing Office [www.gpo.gov]
[FR Doc No: 2011-4475]



[[Page 11891]]

Vol. 76

Thursday,

No. 42

March 3, 2011

Part III





Department of Health and Human Services





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



Food and Drug Administration



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



21 CFR Part 113



Temperature-Indicating Devices; Thermally Processed Low-Acid Foods 
Packaged in Hermetically Sealed Containers; Final Rule

Federal Register / Vol. 76 , No. 42 / Thursday, March 3, 2011 / Rules 
and Regulations

[[Page 11892]]


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

DEPARTMENT OF HEALTH AND HUMAN SERVICES

Food and Drug Administration

21 CFR Part 113

[Docket No. FDA-2007-N-0265; Formerly Docket No. 2007P-0026]


Temperature-Indicating Devices; Thermally Processed Low-Acid 
Foods Packaged in Hermetically Sealed Containers

AGENCY: Food and Drug Administration, HHS.

ACTION: Final rule.

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

SUMMARY: The Food and Drug Administration (FDA) is amending its 
regulations for thermally processed low-acid foods packaged in 
hermetically sealed containers to allow for use of other temperature-
indicating devices, in addition to mercury-in-glass thermometers, 
during processing. This final rule also establishes recordkeeping 
requirements relating to temperature-indicating devices and reference 
devices maintained by the processor and allows for the use of advanced 
technology for measuring and recording temperatures during processing. 
Finally, this final rule includes metric equivalents of avoirdupois 
(U.S.) measurements where appropriate. This final rule will allow low-
acid canned food processors to transition from mercury-in-glass 
thermometers to alternative temperature-indicating devices. Use of 
temperature-indicating devices that do not contain mercury will 
eliminate concerns about potential contamination of the food or the 
processing environment from broken mercury-in-glass thermometers. 
Elsewhere in this issue of the Federal Register, FDA is publishing a 
30-day notice announcing that it has submitted the information 
collection provisions of this final rule to the Office of Management 
and Budget (OMB) for review and clearance under the Paperwork Reduction 
Act of 1995 (the PRA). The notice also invites the public to submit 
comments on the information provisions to OMB. Prior to the effective 
date of this final rule, FDA will publish a notice in the Federal 
Register announcing OMB's decision to approve, modify, or disapprove 
the information collection provisions of the final rule.

DATES: This final rule is effective March 5, 2012.

FOR FURTHER INFORMATION CONTACT: Mischelle B. Ledet, Center for Food 
Safety and Applied Nutrition (HFS-615), Food and Drug Administration, 
5100 Paint Branch Pkwy., College Park, MD 20740, 301-436-2070.

SUPPLEMENTARY INFORMATION:

Table of Contents

I. Background
II. Comments on the Proposed Rule
III. Minor Revisions in Regulations
IV. Analysis of Economic Impacts
    A. Final Regulatory Impact Analysis
    B. Regulatory Flexibility Analysis
    C. Unfunded Mandate Analysis
V. Paperwork Reduction Act of 1995
VI. Federalism
VII. References

I. Background

    In the Federal Register of March 14, 2007 (72 FR 11990), FDA 
published a proposed rule entitled ``Temperature-Indicating Devices; 
Thermally Processed Low-Acid Foods Packaged in Hermetically Sealed 
Containers'' (the proposed rule). We proposed to revise Sec.  113.40 
(21 CFR 113.40) to provide for use of temperature-indicating devices 
that accurately indicate the temperature during processing. We proposed 
that temperature-indicating devices shall be tested for accuracy 
against an accurate calibrated reference device upon installation and 
at least once a year thereafter, or more frequently if necessary, to 
ensure accuracy during processing. We also proposed that the design of 
the temperature-indicating device shall ensure that the accuracy of the 
device is not affected by electromagnetic interference and 
environmental conditions.
    We proposed to require that each temperature-indicating device have 
a tag, seal, or other means of identity that will be used by the 
processor to identify the temperature-indicating device, and that each 
reference device have a tag, seal, or other means of identity that will 
be used by the processor to identify the reference device. We proposed 
the establishment and maintenance of written records to document the 
accuracy for each temperature-indicating device and each reference or 
standard device.
    We also proposed to provide for the use of metric equivalents of 
avoirdupois (U.S.) measurements for temperature-indicating devices, to 
provide for use of temperature-recording devices that create analog, 
graphical, or digital recordings, and to clarify various operational 
and record requirements of the regulations.
    In the preamble to the proposed rule, FDA stated that, pending 
issuance of a final rule, we intended to consider the exercise of our 
enforcement discretion on a case-by-case basis when processors of low-
acid canned food elect to replace mercury-in-glass thermometers with 
alternative temperature-indicating devices in a manner that was 
consistent with the proposed rule (72 FR 11990 at 11999, March 14, 
2007). The Federal Food, Drug, and Cosmetic Act's (the FD&C Act) 
enforcement provisions commit complete discretion to the Secretary of 
Health and Human Services (and by delegation to FDA) to decide how and 
when they should be exercised (see Heckler v. Chaney, 470 U.S. 821, 835 
(1985); see also Schering Corp. v. Heckler, 779 F.2d 683, 685-86 (DC 
Cir. 1985) (stating that the provisions of the act ``authorize, but do 
not compel FDA to undertake enforcement activity'')). FDA will continue 
to consider the exercise of our enforcement discretion on a case-by-
case basis when processors of low-acid canned food elect to replace 
mercury-in-glass thermometers with alternative temperature-indicating 
devices in a manner that is consistent with the proposed rule until the 
effective date of the final rule. In addition, we will consider the 
exercise of our enforcement discretion on a case-by-case basis for 
processors who comply with the provisions of this final rule prior to 
the effective date. All low-acid canned food processors must comply 
with the requirements of this final rule on and after the effective 
date.

II. Comments on the Proposed Rule

    FDA received six letters, each containing one or more comments, to 
the proposed rule. The comments were from industry, a trade 
association, and individuals. Most of the letters generally supported 
the proposed rule, but provided some comments that suggested 
modifications to the proposed rule. Some of the comments addressed 
issues outside the scope of this rulemaking and will not be addressed 
in this document. A summary of the comments and FDA's responses 
follows.
    (Comment 1) One comment requested that the effective date of this 
final rule be not less than 1 year from the date of publication. The 
comment indicated that companies that are continuing to use mercury-in-
glass thermometers will need time to comply with the additional 
recordkeeping requirements for accuracy checks. Furthermore, companies 
with existing water retorts will need at least 1 year to comply with 
the additional equipment requirements of the regulation. The comment 
also indicated that firms that currently reprocess products or rework 
previously processed product into a new formulation need at least 1 
year to review existing process schedules and conduct confirmatory 
testing if necessary, to comply with Sec.  113.83 (21 CFR 113.83).

[[Page 11893]]

    (Response) We agree with the comment's request to allow 1 year for 
processors to comply with recordkeeping requirements relating to use of 
mercury-in-glass thermometers and to other requirements relating to 
temperature-indicating devices established in this final rule. Thus, 
the effective date of this final rule is 1 year from the date of 
publication in the Federal Register. However, FDA does not agree with 
the comment's suggestion that processors need a year to comply with 
Sec.  113.83 for reprocessed or reworked product. As discussed in our 
response to comment 38, although we clarified the requirements in final 
Sec.  113.83, we did not propose new requirements for reprocessed or 
reworked products in the proposed rule or establish new requirements 
for reprocessed or reworked products in this final rule.
    (Comment 2) One comment recommended defining the term 
``temperature-indicating device'' as the entire system, including the 
sensor(s) and the temperature-indicating device display. The comment 
noted that separate references to the ``temperature-indicating device'' 
and the ``sensor of the temperature-indicating device'' could be 
interpreted to mean that the sensor is not part of the temperature-
indicating device and thus does not have to be calibrated. The comment 
suggested using the term ``temperature-indicating device display'' to 
refer to the electronics/display portion only and to define 
``temperature-indicating device'' to mean the entire system.
    (Response) We agree that the term ``temperature-indicating device'' 
includes the temperature-indicating device sensor and the temperature-
indicating device display. Accordingly, we revised the proposed 
requirements to clarify that each temperature-indicating device must 
have a sensor and a display (final Sec.  113.40(a)(1), (b)(1), (c)(1), 
(d)(1), (e)(1), (f)(1), and (g)(1)(i)(A)). As appropriate, we replaced 
the terms ``sensors of temperature-indicating devices'' and ``sensor of 
the temperature-indicating device'' with ``temperature-indicating 
device sensor'' (final Sec.  113.40(a)(1)(v), (b)(1)(v), (c)(1)(v), 
(d)(1)(v), and (e)(1)(v)). In final Sec.  113.40(f)(1)(v), we clarified 
that the temperature-indicating device sensor, rather than the 
temperature-indicating device, must be located in the steam dome near 
the steam water interface or, when applicable, in each hydrostatic 
water leg.
    Although the comment did not request similar clarification for 
temperature-recording devices, in this final rule we also clarified 
that each temperature-recording device must have a sensor and a 
mechanism for recording temperatures to a permanent record, such as a 
temperature-recording chart (final Sec.  113.40(a)(2), (b)(2), (c)(2), 
(d)(2), (e)(2), (f)(2), and (g)(1)(i)(B)).
    (Comment 3) One comment indicated that the mercury-in-glass 
thermometer originally was used for three important reasons, i.e., 
permanent accuracy, no drift over time, and reliability. According to 
the comment, reliability means ``it works or it doesn't work and you 
know when it doesn't work.'' The comment suggested that these factors 
should be characteristics of any alternative temperature-indicating 
device. Another comment suggested revising proposed Sec.  113.40(a)(1) 
to require alternative temperature-indicating devices to meet or exceed 
the accuracy and reliability of mercury-in-glass thermometers.
    (Response) The Agency recognizes that accuracy, drift, and 
reliability are important considerations for any temperature-indicating 
device. However, the comment does not specify any unique problems that 
may be associated with these factors that were not addressed by the 
proposed codified language. Thus, the Agency is not making any changes 
to the proposed codified in response to this comment.
    The comment's reference to ``permanent accuracy'' is not clear. 
Perpetual and unfailing accuracy cannot be guaranteed for any 
temperature-indicating device, including mercury-in-glass thermometers. 
Each temperature-indicating device must be tested for accuracy, as 
required in final Sec.  113.40(a)(1), (b)(1), (c)(1), (d)(1), (e)(1), 
(f)(1), and (g)(1)(i)(A) of this final rule. A temperature-indicating 
device that is defective or cannot be adjusted to the accurate 
calibrated reference device must be repaired before further use or 
replaced (final Sec.  113.40(a)(1)(iii), (b)(1)(iii), (c)(1)(iii), 
(d)(1)(iii), (e)(1)(iii), (f)(1)(iii), and (g)(1)(i)(A)(3)).
    We use the terms ``accurate'' and ``accuracy'' in this final rule 
to refer to ``measurement accuracy.'' Measurement accuracy is defined 
in the International Vocabulary of Metrology as ``closeness of 
agreement between a measured quantity value and a true quantity value 
of a measurand'' (Ref. 1). For a temperature-indicating device, the 
temperature shown on the display is the ``measured quantity value'' and 
the actual or true temperature is the ``true quantity value.'' As 
discussed in our response to Comment 9, this final rule provides that 
the measurement accuracy for a temperature-indicating device must be 
within 1 [deg]F (0.5 [deg]C) of the true quantity value, i.e., the 
temperature-indicating device must be accurate to 1 [deg]F (0.5 [deg]C) 
(final Sec.  113.40(a)(1)(iv), (b)(1)(iv), (c)(1)(iv), (d)(1)(iv), 
(e)(1)(iv), (f)(1)(iv), and (g)(1)(i)(A)(4)).
    We agree that ``drift over time'' is a factor that must be 
considered to assure that the temperature-indicating device is accurate 
during processing. However, because an absolute requirement for no 
drift over time may prevent use of an otherwise appropriate 
temperature-indicating device, we do not agree that this characteristic 
should be specified in this final rule. We believe the requirement of 
this final rule for the temperature-indicating device to be accurate 
encompasses considerations relating to drift. If the accuracy of the 
temperature-indicating device may be affected by drift, it is our 
expectation that an appropriate calibration interval (i.e., more 
frequently than once per year) or other appropriate mechanism will be 
established by the processor to ensure that the temperature-indicating 
device is accurate during processing.
    The reliability of a temperature-indicating device is determined 
based on evaluation of past performance of the specific temperature-
indicating device or similar temperature-indicating devices. Past 
performance may be used as an indicator, but not as an absolute 
guarantee or predictor, of future performance. Although we agree that 
warranties and predictions of reliability are important considerations 
for processors when choosing a temperature-indicating device, they do 
not ensure accuracy during processing or alleviate the processors' 
responsibility to ensure that the temperature-indicating device 
provides an accurate temperature reading during processing (final Sec.  
113.40(a)(1), (b)(1), (c)(1), (d)(1), (e)(1), (f)(1), and 
(g)(1)(i)(A)). A temperature-indicating device that does not accurately 
indicate the temperature during processing does not comply with the 
requirements of this final rule.
    We believe that the requirement in this final rule for the 
temperature-indicating device to accurately indicate the temperature 
during processing (final Sec.  113.40(a)(1), (b)(1), (c)(1), (d)(1), 
(e)(1), (f)(1), and (g)(1)(i)(A) is adequate to ensure the accuracy and 
reliability of the temperature-indicating device, and that it is not 
necessary to revise the regulation to require that alternate 
temperature-indicating devices meet or exceed the accuracy and 
reliability of mercury-in-glass thermometers, as suggested by the 
comment.
    (Comment 4) One comment recommended revising proposed Sec.  
113.40(a)(1) to require temperature-indicating devices to be tested for

[[Page 11894]]

accuracy against a reference device for which the accuracy is traceable 
to a National Institute of Standards and Technology (NIST), or 
equivalent, standard reference device.
    (Response) We agree with the comment. We revised the applicable 
proposed requirements to clarify that each temperature-indicating 
device and each reference device that is maintained by the processor 
must be tested for accuracy against a reference device for which the 
accuracy is traceable to a NIST, or other national metrology institute, 
standard reference device (final Sec.  113.40(a)(1), (b)(1), (c)(1), 
(d)(1), (e)(1), (f)(1), and (g)(1)(i)(A)). The term ``reference device 
maintained by the processor'' refers to the reference device used by a 
processor who performs the accuracy tests at the processor's own 
facility or facility laboratory. For such reference device, the 
processor, rather than a third party laboratory, is responsible for 
ensuring accuracy of the reference device when it is used for the 
accuracy test and for ensuring that its accuracy is traceable to a 
NIST, or other national metrology institute, standard reference device. 
The term ``traceable'' refers to ``metrological traceability,'' which 
is defined in the International Vocabulary of Metrology as the 
``property of a measurement result whereby the result can be related to 
a reference through a documented unbroken chain of calibrations, each 
contributing to the measurement uncertainty'' (Ref. 2). ``Measurement 
result'' is defined as a ``set of quantity values being attributed to a 
measurand together with any other available relevant information'' 
(Ref. 3) and ``measurement uncertainty'' is defined as ``the non-
negative parameter characterizing the dispersion of the quantity values 
being attributed to a measurand, based on the information used'' (Ref. 
4).
    This final rule also clarifies that the record of the accuracy test 
for a temperature-indicating device or a reference device maintained by 
the processor must include documentation of the traceability of the 
accuracy of the reference device to a NIST, or other national metrology 
institute, standard reference device (final Sec.  113.100(c) and (d) 
(21 CFR 113.100(c) and (d))). For an accuracy test performed by the 
processor and, thus, for which the processor maintains the reference 
device, the documentation of traceability must be a guarantee, 
certificate of accuracy, certificate of calibration, or other document 
from the manufacturer or other source of the reference device. For an 
accuracy test performed by an outside facility, the documentation of 
traceability must be a guarantee, certificate of accuracy, certificate 
of calibration, or other document from the facility that includes a 
statement or other documentation regarding the traceability of the 
accuracy to a NIST, or other national metrology institute, standard 
reference device.
    The information required to be included in the records of accuracy 
for temperature-indicating devices and reference devices was set forth 
in proposed Sec.  113.40(a)(1)(ii), (b)(1)(ii), (c)(1)(ii), (d)(1)(ii), 
(e)(1)(ii), (f)(1)(ii), and (g)(1)(i)(A)(2). To eliminate redundancy, 
we moved the information requirements for the records of accuracy for 
temperature-indicating devices and reference devices maintained by the 
processor from each of these sections to final Sec.  113.100(c) and (d) 
of Subpart F--Records and Reports. We redesignated proposed Sec.  
113.100(c), (d), and (e), as final Sec.  113.100(e), (f), and (g), 
respectively. We also revised proposed Sec.  113.87(c) (21 CFR 
113.87(c)) to clarify that the records of accuracy tests for 
temperature-indicating devices used to determine the initial product 
temperature and reference devices maintained by the processor must be 
maintained in accordance with Sec.  113.100(c) and (d).
    (Comment 5) One comment expressed concern about the proposed 
requirement that the design of the temperature-indicating device ensure 
that the accuracy of the device is not affected by electromagnetic 
interference and environmental conditions (proposed Sec.  
113.40(a)(1)(i), (b)(1)(i), (c)(1)(i), (d)(1)(i), (e)(1)(i), (f)(1)(i), 
and (g)(1)(i)(A)(1)). According to the comment, the proposed language 
focuses on only a few of the considerations that a processor must take 
into account when selecting a temperature-indicating device and the 
considerations in the proposed language may not be applicable to future 
temperature-indicating technologies. The comment pointed out that a 
temperature-indicating device that is very robust in terms of the 
electromagnetic interference and environmental conditions could provide 
unreliable temperature readings because of other aspects of the design 
and installation. However, a temperature-indicating device that is less 
robust in terms of electromagnetic interference and environmental 
conditions could provide reliable and accurate readings due to good 
design and installation practices. The comment stated that the end goal 
of any temperature-indicating device is reliable and accurate readings. 
The comment suggested that it would be more effective to state that: 
``The design, installation, and operation of the temperature-indicating 
device shall be such that the accuracy and reliability of the device is 
ensured.''
    (Response) We do not agree that the language recommended by the 
comment provides clarity or value to the regulation. The requirements 
in the regulation for the temperature-indicating device to be accurate 
upon installation and during processing (final Sec.  113.40(a)(1), 
(b)(1), (c)(1), (d)(1), (e)(1), (f)(1), and (g)(1)(i)(A)) encompass 
design, installation, operation, and reliability considerations 
traditionally associated with mercury-in-glass thermometers and that 
must be considered for other temperature-indicating devices. However, 
we believe it is necessary to emphasize in this final rule that the 
design of the temperature-indicating device must assure that accuracy 
is not affected by electromagnetic interference and environmental 
conditions because these factors are not traditionally associated with 
mercury-in-glass thermometers. As discussed in the preamble to the 
proposed final rule, although electromagnetic energy does not affect 
the accuracy of mercury-in-glass thermometers, temperature-indicating 
devices with electronic or electromagnetic components are vulnerable to 
electromagnetic interference. Electromagnetic energy may vary in the 
area where a temperature-indicating device is located as electronics 
are turned on and off, introduced into, and removed from the area. 
Electromagnetic energy exposure may also vary when a temperature-
indicating device is moved from one location to another, e.g., from one 
retort to another. Thus, unlike a mercury-in-glass thermometer, a 
temperature-indicating device that may be affected by electromagnetic 
energy must be designed based on consideration of that factor, i.e., 
the temperature-indicating device must be designed to ensure that its 
accuracy during processing is not compromised by exposure to 
electronics that generate or cause fluctuations in electromagnetic 
energy. Similarly, some environmental conditions, such as humidity, 
vibrations, and air pressure, that do not affect the accuracy or 
performance of mercury-in-glass thermometers must be considered and 
addressed in the design of other temperature-indicating devices.
    (Comment 6) One comment objected to the proposed requirement that 
the design of the temperature-indicating device ensure that accuracy is 
not affected by environmental conditions because it does not clearly 
state which

[[Page 11895]]

environmental conditions are important and which are not (proposed 
Sec.  113.40(a)(1)(i), (b)(1)(i), (c)(1)(i), (d)(1)(i), (e)(1)(i), 
(f)(1)(i), and (g)(1)(i)(A)(1)). The comment expressed concern that 
some important environmental factors may not be adequately considered. 
The comment noted that there is a difference between environmental 
considerations for mechanical and electronic instruments. According to 
the comment, moisture is an important environmental concern with 
electronic instruments. The comment noted that condensation on a 
computer board or wiring terminals can be detrimental to making a 
measurement and can cause errors. The comment suggested requiring the 
use of temperature-indicating devices with an Ingress Protection code 
suitable for the environment. The comment also indicated concern about 
ambient temperature and vibration, either or both of which may affect 
some electronic and mechanical technologies. According to the comment, 
the ambient temperature coefficient, which is usually expressed as 
degrees of error per degree of change from a specified ambient 
temperature, may not be specified for some temperature-indicating 
devices. The comment expressed concern that most users will not have 
the ability to evaluate the impact of ambient temperature and may not 
be aware that the ambient temperature coefficient is important. The 
comment emphasized that design and installation are essential 
components in vibration resistance.
    (Response) Processors are responsible for ensuring that 
environmental factors, including those expressed in the comment, are 
adequately considered. Processors must use temperature-indicating 
devices appropriate for the processing environment and take appropriate 
steps to evaluate environmental factors that may affect the accuracy of 
the temperature-indicating device. Processors who do not have specific 
expertise for evaluating the effect of environmental factors on 
temperature-indicating devices may need to obtain advice from a 
thermometry expert or obtain a manufacturer's guaranty or warranty 
regarding use of a specific temperature-indicating device in their 
specific food processing environment.
    (Comment 7) One comment requested clarification of proposed Sec.  
113.40(a)(1)(i), which requires that the design of the temperature-
indicating device ensure that the accuracy of the device is not 
affected by electromagnetic interference and environmental conditions. 
The comment questioned whether mechanical thermometers are exempt from 
this requirement. The comment stated that most processors will have no 
way to determine the effects of electromagnetic interference on an 
electronic thermometer design. The comment suggested that the 
regulation should state that temperature-indicating devices should 
comply with an electromagnetic interference standard that is current at 
the time they are designed. According to the comment, this would 
eliminate issues associated with changes to standards that make 
existing temperature-indicating devices noncompliant. The comment 
suggested that temperature-indicating devices should comply with the 
European standards EN 61326-1:2006 Electrical equipment for 
measurement, control and laboratory use; EN 61000-4-2 Personnel 
Electrostatic Discharge Immunity; EN 61000-4-3 Electromagnetic 
compatibility (EMC); and EN 61000-4-6 Conducted disturbances immunity.
    (Response) This final rule does not exempt mechanical thermometers, 
e.g., mercury-in-glass thermometers, from the requirement that the 
design ensure that accuracy is not affected by electromagnetic 
interference and environmental conditions. However, although the 
accuracy of mechanical thermometers may be affected by environmental 
conditions, they generally are not susceptible to the affects of 
electromagnetic interference as are electronic devices.
    FDA is providing flexibility to processors with respect to this 
requirement and is not limiting processors to specific standards with 
which they must comply. Processors, in conjunction with temperature-
indicating device manufacturers and appropriate thermometry experts, 
should ensure that the temperature-indicating devices that processors 
use are accurate during processing. A processor may elect to use an 
appropriate electronic standard, such as those established by the 
European Union, to ensure compliance with final Sec.  113.40(a)(1)(i), 
(b)(1)(i), (c)(1)(i), (d)(1)(i), (e)(1)(i), (f)(1)(i), and 
(g)(1)(i)(A)(1).
    (Comment 8) One comment stated that electronic thermometers are not 
capable of communicating that there is an accuracy problem. The comment 
stated that it is risky to rely on the history of calibration to prove 
an instrument's accuracy because the temperature-indicating device may 
perform properly for years and then fail without warning. The comment 
pointed out that a failure that occurs between calibration cycles may 
not be detected for a significant period of time. The comment suggested 
that additional features are needed to ensure that a temperature-
indicating device retains its accuracy, will not drift, and will report 
any potential errors. The comment indicated that a system with internal 
diagnostics and error reporting to the operator would be one way of 
providing this evidence. The comment suggested that FDA require that an 
electronic temperature-indicating device incorporate technology to 
alert the operator of measurement errors.
    (Response) Processors must ensure that temperature-indicating 
devices are accurate during processing (final Sec.  113.40(a)(1), 
(b)(1), (c)(1), (d)(1), (e)(1), (f)(1), and (g)(1)(i)(A)). Processors 
must test the temperature-indicating device for accuracy upon 
installation and at least once per year thereafter, or more frequently 
if necessary, to ensure accuracy (emphasis added) (see, e.g., final 
Sec.  113.40(a)(1)). These requirements for accuracy for all 
temperature-indicating devices make it unnecessary for this final rule 
to require specific mechanisms to alert the operator of measurement 
errors. Processors should adopt whatever features or systems are 
appropriate to ensure the accuracy of a given temperature-indicating 
device, and to detect defects or failures that may cause a temperature-
indicating device to be inaccurate. For mercury-in-glass thermometers, 
the process for detecting failure may include periodic visual 
examinations and appropriate followup based on findings of defects or 
potential for failure. Electronic devices may have hardware and 
software components with built-in diagnostic and alarm features. 
Processors also may use backup or duplicate devices to detect defects 
or failures. In addition, when adjustments are made to the temperature-
recording device so that it agrees as nearly as possible with, but to 
be in no event higher than, the temperature-indicating device during 
the process time, as required by this regulation (final Sec.  
113.40(a)(2)(iii), (b)(2)(iii), (c)(2)(iii), (d)(2)(iii), (e)(2)(iii), 
(f)(2)(iii), and (g)(1)(i)(B)(3)), the need for such adjustment may be 
used as a signal for determining whether a temperature-indicating 
device failure occurred. Thus, features or systems for ensuring 
accuracy or for detecting inaccuracies may be different for different 
types of temperature-indicating devices, as well as subject to 
technological advancements that we may not anticipate at this time. To 
ensure processors have flexibility to adopt future technologies to 
detect

[[Page 11896]]

defects or failures of temperature-indicating devices, we have not 
required in this final rule specific features or systems to detect such 
defects or failures.
    (Comment 9) One comment expressed concern that the proposed rule 
did not mention measurement uncertainties or test accuracy ratio, which 
are essential parameters for assuring an accurate calibration that are 
specified in standards issued by the American National Standards 
Institute (ANSI) and the International Organization for Standardization 
(ISO) for certification of calibration laboratories. The comment stated 
that the ANSI and ISO standards provide a limit for measurement 
uncertainty and establish a minimum test accuracy ratio that is 
commonly used by calibration facilities. According to the comment, 
although the proposed rule requires use of a calibrated accurate 
reference device, the lack of specific calibration parameters may lead 
to inaccurate calibrations for temperature-indicating devices.
    (Response) Measurement uncertainty is inherent in the proposed 
requirement that the temperature-indicating device be easily readable 
to 1 [deg]F (0.5 [deg]C), (i.e., the dispersion of the quantity values 
for the temperature must be within 1 [deg]F (0.5 [deg]C) of the actual 
temperature) (proposed Sec.  113.40(a)(1)(iv), (b)(1)(iv), (c)(1)(iv), 
(d)(1)(iv), (e)(1)(iv), (f)(1)(iv), and (g)(1)(i)(A)(4)). However, we 
acknowledge that the term ``easily readable'' is readily understood for 
a mercury-in-glass thermometer, which has a visible scale of 
temperature gradations, but it may not be clear for other temperature-
indicating devices, such as those that display a digital reading of the 
temperature. Therefore, we removed the term ``easily readable'' and 
clarified in this final rule that a temperature-indicating device must 
be accurate to 1 [deg]F (0.5 [deg]C) (final Sec.  113.40(a)(1)(iv), 
(b)(1)(iv), (c)(1)(iv), (d)(1)(iv), (e)(1)(iv), (f)(1)(iv), and 
(g)(1)(i)(A)(4)).
    We do not agree that the regulations should specify calibration 
parameters, such as those relating to measurement uncertainties or test 
accuracy ratio, or require use of specific calibration standards, such 
as the ANSI and ISO standards suggested by the comment. Metrology 
authorities, in addition to ANSI and ISO, issue calibration standards, 
which may be revised or replaced. It would be impractical for FDA to 
maintain in the regulations a current list of acceptable calibration 
standards. Processors are responsible for ensuring that the 
temperature-indicating device is accurate during processing and for 
testing each temperature-indicating device for accuracy against a 
reference device for which the accuracy is traceable to a NIST, or 
other national metrology institute, standard reference device by 
appropriate standard procedures, upon installation and at least once a 
year thereafter, or more frequently if necessary (final Sec.  
113.40(a)(1), (b)(1), (c)(1), (d)(1), (e)(1), (f)(1), and 
(g)(1)(i)(A)). Thus, processors are responsible for ensuring that 
accuracy tests are performed by appropriate standard procedures or by 
calibration facilities that use appropriate standard procedures.
    (Comment 10) One comment recommended revising proposed Sec.  
113.40(a)(1) to clarify that the identity of each temperature-
indicating device and reference device must be ``unique.''
    (Response) We do not agree that the term ``unique'' is necessary 
because each temperature-indicating device and each reference device 
that is maintained by the processor must have a tag, seal, or other 
means of identity (final Sec.  113.40(a)(1), (b)(1), (c)(1), (d)(1), 
(e)(1), (f)(1), and (g)(1)(i)(A)). The purpose of a tag, seal, or other 
means of identity is, in part, to uniquely identify each temperature-
indicating device and each reference device that is maintained by the 
processor so that one temperature-indicating or reference device can be 
distinguished from another and so that appropriate records can be 
associated with each temperature-indicating device or reference device.
    (Comment 11) One comment expressed concern about the information 
required in proposed Sec.  113.40(a)(1)(ii)(A) and (a)(1)(ii)(B) for 
documentation of accuracy of temperature-indicating devices and 
reference devices. The comment suggested that the final rule should 
instead require documentation that conforms to the standards 
established by the American National Standards Institute, National 
Conference of Standards Laboratories (ANSI/NCSL) or the International 
Organization for Standardization, International Electrotechnical 
Commission (ISO/IEC) for accrediting calibration laboratories. The 
comment stated that the laboratory accreditation standards indicate 
acceptable reporting practices. The comment acknowledged that the 
standards may be too prescriptive for food processors who perform their 
own calibrations.
    (Response) We do not agree that the regulation should require the 
documentation of accuracy of temperature-indicating devices and 
reference devices to conform to the standards specified in the comment 
for accrediting calibration laboratories. Although FDA supports use of 
accredited calibration laboratories and recognizes that the 
laboratories must maintain certain documentation for the accreditation, 
the records required by this final rule are appropriately limited to 
those necessary to document that the temperature-indicating device was 
tested for accuracy at sufficient frequency to ensure accuracy during 
processing. As acknowledged by the comment, a requirement for 
processors to adhere to accreditation standards would impose an 
unnecessary burden on those who successfully perform their own 
calibrations but are not accredited by ANSI/NCSL or ISO/IEC.
    (Comment 12) One comment recommended revising proposed Sec.  
113.40(a)(1)(ii)(A) and (a)(1)(ii)(B) to require that documentation of 
the results of the accuracy test include before and after data, i.e., 
the temperature reading of the temperature-indicating device compared 
to the accurate calibrated reference device, before and after the 
calibration. The comment indicated that the before data is needed 
because it is the basis for determining whether the device was accurate 
at the time of calibration and for documenting any adjustment that was 
made.
    (Response) Proposed Sec.  113.40(a)(1)(ii)(A) and (a)(1)(ii)(B) 
require that the results of each accuracy test be documented. Although 
not explicitly stated in the proposed rule, we would expect 
documentation of the results of the accuracy test to include 
information about the amount of calibration adjustment that was 
necessary. The ``before and after data'' suggested by the comment would 
be reflected in the amount of calibration adjustment. The amount of 
calibration adjustment is an indication of whether the temperature-
indicating device was accurate at the time of the calibration. If an 
adjustment is required, the processor should evaluate the need for more 
frequent accuracy tests and also determine whether food processed prior 
to the adjustment is under processed. To provide clarity in the 
regulation regarding the requirement to record the amount of 
calibration adjustment that was necessary for a temperature-indicating 
device, we are revising final Sec.  113.100 ``Processing and production 
records'' to indicate that the record of each accuracy test for each 
temperature-indicating device and for each reference device that is 
maintained by the processor must include the results of each accuracy 
test, including the amount of calibration adjustment (final Sec.  
113.100(c)(5) and (d)(5)).

[[Page 11897]]

    Other information relating to the results of the accuracy test that 
should be recorded when it is relevant includes information about the 
condition of the temperature-indicating device (i.e., intact or broken 
mercury column, worn or broken components) and disposition of the 
temperature-indicating device if it cannot be calibrated (i.e., 
destroyed, repaired, or replaced).
    (Comment 13) One comment addressed the proposed requirement that 
records of the accuracy test for the temperature-indicating device 
include the date of the next scheduled accuracy test (proposed Sec.  
113.40(a)(1)(ii)(A), (b)(1)(ii)(A), (c)(1)(ii)(A), (d)(1)(ii)(A), 
(e)(1)(ii)(A), (f)(1)(ii)(A), and (g)(1)(i)(A)(2)(i)). One comment 
interpreted this requirement to imply that the test must be conducted 
on that specific date. The comment suggested removing the requirement 
or changing the language to ``the date of the calibration expiration.''
    (Response) We acknowledge that the proposed requirement concerning 
the date of the next scheduled accuracy test may be misinterpreted to 
mean that the next accuracy test must be conducted on that specific 
date. However, we do not agree that the revised language recommended by 
the comment, i.e., the date of the calibration expiration, adequately 
clarifies that the next accuracy test must be conducted on or before 
the specified date. In this final rule, we require that the record of 
accuracy for a temperature-indicating device and a reference device 
maintained by the processor include the date on or before which the 
next accuracy test must be performed (final Sec.  113.100(c)(6) and 
(d)(6)).
    (Comment 14) One comment recommended placing on each temperature-
indicating device a calibration sticker that indicates the date of the 
last calibration and the date the next calibration is due. According to 
the comment, the calibration standard ISO/IEC 17025 does not require 
the calibration due date to be recorded on the certificate issued by 
the calibration facility, which may have no knowledge of the 
calibration interval for the specific device.
    (Response) We recognize that outside calibration facilities are not 
responsible for determining the frequency of the accuracy tests for 
temperature-indicating devices and, thus, are not required to record 
the frequency on a calibration certificate. We do not agree with the 
comment's recommendation to require a sticker on each temperature-
indicating device with the date of the last calibration and date the 
next calibration is due. Although we do not object to processors using 
stickers or similar mechanisms on temperature-indicating devices to 
emphasize when the next accuracy test for a temperature-indicating 
device must be performed, we consider it sufficient to require that 
information relating to the accuracy test, such as the date on or 
before which the next accuracy test must be performed, be included in 
the processor's records of the accuracy test (final Sec.  113.100(c)).
    (Comment 15) One comment questioned why the documentation 
requirements for accuracy tests in proposed Sec.  113.40(a)(1)(ii)(B) 
apply to reference devices. The comment pointed out that the reference 
device may be located in a third party calibration laboratory.
    (Response) Accuracy tests for temperature-indicating devices may be 
performed by the processor or by a third party calibration laboratory. 
Processors who perform their own accuracy test must ensure that the 
reference device they use is accurate and must maintain records to 
document that accuracy. In this final rule, we clarify that the 
required records of the accuracy tests for reference devices are for 
reference devices maintained by the processor (final Sec. Sec.  
113.40(a)(1), (b)(1), (c)(1), (d)(1), (e)(1), (f)(1), and (g)(1)(i)(A), 
113.87(c), and 113.100(d)).
    (Comment 16) One comment recommended that processors be required to 
implement a method or process for identifying when a temperature-
indicating device needs to be calibrated. The comment pointed out that 
inexpensive software packages are readily available for this purpose.
    (Response) We recognize that processors may desire to establish a 
system to prompt them when scheduled activities, such as calibrations, 
need to be performed. Although available software may be appropriate 
for that purpose, we do not agree that the regulations should require 
processors to develop or use existing software or any other specific 
method or system to identify when a temperature-indicating device needs 
to be calibrated. Processors must test temperature-indicating devices 
for accuracy upon installation and at least once a year thereafter, or 
more frequently if necessary (final Sec.  113.40(a)(1), (b)(1), (c)(1), 
(d)(1), (e)(1), (f)(1), and (g)(1)(i)(A)). The appropriate frequency 
for the accuracy test should be determined based on previous accuracy 
test results, evidence of damage, and other factors or situations that 
cause the accuracy of the temperature-indicating device to be 
questionable.
    (Comment 17) One comment objected to the preamble statement, ``FDA 
recommends, but is not proposing to require, a dual probe design.'' (72 
FR 11989 at 11993). According to the comment, FDA's recommendation for 
a dual probe design will lead companies to purchase a dual probe unit 
to reduce any potential conflict with FDA. The comment stated that the 
dual probe design is a patented technology and other designs or 
mechanisms may be used for detecting malfunctions.
    (Response) In the preamble to the proposed rule, FDA stated, ``The 
design of the mercury-in-glass thermometer makes it relatively easy to 
detect a malfunction, including those caused by environmental 
conditions, because most are associated with a broken thermometer, 
separated column, or scale slippage. However, malfunction of other 
temperature-indicating devices may need to be detected by means other 
than observation. For example, a temperature-indicating device could be 
designed with a dual probe sensor that would enable detection of loss 
of accuracy of one of the probes when the probe readings do not agree. 
FDA recommends, but is not proposing to require, a dual probe design. 
FDA recognizes that specific design specifications for temperature-
indicating devices may limit the flexibility of the regulation for 
current and future technologies'' (72 FR 11990 at 11993). Thus, in the 
preamble to the proposed rule, we discussed a dual probe sensor as one 
means to detect a malfunction of a temperature-indicating device. We 
agree that a dual probe sensor is not the only design, mechanism, or 
process that may help detect temperature-indicating device failures. 
Therefore, this final rule does not require a dual probe design to 
detect malfunctions or failures of a temperature-indicating device.
    (Comment 18) One comment objected to the requirement for ``written 
documentation,'' found in proposed Sec. Sec.  113.40(a)(1)(ii), 
(b)(1)(ii), (c)(1)(ii), (d)(1)(ii), (e)(1)(ii), (f)(1)(ii), and 
(g)(1)(i)(A)(2). The comment indicated that the term ``written'' 
implies hand-written documentation and will limit new documentation 
technologies. The comment stated that the term ``written'' should be 
removed to allow for means of documentation other than just written 
records, especially since the Agency proposed in Sec.  113.100(f) to 
allow electronic records. The comment also stated that the term 
``written'' should be removed from other sections of the regulations 
that apply to records.
    (Response) We do not agree that the term ``written'' implies that 
the documents are hand-written. Written documentation may be generated

[[Page 11898]]

mechanically, such as when a stylus generates a tracing onto a paper 
chart, or electronically, including computer generated documents. 
However, we do agree that the term is not necessary for describing the 
requirements for establishing and maintaining records. Therefore, in 
this final rule, we used the term ``record'' or ``records'' without the 
qualifying term ``written'' (final Sec. Sec.  113.87(e) and 113.100(b) 
and (e)). For consistency, we also removed the qualifying term 
``written'' from Sec.  113.87(b). In addition, where the term ``written 
documentation'' is intended to mean ``records'' that must be 
established and maintained, we changed the term ``written 
documentation'' to ``records'' (final Sec.  113.40(a)(1)(ii), 
(b)(1)(ii), (c)(1)(ii), (d)(1)(ii), (e)(1)(ii), (f)(1)(ii), and 
(g)(1)(i)(A)(2)).
    (Comment 19) One comment recommended that proposed Sec.  
113.40(b)(6)(ii) on water circulation be redesignated as new Sec.  
113.40(b)(9). The comment suggested that it was inappropriate to place 
the requirements for water circulation and for air supply in the same 
section, specifically proposed Sec.  113.40(b)(6)(i) and (b)(6)(ii), 
which, according to the comment, respectively addressed air supply and 
water control. The comment stated that, for discontinuous water retort, 
air supply and water circulation are not related functions as they are 
for vertical water retorts covered in Sec.  113.40(b).
    (Response) The proposed rule does not have a Sec.  
113.40(b)(6)(ii). Because the comment was related to water circulation 
for discontinuous agitating retorts, we assume the comment was 
requesting redesignation of proposed Sec.  113.40(e)(6)(ii). We also 
assume the comment was comparing proposed Sec.  113.40(e)(6)(ii), 
related to water circulation in discontinuous agitating retorts, to 
proposed Sec.  113.40(b)(10)(ii), related to water circulation in still 
retorts, including vertical still retorts. We reviewed the structure of 
proposed Sec.  113.40(b)(10) and (e)(6) and agree that separating the 
requirements for the air supply and controls and the water circulation 
functions into distinct paragraphs for both discontinuous agitating and 
still retorts enhances the clarity of the regulation. We also 
determined that, based on changes to proposed Sec.  113.40(e)(8), as 
explained in response to Comment 20, proposed Sec.  113.40(b)(9) and 
(e)(8), relating to the water level indicator, should be redesignated 
to immediately precede proposed Sec.  113.40(b)(10)(ii) and (e)(6)(ii), 
respectively, relating to water circulation. Thus, in this final rule, 
we redesignated proposed Sec.  113.40(b)(9), (b)(10)(i), and 
(b)(10)(ii) as final Sec.  113.40(b)(10), (b)(9), and (b)(11), 
respectively. We redesignated proposed Sec.  113.40(e)(6)(ii) and 
(e)(8) as final Sec.  113.40(e)(7) and (e)(6)(ii), respectively. We 
made conforming changes to the numbering of proposed Sec.  
113.40(b)(11), (b)(12), (b)(13), and (b)(14), which is now final Sec.  
113.40(b)(12), (b)(13), (b)(14), and (b)(15), respectively. Similarly, 
we redesignated proposed Sec.  113.40(e)(6)(ii) and (e)(7), as final 
Sec.  113.40(e)(7) and (e)(8), respectively.
    (Comment 20) One comment suggested revising proposed Sec.  
113.40(b)(6), relating to air supply and controls, to clarify that the 
requirements apply only if air is used for providing overpressure. The 
comment also suggested revising proposed Sec.  113.40(e)(8), which 
requires a water level indicator and operator checks of the water level 
to ensure that water covers the top layer of containers during the 
entire come-up time and processing periods. The comment requested 
revisions to clarify that the requirements of proposed Sec.  
113.40(e)(8) apply only if water level is determined to be a critical 
factor in the scheduled process or retort operating procedures. 
According to the comment, these revisions would accommodate current 
systems for pressure processing in discontinuous agitating retorts that 
utilize steam as the source of overpressure. The comment stated that 
for such systems, the processing authority may have determined that 
water level is not critical to the scheduled process because of the 
influences of steam in the retort headspace area and the continuous 
rotation of the retort baskets.
    (Response) Because proposed Sec.  113.40(b)(6) does not relate to 
air supply and controls, but is instead about crate supports, we assume 
here as we did in our response to Comment 19 that the comment is 
referring to proposed Sec.  113.40(e)(6)(i), relating to air supply and 
controls for pressure processing in water in discontinuous agitating 
retorts. Proposed Sec.  113.40(e)(6)(i) requires that a means be 
provided for introducing compressed air at the proper pressure and 
rate. We agree with the comment that the requirement of proposed Sec.  
113.40(e)(6)(i) applies only if air is used for providing overpressure. 
We also agree that the requirement of proposed Sec.  113.40(e)(6)(ii) 
for a water level indicator and recorded checks of the water level 
during processing should be revised to accommodate discontinuous 
agitating retorts that utilize steam as the source of overpressure. 
Accordingly, in final Sec.  113.40(e)(6)(i) and (e)(6)(ii), we 
clarified that the requirements relating to air supply and controls and 
to the water level indicator apply only if air is used for providing 
overpressure.
    (Comment 21) One comment suggested revising proposed Sec.  
113.40(b)(10)(ii), which requires the water circulation pump to be 
equipped with a bleeder to remove air when starting operations. The 
comment suggested revising this requirement to allow for use of other 
suitable devices for air removal.
    (Response) We agree that proposed Sec.  113.40(b)(10)(ii), 
redesignated as Sec.  113.40(b)(11) in this final rule, should be 
revised to allow for use of water circulation pumps, other than a water 
circulation pump with a bleeder, designed to ensure proper heat 
distribution. To ensure proper heat distribution, the water circulation 
pump must be designed to properly start the flow of water and to 
maintain the flow of water at the appropriate flow rate. To obtain the 
appropriate flow rate, the water circulation pump must be designed or 
equipped with a suitable means, such as a bleeder, to remove air from 
the pump chamber or the pump must be self priming. In addition, the 
pumping system must ensure that it avoids cavitation, i.e., changes in 
water pressure caused by the formation of cavities or voids within the 
circulating water. Water circulation pumps that use mechanisms other 
than bleeders to remove air must be designed to ensure appropriate 
water circulation and to prevent cavitation.
    To clarify this requirement, in Sec.  113.40(b)(11) of this final 
rule we specify that the water circulation pump must be designed to 
provide proper flow on startup and during operation, such as with a 
bleeder or other suitable means to remove air during startup and with 
an appropriate device or design to prevent pump cavitation during 
operation. In addition, the pump must be equipped with a signaling 
device to warn the operator when it is not running. For consistency, we 
made similar changes to proposed Sec.  113.40(e)(6)(ii) (redesignated 
as Sec.  113.40(e)(7) in this final rule). In final Sec.  113.40(b)(11) 
and (e)(7), we removed the reference to ``pilot light'' as the example 
of a signaling device to avoid the appearance of preference for a pilot 
light signaling device and to provide flexibility for processors to 
determine an appropriate signaling device.
    (Comment 22) One comment agreed with the provision of proposed 
Sec.  113.40(b)(1)(v) that allows a temperature-indicating device to be 
installed in a separate well or sleeve, i.e., ``If a separate well or 
sleeve is used, there must be adequate circulation to

[[Page 11899]]

ensure accurate temperature measurement.'' However, the comment 
indicated that the provision appears to conflict with another 
requirement in proposed Sec.  113.40(b)(1)(v) for the temperature-
indicating device sensor to extend directly into the water a minimum of 
at least 2 inches (5.1 centimeters) without a separable well or sleeve.
    (Response) We agree that additional clarification is needed. In 
this final rule, we revised proposed Sec.  113.40(b)(1)(v) and a 
similar requirement in proposed Sec.  113.40(e)(1)(v) to clarify that 
the temperature-indicating device sensor must be installed directly 
into the retort shell or in a separate well or sleeve attached to the 
retort. In addition, for all retorts covered by these sections, the 
temperature-indicating device sensor must be located so that it is 
beneath the surface of the water throughout the process and where there 
is adequate circulation to ensure accurate temperature measurement. We 
also removed the requirement for the temperature-indicating device 
sensor to extend at least 2 inches (5.1 centimeters) directly into the 
water when the temperature-indicating device sensor is not located in a 
separate well or sleeve. We believe the requirement for adequate water 
circulation to ensure accurate temperature measurement obviates the 
need to specify how far the temperature-indicating device sensor must 
extend into the water and allows for use of alternative technologies.
    (Comment 23) One comment noted that proposed Sec.  113.40(f)(1)(v) 
should be revised to clarify that placement requirements in the steam 
dome and the hydrostatic water leg are for the temperature-indicating 
device sensor.
    (Response) We agree. In this final rule, we revised proposed Sec.  
113.40(f)(1)(v) to clarify that the placement requirements in the steam 
dome and the hydrostatic water leg apply to the temperature-indicating 
device sensor, rather than the entire temperature-indicating device.
    (Comment 24) One comment stated that the requirement for the 
temperature-recording device sensor to be installed either within the 
retort shell or in a well attached to the shell is misplaced in the 
paragraph heading, Temperature controller (proposed Sec.  
113.40(a)(2)(iv), (c)(2)(iv), (d)(2)(iv), (e)(2)(iv), and (f)(2)(iv)). 
The comment indicated that the statement applies to all temperature-
recording device sensors, but its placement in the regulations implies 
that it applies only to combination recording-controlling devices. The 
comment suggested moving the statement relating to installation of the 
sensor, along with the requirement for the temperature-recording device 
sensor well to have a \1/16\-inch (1.5 millimeters) or larger bleeder, 
to a separate paragraph.
    (Response) We agree. In this final rule, we moved the statements 
relating to installation of the sensor and, where relevant, the 
requirement for the temperature-recording device sensor well to have a 
\1/16\-inch (1.5 millimeters) or larger bleeder to the paragraph 
heading, Temperature-recording device (final Sec.  113.40(a)(2), 
(c)(2), (d)(2), (e)(2), and (f)(2)).
    (Comment 25) One comment objected to the requirement in proposed 
Sec.  113.40(e)(1)(v) for the temperature-indicating device sensor to 
be installed either within the retort shell or in an external well 
attached to the retort. The comment indicated that placement of the 
temperature-indicating device in the suction manifold shows good 
agreement with temperatures inside the retort once the Cook Hold step 
begins. According to the comment, this placement is an improvement over 
using a thermometer well, since the water line for a partial immersion 
process is normally below the feed leg of the thermometer well and the 
temperature at that location may not be representative of the retort 
temperature. The comment suggested revising Sec.  113.40(e)(1)(v) by 
adding the following language to permit alternative sensor placement, 
if appropriately documented: ``Other installations deviating from these 
sensor locations may be used if the processor has evidence, on file, in 
the form of heat distribution data that its installation accomplishes 
adequate heat distribution. Such documentation is likely to include 
heat distribution studies conducted and documented by the processor to 
show that the process temperature will be reached once the Cook Hold 
time begins.''
    (Response) We do not agree with the comment's recommendation that 
Sec.  113.40(e)(1)(v) should state that process deviations relating to 
placement of temperature-indicating device sensors may be acceptable if 
supported by heat distribution data. Section 108.35 states the 
requirements for submitting information to demonstrate process adequacy 
for a system design that deviates from the requirements of the 
regulations. A change in the design of a system for processing in water 
in discontinuous agitating retorts, such as placement of a temperature-
indicating device sensor in a suction manifold rather than within the 
retort shell or in an external well attached to the retort, would 
require substantiation by qualified scientific authority as to its 
adequacy, including, for example, heat distribution studies as 
suggested by the comment. Such information must be submitted to FDA 
(Sec.  108.35(c)(2)(ii) (21 CFR 108.35(c)(2)(ii))).
    (Comment 26) One comment expressed concern that proposed Sec.  
113.40(a)(2), which requires each retort to have an accurate 
temperature-recording device, does not define the term ``accurate'' or 
state how to determine that a temperature-recording device is accurate. 
The comment suggested using the same calibration method for 
temperature-recording devices as used for temperature-indicating 
devices and reference devices by requiring annual calibrations of 
temperature-recording devices with NIST traceability. The comment 
stated that this would effectively allow the temperature-recording 
device to be used as a secondary component of a ``redundant system'' to 
verify the accuracy of the temperature-indicating device. Accordingly, 
the temperature-indicating device would still be ``the standard'' 
device and should still be required to have the characteristics of high 
accuracy and reliability. The comment indicated that if the 
temperature-recording device is adjusted to the temperature-indicating 
device and the temperature-indicating device slowly drifts, this may 
not be known until the next calibration cycle, which could be up to a 
year later. However, according to the comment, if the devices are 
allowed to vary within their individual established calibration 
tolerances, it will be known if one device drifts out of its tolerance. 
The comment stated that adjusting the temperature-recording to the 
temperature-indicating device does not ensure the accuracy of the 
temperature-recording device or the recorded data.
    (Response) This final rule requires the temperature-recording 
device to be adjusted to agree as nearly as possible with, but to be in 
no event higher than, the temperature-indicating device during the 
process time (final Sec.  113.40(a)(2)(iii), (b)(2)(iii), (c)(2)(iii), 
(d)(2)(iii), (e)(2)(iii), (f)(2)(iii), and (g)(1)(i)(B)(3)). Processors 
must ensure that the temperature-indicating device is accurate during 
processing and that the recording mechanism of the temperature-
recording device is adjusted to and reflects the temperature indicated 
by the temperature-indicating device. For some temperature-recording 
devices, such as those that record to a chart, adjustments to the 
mechanism that draws onto the chart are made by hand based on visually 
determining where the mechanism should be placed in contact with the 
chart. Unavoidable

[[Page 11900]]

imprecision relating to, for example, manual placement of the recording 
mechanism onto a chart, must result in recording a temperature that is 
not greater than the actual processing temperature. A recorded 
temperature that is higher than the actual processing temperature may 
mean that the product was not processed at or above the required 
processing temperature (i.e., the product was under processed) and may 
pose a health hazard. However, if the temperature-recording device 
records a temperature that is lower than the actual processing 
temperature, although the quality of the product may be affected, 
processing at a higher temperature than recorded (i.e., over 
processing) does not create a health hazard. Thus, although the 
recorded temperature should reflect the actual processing temperature 
as precisely as possible, we believe the requirement to not record a 
temperature that is higher than the temperature-indicating device, 
which must be accurate, provides an appropriate parameter for ensuring 
that the product is not under processed.
    We believe processors should adjust the temperature-recording 
device mechanism for each batch at least at the beginning of the 
process and, as necessary, check the adjustment during the process time 
to ensure compliance with the regulation and to ensure that the batch 
is processed at or above the scheduled process temperature. To 
emphasize that the adjustment must occur with sufficient frequency to 
ensure that the temperature-recording device record reflects the 
temperature indicated by the temperature-indicating device, we revised 
the final rule to require the temperature-recording device to be 
adjusted with sufficient frequency to ensure agreement as nearly as 
possible with, but to be in no event higher than, the temperature-
indicating device during processing (final Sec.  113.40(a)(2)(iii), 
(b)(2)(iii), (c)(2)(iii), (d)(2)(iii), (e)(2)(iii), (f)(2)(iii), and 
(g)(1)(i)(B)(3)).
    (Comment 27) One comment suggested replacing the term ``recording 
chart'' with ``temperature-recording device record'' in proposed Sec.  
113.40(c)(8)(ii).
    (Response) We agree. In Sec.  113.40(c)(8)(ii) of this final rule, 
we replaced the term ``recording chart'' with ``temperature-recording 
device record.'' Also, because the term ``marked'' may be interpreted 
to mean a manual action, for clarity and to allow for use of 
alternative technologies, we replaced the term ``marked'' with 
``indicated'' in Sec.  113.40(c)(8)(ii) and (c)(9).
    (Comment 28) One comment suggested that the statement that air-
operated temperature controllers should have adequate filter systems to 
ensure a supply of clean, dry air is misplaced in the regulations 
(proposed Sec.  113.40(a)(2)(iv), (b)(2)(iv), (c)(2)(iv), (d)(2)(iv), 
(e)(2)(iv), and (f)(2)(iv)). The comment stated that, because this 
statement applies to