DEPARTMENT OF TRANSPORTATION

Federal Aviation Administration

14 CFR Part 39

[Docket No. FAA-2012-0002; Directorate Identifier 2011-NE-42-AD;
Amendment 39-18610; AD 2016-16-12]
RIN 2120-AA64

Airworthiness Directives; Continental Motors, Inc. Reciprocating
Engines

AGENCY: Federal Aviation Administration (FAA), DOT.

ACTION: Final rule.

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

SUMMARY: We are adopting a new airworthiness directive (AD) for certain
Continental Motors, Inc., (CMI) San Antonio (formerly known as
Airmotive Engineering Corp. (AEC)), replacement parts manufacturer
approval (PMA) cylinder assemblies marketed by Engine Components
International Division (ECi). On July 17, 2015, AEC was purchased by
CMI and is now operating as ``Continental Motors--San Antonio.'' These
cylinder assemblies are used on all CMI model -520 and -550
reciprocating engines, and on all other CMI engine models approved for
the use of model -520 and -550 cylinder assemblies, such as the CMI
model -470 when modified by supplemental type certificate (STC). This
AD was prompted by reports of multiple cylinder head-to-barrel
separations and cracked and leaking aluminum cylinder heads. This AD
requires removal of the affected cylinder assemblies, including
overhauled cylinder assemblies, according to a phased removal schedule.
We are issuing this AD to prevent failure of the cylinder assemblies,
which could lead to failure of the engine, in-flight shutdown, and loss
of control of the airplane.

DATES: This AD is effective September 15, 2016.

ADDRESSES: For service information identified in this AD, contact
Continental Motors, Inc., San Antonio, 9503 Middlex Drive, San Antonio,
TX 78217; phone: 210-820-8100; Internet: http://www.continentalsanantonio.
com. You may view this service information
at the FAA, Engine & Propeller Directorate, 1200 District Avenue,
Burlington, MA. For information on the availability of this material at
the FAA, call 781-238-7125. It is also available on the Internet at
http://www.regulations.gov by searching for and locating Docket No.
FAA-2012-0002.

Examining the AD Docket

You may examine the AD docket on the Internet at http://www.regulations.gov
by searching for and locating Docket No. FAA-2012-
0002; or in person at the Docket Management Facility between 9 a.m. and
5 p.m., Monday through Friday, except Federal holidays. The AD docket
contains this AD, the regulatory evaluation, any comments received, and
other information. The address for the Docket Office (phone: 800-647-
5527) is Document Management Facility, U.S. Department of
Transportation, Docket Operations, M-30, West Building Ground Floor,
Room W12-140, 1200 New Jersey Avenue SE., Washington, DC 20590.

FOR FURTHER INFORMATION CONTACT: Jurgen E. Priester, Aerospace
Engineer, Delegation Systems Certification Office, FAA, Rotorcraft
Directorate, 10101 Hillwood Parkway, Fort Worth, TX 76177; phone: 817-
222-5190; fax: 817-222-5785; email: jurgen.e.priester@faa.gov.

SUPPLEMENTARY INFORMATION:

Discussion

August 12, 2013--NPRM

We issued a notice of proposed rulemaking (NPRM) to amend 14 CFR
part 39 by adding an AD that would apply to certain CMI San Antonio
replacement PMA cylinder assemblies marketed by ECi. These assemblies
are used on CMI model -520 and -550 reciprocating engines, and all
other CMI engine models approved for the use of models -520 and -550
cylinder assemblies such as the CMI model -470 when modified by STC.
The NPRM published in the Federal Register on August 12, 2013 (78 FR
48828) (referred to herein after as the ``August 12, 2013, NPRM''). The
August 12, 2013, NPRM proposed to require initial and repetitive
inspections, immediate replacement of cracked cylinder assemblies, and
replacement of cylinder assemblies at reduced times-in-service (TIS)
since new. The August 12, 2013, NPRM also proposed to prohibit the
installation of affected cylinder assemblies into any engine.

September 26, 2013-March 12, 2014--Posting Technical Documents/
Extension of Comment Period/Initial Regulatory Flexibility Analysis
(IRFA)

We received several hundred comments to our August 12, 2013, NPRM.
In response to this high-level of public interest, we undertook several
actions to help the public understand and provide further comment on
our proposed rule. These actions included:
Extending the comment period to the August 12, 2013, NPRM;
publishing an IRFA; and
adding several technical documents that were posted to
Docket No. FAA-2012-0002 (see Addresses section of this final rule for
information on locating the docket) on September 20, 2013.
Documents added to the docket include:
(1) FAA Safety Recommendations 08.365, 08.366, and 11.216, which
were written against the subject ECi cylinder assemblies;
(2) NTSB Safety Recommendation A-12-7, also written against the
subject ECi cylinder assemblies;
(3) The original ECi AD worksheet for 2011-NE-42-AD, which
documents the reasons for the proposed rule;
(4) A list of separations of ECi cylinder assemblies;
(5) A white paper \1\ on failures of ECi cylinders by the FAA Chief
Scientific and Technical Adviser (CSTA) for Engine Dynamics;
---------------------------------------------------------------------------

\1\ An authoritative report that informs readers about a complex
issue.
---------------------------------------------------------------------------

(6) Figures showing ECi Dome Separation Failures;
(7) A briefing on ``ECi Cylinder Head Failures on Continental IO
520 & 550 Engines''; and
(8) FAA Policy Memorandum on ``Risk Assessment for Reciprocating
Engine Airworthiness Directives,'' dated May 24, 1999.
We notified the public of these actions on September 26, 2013, via
the Federal Register (78 FR 59293). In that notification, we extended
the comment period for the August 12, 2013, NPRM to December 11, 2013.
This extension allowed the public additional time to comment on our
August 12, 2013, NPRM and the additional information we had added to
the docket.
We also determined that we needed to add to the docket a detailed
regulatory flexibility analysis to estimate the effects of the proposed
rule on small business entities. We published an Initial Regulatory
Flexibility Analysis in Docket FAA-2012-0002 on March 12, 2014 (79 FR
13924).

September 3-4, 2014--Challenge Team's Review of August 12, 2013, NPRM

Because the response to our August 12, 2013, NPRM was so negative--
we received over 500 comments, most disagreeing with the NPRM--we
established a Challenge Team to review our proposed AD. The Challenge
Team was an independent, multi-disciplinary team, consisting of three
FAA CSTAs, FAA Aircraft Certification Service (AIR) managers, and other
FAA technical experts from all four Directorates.
The Challenge Team reviewed the technical information that formed
the basis for our proposed AD and the public comments we had received
concerning our proposal. The CSTA for Aircraft Safety Analysis also
independently computed a new risk assessment using the earlier failure
reports, and the additional failure reports that we received from the
public as comments to our August 12, 2013, NPRM.
Based on their review of this data and the new risk assessment of
failures of affected cylinder assemblies, the Challenge Team determined
that an AD was still required. But, they suggested changes to make
compliance less aggressive and substantially reduce cost. Their
recommended changes included revising the compliance schedule in favor
of a phased removal schedule, clarifying that overhauled cylinder
assemblies are included in the proposed phased removal schedule,
eliminating the reporting requirement for removed cylinder assemblies,
and removing the requirement for initial and repetitive inspection.

January 8, 2015--First Supplemental Notice of Proposed Rulemaking
(SNPRM)

We adopted the Challenge Team's recommendations, and we then
published them as an SNPRM in the Federal Register on January 8, 2015
(80 FR 1008) (referred to herein after as the ``January 8, 2015,
SNPRM''). The January 8, 2015, SNPRM proposed to modify the schedule
for removal of the affected cylinder assemblies, added that overhauled
affected cylinder assemblies be removed within 80 hours, eliminated a
reporting requirement, and removed a requirement for initial and
repetitive inspections.
We also responded in our January 8, 2015, SNPRM, to the several
hundred comments that we received to the August 12, 2013, NPRM. Many of
these comments were repetitious, so we grouped the comments and
provided our responses to the different groups,
depending on the nature of the comment. For example, some comments
claimed that airplanes can operate safely with a separated cylinder
head; others suggested that pilot error was causing cylinder head
separations; and others recommended adopting less stringent compliance
requirements. Each of these groups received our response to the group's
comment.

June 9, 2015--Meeting With National Transportation Safety Board (NTSB)

The NTSB, in its comments to our August 12, 2013, NPRM; January 8,
2015, SNPRM; and in its Safety Recommendation A-12-07, did not fully
support our approach to resolving the unsafe condition that is the
subject of this final rule. Therefore, we met with the NTSB on June 9,
2015 to understand the technical basis for their recommendation and
their technical objections to our proposed AD. At this meeting, we
presented the NTSB the technical information upon which we based our AD
as amended. Information that was reviewed included failure reports, the
risk assessment by the FAA's CSTA for Aircraft Safety Analysis, FAA
safety recommendations, and the data supporting our conclusion that
field inspections had an insufficient probability of cylinder failure
detection.
The NTSB noted in this meeting that Safety Recommendation A-12-7,
and the NTSB's comments to the August 12, 2013, NPRM and the January 8,
2015, SNPRM, were based on the information available to them at that
time. The NTSB also indicated it would reassess its recommendation and
comments to our proposed rule based on the presentations and the
supporting data that we had presented.

June 23, 2015--Additional Technical Documents Posted

We received additional comments to our August 12, 2013, NPRM and
our January 8, 2015, SNPRM, requesting that we provide additional
information that supports this AD. Commenters also requested that we
identify the data that we relied on in drafting this AD and to explain
why that data supported our conclusion that an unsafe condition exists.
Based on these comments, we concluded that further additional public
participation in our proposed AD was appropriate. Specifically, we
concluded that we would post to the docket the additional technical
information responsive to the comments. So, on June 23, 2015, we posted
the additional technical information to Docket No. FAA-2012-0002 (see
ADDRESSES section of this final rule for information on locating the
docket). These documents provide further technical rationale for this
AD. This additional technical information included:
(1) The risk analysis process conducted by the FAA's CSTA for
Aircraft Safety Analysis--referenced in Docket No. FAA-2012-0002 as the
``Proposed Airworthiness Directive for ECi Cylinders Risk Analysis
Process,'' referred to herein as the ``risk analysis'';
(2) A risk analysis using the Small Airplane Risk Analysis (SARA)
methods used by the FAA's Small Airplane Directorate (SAD)--referenced
in Docket No. FAA-2012-0002 as ``SARA Worksheet Systems/Propulsion'';
(3) A June 2011, presentation by AEC to the FAA concerning its ECi
cylinder assemblies;
(4) A list of ECi cylinder assembly failure reports consisting of
only those reports where both cylinder serial number and time in
service are included in the reports;
(5) A list of additional failures of ECi cylinder assemblies
reported by a maintenance organization; and
(6) AEC Technical Report 1102-13, dated April 30, 2011.

August 28, 2015--2nd SNPRM

We published a second SNPRM in the Federal Register on August 28,
2015 (80 FR 52212, referred to herein after as the ``August 28, 2015,
SNPRM''). The August 28, 2015, SNPRM retained the compliance
requirements proposed by the January 8, 2015, SNPRM. We published the
August 28, 2015, SNPRM to provide the public a final opportunity to
comment on the proposed AD and the additional technical documentation
we had added to the docket on June 23, 2015.
Also, since many commenters had cited NTSB support for their
positions, we wanted to clarify our rationale for disagreeing with the
compliance actions proposed by the NTSB in its Safety Recommendation A-
12-7, and the NTSB's comments to the August 12, 2013, NPRM and the
January 8, 2015, SNPRM.
The NTSB did submit a final comment to our August 28, 2015, SNPRM,
that was posted to the docket on November 23, 2015. In the NTSB's final
comment, the NTSB indicated that it now considers that our proposed
compliance actions satisfy the intent of Safety Recommendation A-12-7.
The information we covered with the NTSB, including copies of FAA
presentations to the NTSB, were subsequently posted to Docket No. FAA-
2012-0002 (see ADDRESSES section of this final rule for information on
locating the docket) on April 6, 2016.

Comments

Introduction

We have, through the August 12, 2013, NPRM; the September 26, 2013,
posting of additional information; our extension of the August 12,
2013, NPRM comment period; the January 8, 2015, SNPRM; and August 28,
2015, SNPRM, given the public the opportunity to participate in
developing this AD. The public, as noted already, has participated
deeply in this rule making; providing hundreds of comments.
This final rule includes our responses to any previously
unaddressed comments to the August 12, 2013, NPRM and to the January 8,
2015, SNPRM, that we may have left without response, and to the August
28, 2015, SNPRM.
To organize comments and facilitate their review, we again grouped
like comments and responses. These groupings in this final rule's
comments section are:
(1) Comments to withdraw or revise the SNPRMs for technical
reasons--these comments, and the resulting groupings, were similar to
those we used in responding to the August 12, 2013, NPRM. They include,
for example, requests to withdraw the SNPRM because the commenters
claim that ECi cylinder assemblies are not unsafe; airplanes can
operate safely with a separated cylinder head; or the root cause of
cylinder failure is unknown.
(2) Comments to the FAA's risk assessment processes and policies--
these comments generally asserted that the SNPRMs should be withdrawn
because the FAA had not appropriately followed its risk assessment
processes and policies in determining that the failure of ECi cylinder
assemblies represents an unsafe condition.
(3) Comments to the FAA's rulemaking processes--these comments
generally requested that the SNPRMs be withdrawn, alleging that the FAA
had failed to follow its rulemaking processes and was adopting a rule
that is ``arbitrary and capricious.''
(4) Comments to the cost of compliance--these comments indicated
that the cost of compliance to this AD was higher than the FAA has
estimated and will have a substantial effect on small entities.
(5) Administrative comments--these were generally comments that did
not pertain to the substance of this AD, such as requests for names and
phone numbers of FAA personnel involved in this rulemaking.
(6) Support for the SNPRMs--these were comments in support of
issuing the SNPRMs.

A. Comments To Withdraw or Revise the SNPRMs for Technical Reasons

Request To Withdraw the SNPRMs Because ECi Cylinder Assemblies Are
Not Unsafe

Comment. Several organizations and individuals, commenting to the
August 12, 2013, NPRM, commented also to the January 8, 2015, and
August 28, 2015, SNPRMs, that the affected ECi cylinder assemblies have
an equivalent, or lower, failure rate than that of cylinder assemblies
manufactured by the original equipment manufacturer (OEM). The
commenters also indicated that there have been no failures of ECi
cylinder assemblies in the last 3 years. These commenters request the
FAA withdraw this AD because they believe that the ECi cylinder
assemblies are not unsafe.
Response. We disagree. The rate of separation for the affected ECi
cylinder assemblies is at least 32 times greater than that of OEM
cylinder assemblies over the same period. Although there are
approximately four times as many OEM cylinder assemblies in service
than ECi cylinder assemblies, the ECi cylinder assemblies suffered more
cylinder head separations than OEM cylinder assemblies since 2004. This
data is available for review in Docket No. FAA-2012-0002 (see ADDRESSES
section of this final rule for information on locating the docket). In
addition, we have continued to receive field reports of failures of the
affected cylinders in the past three years. We did not withdraw the
August 28, 2015, SNPRM.
Comment. Commenters also questioned the validity of the data that
the FAA used to justify the proposed AD.
Response. We interpret the comment as suggesting that the data used
to justify the rule is not valid. We disagree. We used warranty reports
from ECi and RAM Aircraft, which is a major overhauler of CMI engines,
STC holder for an increased horsepower version of the affected model
engine, and the largest user of the affected ECi cylinders. We also
used service difficulty reports (SDRs), and other field service reports
regarding ECI cylinder separations. We did not withdraw the August 28,
2015, SNPRM.
Comment. The IPL Group LLC (IPL Group) commented that the FAA has
mischaracterized ``quality enhancements'' in production as ``design
changes.'' IPL Group noted that ECi had applied experience gained
during manufacturing, as well as through service feedback, to make
quality improvements in production and the changes made to the design
data were not due to design deficiencies.
Response. We disagree. We correctly stated that ECi has made
increases in the dome transition radius through cylinder serial number
33697, and has made incremental increases in the head-to-barrel
interference fit at least through cylinder serial number 61177 (see
Airmotive Engineering Technical Report 1102-13) to address the two
identified inherent design deficiencies associated with the effected
cylinder assemblies. These changes are design changes. We did not
withdraw the August 28, 2015, SNPRM.
Comment. RAM Aircraft commented that when it submitted its December
9, 2013, comment, it calculated the likelihood of a cylinder
separation. RAM Aircraft indicated it provided a significant amount of
data that proves that the likelihood of a cylinder separation is
``extremely remote.'' RAM commented that at that time their data showed
one cylinder separation for every: 21,808 multi-engine aircraft flight
hours, or 172 average years of active service; and 42,057 single engine
aircraft flight hours, or 455 average years of active service. Further,
that the fleet of aircraft using the cylinders subject to the January
8, 2015, SNPRM have continued to fly for an additional 14 months since
December 9, 2013. RAM Aircraft indicated that there is no doubt that
both the 21,808 multi-engine aircraft flight hour number, and the
42,057 single engine aircraft flight hour number, would both be now
much larger, thereby, further reducing the likelihood of a cylinder
separation.
Response. We disagree. RAM Aircraft's data does not substantiate
its claimed failure rate. Without knowing the total number of hours
flown on all affected cylinders, it is not possible to accurately
calculate an hours-based failure rate. This data is not available for
general aviation aircraft. We, therefore, find RAM Aircraft's estimate
to be unreliable. We did not withdraw the August 28, 2015, SNPRM.
Comment. RAM Aircraft also indicated that a statement by the FAA in
the January 8, 2015, SNPRM regarding numbers of failures of affected
cylinder assemblies was grossly misleading. RAM Aircraft assumes that
the FAA is referring to reports entered via the SDR system. RAM
Aircraft indicated that it has provided evidence in an earlier comment
that not every piece of information in the SDR system can be taken at
face value. With respect to this SNPRM, RAM suggested that it is very
important to distinguish between the ``SNPRM failure modes''
(quotations not in original comment) and other types of ``nuisance''
cracks that are common occurrences in all manufacturer's air-cooled
aircraft cylinders. The SNPRM failure modes do not include cracks
between spark plug holes, valve seats, injector ports, etc. There is no
doubt that the ``hundreds of failures'' referenced by the FAA were
never researched to determine which were of the SNPRM failure mode and
which were of the ``nuisance'' variety.
Response. We disagree. Our response in the January 8, 2015, SNRPM
is not misleading. On the contrary, under-reporting of cylinder
assembly cracks in the SDR system further reinforces the need for this
AD. Further, the FAA did not include the SDR failure reports referred
to by the commenter as of the ``nuisance'' variety in the list of
separations that were used to substantiate the need for this AD. We did
not base this AD on nuisance cracks in the affected cylinder
assemblies. We did not withdraw the August 28, 2015, SNPRM.
Comment. One commenter stated that the separated cylinders that
were determined to be the precipitating root cause events for the two
fatal accidents cited by the FAA in the January 8, 2015, SNPRM were
overhauled cylinders, so they therefore should not be considered in the
determination as to whether or not the proposed corrective action
should be implemented.
Response. We disagree. The ECi cylinder heads, P/N AEC 65385, of
the separated cylinder assemblies that precipitated the two referenced
fatal accidents were of the same type design and within the same
affected cylinder assembly serial number range as are used in new ECi
cylinder assemblies. The cast and then machined aluminum cylinder head
shrink band region has the predominant features that define the final
interference fit of the overall cylinder assembly, not the steel
barrel. This is further supported by the fact that the design changes
that ECi made to the interference fit were accomplished by modification
of the cylinder head. We did not withdraw the August 28, 2015, SNPRM.
Comment. Danbury Holdings commented that the FAA should withdraw
the August 28, 2015 SNPRM because the FAA failed to establish that the
affected product, i.e., the ECi cylinder assemblies, do not meet the
established minimum safety standards established by 14 CFR part 33.
Response. We disagree. The operational history of the affected ECi
cylinder assemblies established that the affected ECi cylinder
assemblies present an unacceptable compromise to safety, an unsafe
condition, when installed in operating aircraft engines. We did not
withdraw the August 28, 2015, SNPRM.

Comment. Danbury Holdings also stated that the ``same unsafe
condition'' that is addressed by this AD is present in the cylinders of
all manufacturers and that the FAA failed to consider similar failures
of the OEM cylinders.
Response. We disagree. The affected ECi PMA cylinders have
separated at a significantly higher rate than the OEM cylinders over
the same service period since the ECi PMA cylinders entered service.
ECi itself identified two root causes for the separations. See AEC
Technical Report 1102-13 in Docket No. FAA-2012-0002 (see ADDRESSES
section of this final rule for information on locating the docket)
which recommends withdrawal from service of the affected ECi cylinders.
We compared the number of separations of these affected ECi PMA
cylinders to the number of OEM separations over the same service
period, since the ECi PMA cylinders entered service in meaningful
numbers. Over the same period of time the affected ECi PMA cylinders
and OEM cylinders were in service, the ECi cylinders experienced eight
times the number of OEM separations, even though only one-quarter as
many ECI cylinders were in service as the OEM's. Further, the SDR
database does not reveal similar separation rates or similar failure
modes for OEM cylinders. Therefore, we have no reason to regard the OEM
cylinder assemblies as subject to the same or similar unsafe condition.
We did not withdraw the August 28, 2015, SNPRM.

Request To Withdraw the SNPRMs Because Airplanes Can Operate Safely
With a Separated Cylinder Head

Comment. Several commenters indicated that we should not issue this
AD because airplanes can continue to operate safely even after a
cylinder head separation.
Response. We disagree. An in-flight cylinder head separation is an
unsafe condition that presents multiple secondary effects. For example,
in-flight fire and loss of aircraft control. Accident data confirms
that separated cylinders have also been a precipitating event in fatal
accidents. Therefore, the safety consequences represented by a cylinder
head separation in flight are significant, and represent an unsafe
condition appropriate for an AD. We did not withdraw the August 28,
2015, SNPRM.
Comment. Several commenters added that airplane engines are
designed and certified to safely operate with one failed cylinder.
Response. We disagree. Applicants are not required to show that
their engines are designed to operate with one cylinder failed or with
a separated cylinder, nor that doing so constitutes safe operation of
an engine. We did not withdraw the August 28, 2015, SNPRM.
Comment. Danbury Aerospace commented that the docket contains
evidence from RAM Aircraft that valid and verifiable testing
establishes that a head-to-barrel separation results in less than 20
percent power loss to the engine.
Response. We disagree. The RAM Aircraft testing that is included in
Docket FAA-2012-0002 only quantified the horsepower output per
cylinder. The RAM Aircraft testing was of an uninstalled engine in a
test cell and RAM Aircraft did not attempt to assess the impact of
reduced engine horsepower output on airplane level performance. We
estimate that a 20% reduction in engine horsepower on a single-engine
airplane results in a nearly 40% reduction in aircraft rate of climb,
which is a hazardous condition. It is also a potentially hazardous
condition for twin-engine airplanes due to the resultant asymmetric
thrust condition. We did not withdraw the August 28, 2015, SNPRM.
Comment. Danbury Aerospace indicated that FAA guidance material
does not define this condition as ``hazardous'' in the certification
process.
Response. We interpret the comment to be that the FAA has no
definition of hazardous event that includes loss of one cylinder in a
six-cylinder engine, within the engine certification regulations (14
CFR part 33). We agree. The certification process does not define
``hazardous events.'' The FAA establishes through the engine
certification process the minimum standards that an engine needs to
meet to be considered airworthy. For example, Sec. 33.19 establishes
durability standards that are designed to minimize the development of
an unsafe condition between overhaul periods. These minimum safety
standards must also be met by PMA parts, either through establishing
identicality or through test and computation. FAA Policy PS-ANE100-
1997-00001, provides guidance for the certification of PMA applications
for reciprocating engine critical, highly stressed or complex parts,
including, but not limited to crankshafts and cylinder heads. We did
not withdraw the August 28, 2015, SNPRM.
Comment. RAM Aircraft commented that it has run tests that
substantiate and document the power loss as a ``minor power loss'' in
the event of a cylinder separation.
Response. We interpret the comment to be that any power loss from
cylinder head separation is only minor. We disagree. The loss of one
cylinder's power would equate to approximately a 17 to 20% reduction in
engine horsepower output. Further, loss of a cylinder at critical
phases of flight, for example, during climb-out where like here, the
failure is at increased probability of occurring, produces a power loss
sufficient to result in a 40% reduction in airplane rate of climb. This
would constitute a hazardous condition during a critical phase of
flight like departure/climb. We did not withdraw the August 28, 2015,
SNPRM.
Comment. RAM Aircraft suggested that this minor power loss would be
classified as a ``minor hazard,'' based on guidance from the FAA's
``Policy Statement on Risk Assessment for Reciprocating Engine
Airworthiness Directives'' (PS-ANE100-1999-00006). According to the FAA
policy statement, minor hazards are candidates for AD action only when
the probability of the event is very high.
Response. We disagree. FAA policy classifies service problems that
do not result in a significant power loss, such as a partial power
loss, rough running, pre-ignition, backfire, single magneto failures,
as ``minor.'' We found that cylinder separations results in a 17 to 20%
reduction in engine horsepower output results in an approximately 40%
reduction in airplane excess power, which translates into a 40%
reduction in airplane rate of climb. This constitutes a hazardous
condition that is not a ``minor hazard.'' We did not change this AD
based on this comment.
Comment. RAM Aircraft commented that Appendix VI of the SAD
Airworthiness Directives Manual Supplement includes examples of
conditions that potentially have a ``minor'' affect. The loss of one
engine (multi-engine aircraft) is listed as a condition with a
``minor'' effect. Given the ``minor'' effect of the loss of one engine
and the likelihood of the cylinder separation being extremely remote,
then this AD should not be issued against multi-engine aircraft.
Response. We disagree. By comparing the risk analysis computed by
the CSTA for Aircraft Safety Analysis with either the Small Airplane
Risk Analysis guidelines used by the SAD or the Engine and Propeller
Directorate (E&PD) Continued Airworthiness Assessment Process (CAAP)
Handbook guidelines, demonstrates that an AD is needed for both single
and twin-engine aircraft. We did not withdraw the August 28, 2015,
SNPRM.
Comment. RAM Aircraft commented that they are not aware of any
substantiated fact of a ``fire,'' or any other significant consequence
of a cylinder head separation. Further, RAM Aircraft noted that in its May
12, 2014, comment, it had documented the research it had done to refute
the ``rumor'' of a fire resulting from a cylinder head separation of an
ECi cylinder.
Response. We disagree. RAM Aircraft itself submitted data to the
FAA indicating that a fire could occur from cylinder head separation.
FAA requested to see that information. FAA's subsequent visit to RAM
Aircraft confirmed that a failed cylinder caused an in-flight fire on a
Cessna 414 airplane. We did not withdraw the August 28, 2015, SNPRM.
Comment. Danbury Aerospace cited FAA documents that indicate that
the design of an aircraft engine, for reciprocating engines, should
incorporate mitigating features. For example, Danbury quoted SAD
Standards Staff (ACE-110) Memorandum, dated May 6, 1986, and an E&PD
Standards Staff (ANE-110) memorandum, dated May 24, 1997.
Response. We agree. However, the regulatory requirement for a
designer to mitigate a possible reciprocating engine failure prior to
certification is different than correcting an unsafe condition found to
exist after certification. This AD addresses an unsafe condition--
cylinder head separation, found after certification. A regulatory
requirement to mitigate in the aircraft design an engine failure is not
the subject of this AD. We did not withdraw the August 28, 2015, SNPRM.
Comment. IPL Group commented that we were misusing the term
``catastrophic'' when describing the effects of potential cylinder
failures.
Response. We disagree. As to the use of ``catastrophic,'' we did
not use the term in the August 12, 2013, NPRM, the two SNPRMs, or in
this final rule AD. We did not change the August 28, 2015, SNPRM based
on this comment.
Comment. IPL Group argued that a cylinder head separation does not
cause an unsafe event and that there is ``zero evidence'' in Docket No.
FAA-2012-0002 to support the showing that a failed cylinder causes an
unsafe condition.
Response. We disagree. Cylinder separations can cause partial or
complete engine failure which can cause a subsequent loss of power and
control of the airplane. Loss of control of the airplane may result in
the loss of the airplane and injuries or death. Additionally, we note
the NTSB has stated that cylinder head separations could result in a
loss of control of the airplane (see NTSB's comment to ``Docket No FAA-
2008-0052; Directorate Identifier 2008-NE-01-AD, dated September 25,
2009''). We did not withdraw the August 28, 2015, SNPRM.
Comment. Danbury Holdings commented that the FAA had not provided
any information to substantiate the FAA's position that cylinder
separations have a ``significant'' effect on airplane safety or that
cylinder separations would result in a fire.
Response. We disagree. The impact of a cylinder separation in-
flight is an unacceptable compromise to safety. To clarify this point,
we changed the AD to use ``unacceptable.'' We disagree that cylinder
head separations might not result in fire. Cylinder separations can
result in engine failure and/or fire. As an example, on November 29,
1987, a Piper PA-46 airplane experienced a cylinder head separation
followed by an in-flight fire. We did not withdraw the August 28, 2015,
SNPRM.
Comment. Danbury Holdings also stated that the FAA did not issue a
similar AD against the OEM cylinder assemblies because the OEM
manufactured more such cylinder assemblies.
Response. We disagree. The FAA did not mandate actions similar to
those specified in this AD against the OEM cylinders because the OEM
cylinders do not have the inherent design deficiencies that the ECi PMA
cylinders have. Also, the service history of the OEM cylinders
indicates that the OEM separation rate is approximately 32 times lower
than the ECi cylinders. We did not withdraw the August 28, 2015, SNPRM.
Comment. Danbury Holdings further commented that ADs are never
justified for any cylinder manufacturer.
Response. We interpret the comment as suggesting that we should not
issue an AD when engine design deficiencies related to cylinders are
found. We disagree. Cylinders are engine parts whose structural failure
can result in a degradation to or total loss of, engine power output,
and loss of control of an airplane. Cylinder separations aloft can also
cause an in-flight fire. We will exercise our regulatory arm to issue
ADs when we determine doing so is necessary to resolve an unsafe
condition in a product. We did not withdraw the August 28, 2015, SNPRM.
Comment. Danbury Aerospace commented that 14 CFR part 33.43
requires assessment of crankshaft vibration for one cylinder not firing
because the condition is not an engine failure event condition.
Response. We disagree. As we noted in our January 8, 2015, SNPRM,
14 CFR part 33 does not require continued safe operation following a
cylinder separation or following any other engine structural failure.
Section 33.43(d), addressing the engine vibration survey of Sec.
33.43(a), requires assessment of crankshaft vibration for an engine
that has one cylinder that ``is not firing.'' We require vibration
testing with a critical cylinder inoperative because it is a failure
condition where stresses may exceed the endurance limit of the
crankshaft material. We need to know the speed ranges where the
excessive stresses occur so operational information may be provided to
flight crews so they can avoid these speed ranges when a cylinder is
inoperative. We did not withdraw the August 28, 2015, SNPRM.

Request To Withdraw the SNPRMs Because Root Cause of Cylinder Failure
Is Unknown

Comment. Several commenters indicated that the FAA has failed to
identify the root cause(s) of cylinder head separations.
Response. We disagree. We have identified the root cause of
cylinder failure as design deficiencies inherent in the affected ECi
cylinder assemblies. These ECi cylinder assemblies have two inherent
design deficiencies: Insufficient dome radius and insufficient head-to-
barrel interference fit. These design deficiencies are identified in
AEC Technical Report 1102-13, dated April 30, 2011, that we posted to
Docket No. FAA-2012-0002 (see ADDRESSES section of this final rule for
information on locating the docket). We did not withdraw the SNPRMs.
Comment. Danbury Aerospace commented that root cause analysis is
absolutely essential to determining compliance with regulations and if
an unsafe condition has been created. Therefore the agency has not
properly identified the unsafe condition.
Response. We disagree. We identified the unsafe condition in the
engine: Cylinder head separation. The purpose of this AD is to correct
that unsafe condition. We also identified that cylinder head
separations are due to at least two inherent design deficiencies. All
cylinders prior to S/N 33697 have insufficient dome transition radius,
and all cylinders prior to S/N 61177 insufficient head-to-barrel
interference fit. ECi characterized both of these as ``inherent design
deficiencies'' in its AEC Technical Report 1102-13. We did not withdraw
the August 28, 2015, SNPRM.

Request To Withdraw the SNPRMs Because Pilot Error Is Causing Cylinder
Head Separations

Comment. Danbury Aerospace and Danbury Holdings commented that
cylinder head separations involving the ECi cylinder assemblies
affected by this AD were caused by excessive CHTs, presumably caused
pilot error, rather than by design deficiencies of the cylinder
assemblies.
One operator observed that operators who use the ECi cylinder
assemblies and operate them within limits and with good instrumentation
are not having issues. This operator noted that everyone, with the
exception of the FAA, believes that overheating beyond CHT limits by
operators has a direct effect on cylinder head separation.
Response. We disagree. Although pilot error may cause excessive
CHT, we have no data to suggest it is the cause of the unsafe condition
that is the subject of this AD. If pilot error results in excessive
CHT, which leads to cylinder head separation, then we would expect to
see similar damage in engines with other than ECi cylinder assemblies
installed where the pilots exceeded the same limitation(s). However, we
do not have any such data. Also, we have no evidence that either
intentional or inadvertent exceedance of CHT limits has caused cylinder
separation. Further ECi identified several design deficiencies in AEC
Technical Report 1102-13, dated April 30, 2011.We did not withdraw the
SNPRMs.

Request To Withdraw the SNPRMs Because of the Risk of Maintenance
Errors

Comment. Several commenters commented that the FAA should withdraw
the SNPRMs because the removal and replacement of affected cylinder
assemblies before time between overhaul (TBO) would result in
maintenance errors that would adversely affect safety. For example, IPL
Group indicated that replacement of the cylinder assemblies would
likely result in events of main bearings losing clamp-up and turning,
resulting in cylinder through-bolt and flange stud failures, which
would likely result in total engine failure.
Response. We disagree. Our regulatory framework presumes that
maintenance will be performed correctly by experienced personnel
authorized by the FAA to return aircraft to service in an airworthy
condition. Further, we have not observed any negative effects on safety
due to removal of these cylinder assemblies during maintenance. Also,
cylinder removal and replacement is a maintenance action addressed in
engine maintenance manuals. We did not withdraw the SNPRMs.

Request To Justify 80-Hour Removal Requirement for Overhauled Cylinder
Assemblies

Comment. Danbury Aerospace and Danbury Holdings requested that the
FAA provide evidence (including engineering analysis) supporting its
conclusion that overhauled cylinder assemblies should be removed within
80 hours after the effective date.
Response. We interpret the comment to be that the commenters
disagree that the phased removal plan required by this AD is
appropriate. We disagree. This AD mandates a phased removal of affected
cylinders with the intent to retire all affected cylinders by initial
TBO. The FAA recognizes that some cylinders in service may already have
exceeded their initial TBO. Metal fatigue damage is cumulative, and the
longer a cylinder head remains in service, the more likely it will fail
due to one of the inherent design deficiencies. Overhauled cylinders
have likely experienced more load and temperature cycles than lower
time cylinders and the total time in service since new of overhauled
cylinders often cannot be determined. Our determination of 80 hours is
supported by our Challenge Team's findings and our risk analysis that
we uploaded to FAA Docket No. FAA-2012-0002 (see Addresses section of
this final rule for information on locating the docket). We did not
change this AD based on this comment.
Comment. Danbury Aerospace and Danbury Holdings also stated that
the FAA had not substantiated that the overhaul of a cylinder does not
reduce the existing fatigue damage that a cylinder may have incurred
while in service.
Response. We disagree. Fatigue strength of metal alloys operated at
high temperatures continuously decreases with cycles until failure.
This is particularly true for aluminum alloys, including the aluminum
alloy used to cast cylinder heads. Metallic structural elements that
are operated at high temperatures are more susceptible to time
dependent fatigue. The overhaul of a cylinder assembly does not reverse
the fatigue damage that had been previously accumulated in the aluminum
cylinder head casting. We did not change the AD based on this comment.

Request To Revise Applicability

Comment. Danbury Holdings commented that the FAA has no evidence
that all cylinders through S/N 61176 are at risk for separation in the
first thread due to insufficient head-to-barrel interference fit.
Response. We disagree. The SDR database and other field reports
document instances of first-thread failures of cylinders manufactured
to design data applicable to all cylinders prior to S/N 61177. For this
reason, all cylinders through S/N 61176 are subject to the corrective
actions of this AD. We did not change this AD based on this comment.
Comment. One commenter stated that he has an O-470 engine converted
by P. Ponk Aviation to the equivalent of an O-520 engine. He indicated
that those engines should not be affected by this AD.
Response. We disagree. The affected S/N cylinders are installed on
-470 engines, as well as the -520 and -550 engine models. Any engine
that uses one of these affected cylinders is at risk. We have received
at least one report of a separation of these affected S/N cylinders on
-470 engines. Although the unmodified -470 engines have lower engine
horsepower output, their brake mean effective pressure (BMEP) is
actually higher than that of the -520 and -550 engines. BMEP is
proportional to the ratio of horsepower per cubic inch of displacement.
Therefore, the actual operating stresses in the same cylinder wall are
even higher when these same cylinders are installed in an unmodified -
470 engines than it would be for either the -520 or the -550 engines.
The P. Ponk Aviation STC increases the displacement of the unmodified -
470 engine to -520 cubic inches by installing the -520 cylinders on the
-470 engine. Given that no valid sensitivity analysis exists showing
the relationship of BMEP to fatigue life of these cylinders, and since
the crack propagation rate is also unknown, we have included all the -
470 engines, including those modified by the P. Ponk Aviation STC, in
the effectivity of this AD. We did not change this AD based on this
comment.

Request To Adopt Less Stringent Compliance Requirements

Comment. AOPA, RAM Aircraft, as well as operators and private
citizens, requested that we adopt less stringent requirements than
those in the proposed AD. The commenters indicated that the affected
cylinder assemblies should be inspected at regular intervals, but
removed at TBO. For example, one
commenter suggested recurring inspections every 60 hours. Several
commenters cited the NTSB in support of its recommendation. RAM
Aircraft commented that the FAA may be jumping to conclusions by
eliminating these inspections. RAM Aircraft noted that the failure of a
compression/soap test to detect a particular crack in a cylinder
assembly on several occasions does not mean that the test will fail to
detect cracked cylinders on most occasions. By their very nature and
design compromises, i.e., steel barrels to contain the forces of
combustion combined with lighter cylinder head alloys to reduce weight
so that aircraft engines have commercial viability and value, and the
harsh conditions, altitudes, and temperatures in which they operate,
reciprocating aircraft engine cylinders will inevitably crack. RAM
Aircraft indicated that there is no question but that some cylinders
are going to crack, and that therefore, they must be properly operated,
maintained, and inspected.
Response. We disagree. Repetitive inspections until TBO, as
suggested by the commenters, do not adequately address the unsafe
condition in this particular case. Repetitive inspections would not
detect cracks until they have already progressed completely across the
cylinder head wall thickness.
Several operators and mechanics have reported that they
successfully passed the compression/soap test with a partially
separated cylinder. Others have reported that they successfully passed
the compression/soap test and then experienced an in-flight separation
before the next scheduled 50-hour inspection.
Therefore, we conclude that these tests are not sufficiently
reliable. Also, engine overhaul is not a requirement for all operators.
Therefore, tying the proposed recurrent inspection to engine overhaul
would not resolve the unsafe condition. Based on its comment to the
August 28, 2015, SNPRM, we know that the NTSB now considers this rule
consistent with the rationale they have provided in the past in support
of NTSB Safety Recommendation A-12-7 regarding these affected cylinder
assemblies (Reference NTSB Comment FAA-2012-0002-0653, dated September
24, 2015 in Docket FAA-2012-0002). We did not change this AD based on
this comment.
Comment. One commenter indicated it was incorrect to apply the same
requirement to remove the cylinders at specified intervals to different
CMI engine models. He noted, for example, that the TSIO-520-J engine
that is allowed to produce 36 inches of manifold pressure and 310
horsepower will produce less stress on a cylinder head than a TSIO-520-
NB engine that is allowed 41 inches of manifold pressure and 325
horsepower, as installed on a Cessna 414 airplane.
Response. We disagree. Service history indicates that the affected
cylinder assemblies have cracked on -470, -520, and -550 engine models.
The AD, therefore, applies to all affected CMI -470, -520 and -550
engine models. We have no engineering analysis or test data to justify
varying compliance times by engine model or applying the corrective
actions of this AD to only the higher power engines. We did not change
this AD based on this comment.
Comment. Danbury Aerospace observed that the average number of
cylinder assemblies, P/N AEC 631397, in the serial number range in the
January 8, 2015, SNPRM that are still in operation have less than 500
hours left to TBO. Danbury Aerospace indicated that the early removal
of these cylinders is not justified by a statistical analysis developed
in accordance with the E&PD CAAP Handbook.
Response. We disagree. We do not know the exact number of total
hours TIS for each affected cylinder assembly. We have no data to
support the claim that the existing fleet of cylinder assemblies
already has accumulated 1,200 or more hours TIS. Service history also
shows that most of the separations occurred well before initial TBO.
Therefore, removal of the affected cylinder assemblies before TBO is
appropriate. We did not change this AD based on this comment.
Comment. Danbury Holdings commented that the FAA had not provided
evidence that there have been separations within the originally
proposed 50-hour recurrent compression test/soap inspection interval.
Response. We disagree. We received several field reports of
cylinder separations occurring within 50 hours of passing either the
originally proposed 50-hour recurrent compression test/soap inspection
in the August 12, 2013, NPRM. SDR report No. SQP2011F00000 was
submitted by a part 135 operator who operated a Cessna T210N with an
affected ECi cylinder assembly installed. The operator reported that on
September 9, 2011, that affected ECI cylinder head separated at the 5th
cooling fin on-head. At the time of the failure, the engine and failed
cylinder had 817.6 hours time since overhaul/time since new, and its
last compression check inspection was at 19.2 hrs. prior. Other field
reports also document separated cylinders (for example, see SDR Report
2010FA0000179) that recently passed the compression test/soap
inspections. We did not change this AD based on this comment.
Comment. One commenter commented that, based on his experience, ECi
has an aluminum head cracking issue and that these cylinders seem to
crack more than CMI cylinders. The commenter further indicated that he
believed the number of cylinder failures is underreported in the SDR
database. The commenter further noted that in his 30 plus years of
aircraft maintenance experience, he has never seen a cylinder failure
rate this high. The commenter welcomed an AD that requires these
cylinders to be inspected at around 100 hours and the reports of cracks
sent to an FAA database.
Response. We note the comment. We agree that the ECi failure rate
is much higher than the OEM failure rate over the same field service
period and that cylinder cracks are under-reported. For example, many
of the RAM failures listed in the docket were not reported under the
SDR system or as required by 14 CFR 21.3. We did not change this AD
based on this comment.
Comment. RAM Aircraft commented that, based on its previous
comments, the FAA should withdraw the SNPRMs. RAM Aircraft recommended
that the FAA consider education and requiring inspections of all
reciprocating airplane engine cylinders on the terms and conditions the
FAA determines to be appropriate.
Response. We disagree. Our analysis indicates that an AD is
required to resolve the unsafe condition presented by installed
affected ECi cylinder assemblies. We did not withdraw the SNPRMs based
on this comment.
Comment. One commenter suggested that users of a JPI or other
engine monitoring system should be subject to a different compliance
interval.
Response. We disagree. As noted previously, the root cause of these
cylinder failures are design deficiencies. The affected cylinders may
fail without overtemping. Therefore, use of an engine monitoring system
like JPI would be insufficient to detect the unsafe condition. We did
not change this AD based on this comment.

Request To Use Mandatory Service Bulletin Instead of This AD

Comment. One commenter requested that the FAA use a mandatory
service bulletin instead of this AD to implement corrective action.
Response. We disagree. Requiring a manufacturer to issue a
mandatory service bulletin is outside the scope of
the FAA's authority. We did not change this AD based on this comment.

B. Comments to the FAA's Risk Assessment Processes and Polices

Request That the FAA Follow Its Own Risk Assessment Policies and
Guidance

Comment. Multiple commenters, including Danbury Aerospace, Danbury
Holdings, and the Aeronautical Repair Station Association (ARSA)
commented that the FAA did not follow its own risk assessment policies
and guidance, such as FAA Order 8110.107A, Monitor Safety/Analyze Data
(MSAD), dated October 1, 2012, and FAA Order 8040.4A, Safety Risk
Management Policy, dated April 30, 2012, and the E&PD CAAP Handbook,
dated September 23, 2010.
Response. We interpret this comment as a comment that we failed to
follow FAA Order 8110.107A, FAA Order 8040.4A, and the CAAP Handbook.
We disagree. We performed the process as required by FAA Order
8110.107A, Monitor Safety/Analyze Data (MSAD), dated October 1, 2012,
to analyze data and determine corrective action for continued
operational safety issues. We acquired the failure event data from the
MSAD, SDR, NTSB databases, ECi, and outside sources. We conducted a
hazard criteria analysis where we filtered the data to identify
relevant events. We performed a qualitative preliminary risk assessment
and determined that this safety problem required corrective action. We
performed risk analyses in conjunction with the E&PD risk assessment
criteria. We identified that the ECi model separations have two
inherent design deficiencies: Insufficient dome radius and insufficient
head-to-barrel interference fit. Finally, we coordinated with our
Corrective Action Review Board, which determined and agreed to the
proposed corrective action in our August 12, 2013, NPRM.
Later, as part of the Challenge Team's meeting in September, 2014,
the CSTA for Aircraft Safety performed a risk analysis that confirmed
the need for this AD and shaped its compliance plan. We compared the
results of the CSTA's risk analysis to the guidelines used by the SAD
in its SARA and to the guidelines in the E&PD's CAAP Handbook and
determined that an AD is required.
FAA Order 8040.4A requires a risk assessment methodology as
outlined in the Order. FAA Order 8040.4A notes that the safety risk is
a composite of two factors: The potential ``severity'' or worst
possible consequence(s) or outcome of an adverse event that is assumed
to occur, and also the expected frequency of occurrence or likelihood
of occurrence (failure rate) for that specific adverse event. Each of
these factors is assessed independent of the other and then entered as
separate inputs into a risk matrix that yields an overall level of risk
for the event.
We performed the risk assessment required by FAA Order 8040.4A and
concluded that this AD was necessary. Therefore, our August 12, 2013,
NPRM, as revised by the January 8, 2015 SNPRM, and as republished on
August 28, 2015, are consistent with FAA Order 8040.4A, FAA Order
8110.107A, and the CAAP Handbook. We did not change this AD based on
this comment.
Comment. Commenters, including Danbury Holdings, commented that the
FAA should not have included the failure rate of the affected ECi
cylinders in the FAA risk assessments that were used to substantiate
the need for the corrective actions in this AD. Danbury Holdings
indicated that the failure rate is irrelevant to the unsafe condition.
Response. We disagree. We did not use the failure rate in the risk
analysis, however, we used the number of reported failures. A risk
analysis involves using past data; both successful operation as well as
failures (including cracks), to develop a relationship between part
parameters, including age and usage, and risk of failure. Therefore,
our use of failures was appropriate in this risk analysis. We did not
change this AD based on this comment.
Comment. Danbury Aerospace commented that the FAA ignored its own
standards for what constitutes an unsafe condition and therefore has
failed to identify one.
Response. We disagree. The FAA followed its standard risk analysis
processes in determining that the unsafe condition represented by the
affected ECi cylinder assemblies exists. 14 CFR part 39 prescribes that
we issue an AD when an unsafe condition exists in a product and that
condition is likely to exist or develop in other products of the same
type design. We did not change this AD based on this comment.
Comment. Danbury Holdings commented that the basis for the FAA's
risk analysis is seriously flawed because the unsafe condition must be
the basis for the failure, not one unsubstantiated fatality.
Response. We disagree. The unsafe condition in the engine presented
by the presence of affected ECi cylinders is the basis of this AD. We
did not change this AD based on this comment.
Comment. Danbury Holdings further commented that the FAA had failed
to establish a connection between the cylinder separation issue
addressed by this AD and the official reports of the two fatal
accidents that the FAA references.
Response. We disagree. Reports by the Bahamas Department of Civil
Aviation and the NTSB establish that these accidents in the Bahamas and
in Swanzey, New Hampshire involved separated ECi cylinders (see Report
AAIPU# A10-01312 and NTSB Accident Report No. NYC02FA178,
respectively). We have determined that the separation of the affected
ECi cylinder assemblies represents an unsafe condition. We are not
required to establish any further connection with these accidents. We
did not change this AD based on this comment.
Comment. Danbury Holdings added that the FAA should not have
included the fatal accident in the Bahamas in the FAA's risk
assessments because the NTSB full narrative for that accident
(ERA11WA008) made no mention of a cylinder separation.
Response. We interpret the comment as the fatal accident in the
Bahamas is not relevant to this AD. We disagree. As noted in the
previous comment response, we have determined that the separation of
the affected ECi cylinder assemblies, as occurred in the accident in
the Bahamas, represents an unsafe condition. We did not change this AD.
Comment. Danbury Holdings also stated that the root cause of the
other fatal accident, the Swanzey, New Hampshire, accident (see NTSB
Accident Report No. NYC02FA178) that the FAA included in its risk
assessments was unsafe and improper operation of the airplane by the
pilot not cylinder separation.
Response. We disagree. As noted in the preceding comment
discussion, we have determined that the separation of the affected ECi
cylinder assemblies, as occurred in the accident in Swanzey, New
Hampshire, represents an unsafe condition and is therefore relevant to
this AD. We did not change this AD based on this comment.
Comment. Danbury Aerospace added that the accident in the Bahamas
should not be included in the FAA's risk analysis because: (1) It did
not concern a U.S.-registered aircraft and therefore cannot be used in
this rulemaking; (2) loss of control and uncontrolled flight was cited
as the cause; and (3) even if the accident could be included, it does
not meet hazard level thresholds required for rulemaking.
Response. The commenter presents three comments, which have three
parts. We disagree with all three parts. As to part one, the Bahamas
accident involved a U.S.-type certificated product, an engine with
affected ECi cylinders installed. Therefore, the product is the proper subject of
this AD. As to part two, the accident involved an engine with an ECI cylinder
separation, a failure of a part of the engine, during flight. A
cylinder separation during flight represents an unsafe condition in the
engine. Therefore, our action in issuing this AD is appropriate. As to
the part three, the cylinder failure presented a hazard to the engine
and an unsafe condition, and therefore, meets the threshold for an AD.
The need for this AD was confirmed by comparing the result of the risk
analysis to the guidelines in the SAD's SARA and the E&PD's CAAP
Handbook. We did not change this AD based on this comment.

Request That the FAA Define Guidelines Used To Define an Unsafe
Condition

Comment. Danbury Holdings commented that the FAA had not defined
the guidelines that it used to establish the existence of an unsafe
condition.
Response. We interpret the comment to be a request to identify what
guidance defines an unsafe condition. The comment therefore, is not to
the technical merits of this AD, but a request for general guidance. As
such, a response is unnecessary per the Administrative Procedures Act
(APA), and we recommend that the commenter seek his answer through a
direct request to the FAA Aircraft Certification Service or Flight
Standards Division. We did not change this AD based on this comment.

Request To Withdraw the August 28, 2015, SNPRM Because Supporting
Documents Do Not Support Issuing This AD

Comment. Danbury Holdings commented that the documents provided by
the FAA in Docket No. FAA-2012-0002 do not support issuance of this AD.
The supporting documents referred to by Danbury Holdings are: (1) The
risk analysis conducted by the FAA's CSTA for Aircraft Safety Analysis;
(2) a risk analysis using the Small Airplane Risk Analysis (SARA)
methods; (3) a June 2011, presentation by Airmotive Engineering to the
FAA concerning its ECi cylinder assemblies; (4) a list of ECi cylinder
assembly failure reports consisting of only those reports where both
cylinder serial number and time in service are included in the reports;
(5) a list of additional failures of ECi cylinder assemblies reported
by a maintenance organization; and (6) Airmotive Engineering
Corporation Technical Report 1102-13, dated April 30, 2011.
Response. We disagree. The supporting documents that Danbury
Holdings referred to, identified above, support that the FAA followed
its process and were used to help determine that an unsafe condition
exists. We have also uploaded additional documents to Docket No. FAA-
2012-0002 on June 23, 2015 (see ADDRESSES section of this final rule
for information on locating the docket).
The risk analysis performed by the FAA's CSTA for Aircraft Safety
Analysis, recommends removal and replacement of the affected ECi
cylinder assemblies as specified in this AD. The SARA applied to
failures of ECi cylinder assemblies confirms that an AD is necessary.
AEC Technical Report 1102-13 states that a root cause for the first
thread separations was an inherent design deficiency in the form of
insufficient head-to-barrel design interference fit. AEC Technical
Report 1102-13 recommended withdrawing these cylinder assemblies from
service. We did not withdraw the August 28, 2015, SNPRM.
Comment. Danbury Holdings commented that that the FAA's risk
analyses and other technical information were ``flawed, improperly
applied, and replete with unsubstantiated conclusions.''
Response. The commenter failed to provide any examples of FAA
technical information that was flawed, improperly applied, or replete
with unsubstantiated conclusions. Without those details, we are unable
to consider the comment as having technical merit. Accordingly, we
interpret the comment as a general objection to the need for the AD. We
disagree. Our Challenge Team applied the risk assessments by the FAA's
CSTA for Aircraft Safety Analysis, against the SAD's SARA guidelines
and the E&PD's CAAP guidelines and independently concluded that an AD
is required to mitigate the unsafe condition presented by installed
affected ECi cylinder assemblies. We presented both risk assessments in
Docket No. FAA-2012-0002 (see ADDRESSES section of this final rule for
information on locating the docket). We did not change this AD based on
this comment.
Comment. Danbury Holdings commented that it found no relationship
between the risk analysis using SARA methods and any analysis or
conclusion provided by the agency in this rulemaking. We interpret
Danbury Holding's comment as suggesting that no relationship exists
between the risk analysis using SARA methods and any analysis or
conclusion provided by the agency in this rulemaking.
Response. We disagree. In comments to the August 12, 2013, NPRM
some commenters requested that we use the SARA to determine if an AD
was warranted. We used the SARA, and it confirmed the need for an AD.
We did not change this AD based on this comment.
Comment. Danbury Holdings commented that RAM Aircraft had
concluded, through its own risk analysis, that ``the probability of a
cylinder separation is extremely remote'' and that ``historical data
and information thus far evident leads to the conclusion that there has
been no physical discomfort to pilots or passengers and no damage to
any aircraft as a result of the subject cylinders.''
Response. We interpret the comment as two parts; first, that our
risk assessment reached a wrong conclusion, and second, that a cylinder
head separation does not result in any discomfort to pilots or
passengers, or damage to the aircraft. We disagree. FAA Order 8040.4A,
``Safety Risk Management Policy'', dated April 30, 2012, FAA Order
8110.107A, ``Monitor Safety/Analyze Data,'' and the guidance in Engine
& Propeller Directorate memorandum ``Risk Assessment for Reciprocating
Engine Airworthiness Directives,'' PS-ANE-100-1999-00006, dated May 24,
1999, direct how we do a risk assessment.
We analyze safety risk, per FAA Order 8040.4A, as a composite of
two factors: The potential ``severity'' or worst possible
consequence(s) or outcome of an adverse event that is assumed to occur,
and also the ``expected frequency of occurrence'' for that specific
adverse event. FAA Order 8040.4A directs us to assess both factors
independently, then enter each as separate inputs into a risk matrix.
The matrix yields an overall level of risk for the event. The overall
risk is then categorized as either ``Unacceptable Risk,'' ``Acceptable
Risk with Mitigation,'' or ``Acceptable Risk.'' The corrective
action(s), if any, is driven by the assessed overall risk. Table C-1 of
Appendix C of FAA Order 8040.4A defines five levels of severity and
Table C-2 defines five levels of event frequency that are used in the
determination of composite risk.
The FAA classification for the ``severity'' of an engine cylinder
head separation event, per FAA Order 8040.4A, is ``hazardous'' for both
single-engine and light-twin airplanes for several reasons. Cylinder
head separations can significantly reduce the power of the airplane
such that under some conditions it may not be able to safely takeoff
and climb out. It could also create a dangerous asymmetric thrust condition
for twin-engine airplanes. If the separation occurs in cruise flight, the airplane
may have insufficient excess power to continue safe flight at any altitude.
Cylinder head separations have also caused in-flight fires. These are
all unsafe conditions that warrant a ``hazardous'' severity level for
risk assessment purposes.
Table C-2 in FAA Order 8040.4A defines ``extremely improbable'' as
``So unlikely that it is not expected to occur, but it is not
impossible.'' It defines ``extremely remote'' as ``Expected to occur
rarely.'' It defines ``Remote'' as ``Expected to occur infrequently.''
It defines ``probable'' as ``Expected to occur often.'' Finally, it
defines ``frequent'' as ``expected to occur routinely.''
Service history failure reports indicate that in a population of
43,000 cylinders, that 1 of every 1,000 cylinders could separate on
average; either in the dome radius or the first thread. A single-engine
airplane has six of these cylinders, so the actual risk of a separation
of any one of those six cylinders for any given airplane is 6/1,000: 1
of every 166 engines. Similarly, a twin-engine airplane will have 12
cylinders, so the risk of experiencing a separation of one cylinder on
a twin-engine aircraft is twice that of a single engine, 12/1,000, 1 of
every 83 twin-engine airplanes that use these model cylinders.
Separation event under-reporting occurs. This is evidenced by RAM
Aircraft's submittal of 23 additional reported failures of the subject
ECi cylinders after the August 12, 2013 NPRM was issued. Photos of
these failures are available in Docket No. FAA-2012-0002 (see ADDRESSES
section of this final rule for information on locating the docket). The
calculated separation rate, therefore, is likely higher than what we
used in our analysis. Also based on service experience, we expect more
ECi cylinder head separations in the future. Therefore, we concluded
that the most appropriate assessment for the frequency of occurrence
for these cylinder separations is ``Remote C''; ``Expected to occur
infrequently.''
Figure C-1 of FAA Order 8040.4A is a risk matrix that yields an
overall risk based on the severity classification and the assessed
frequency of occurrence. Using the FAA severity classification of
``hazardous'' and the FAA assessed frequency of occurrence ``Remote
C'', yields an overall risk that is ``unacceptable.'' The corrective
actions required by this final rule AD are based on and consistent with
this overall risk assessment.
We, therefore, disagree with claims by RAM Aircraft and other
commenters that a cylinder head separation will have a negligible
effect on airplane safety. Also, several documented in-flight fires
were precipitated by a cylinder head separation. We did not change this
AD based on this comment.
Comment. Danbury Holdings also commented that AEC Technical Report
1102-13 was ``disavowed'' by AEC [now CMI San Antonio] since it was
obtained under questionable circumstances and has since been proven
incorrect given its predictions did not come to fruition.
Response. We disagree. AEC originally provided the analysis to the
FAA when it was considering a service bulletin for the affected ECi
cylinder assemblies. ECi requested the FAA return or destroy ECi
Technical Report 1102-13 after they learned the FAA was considering an
AD. We found the data in this report useful in our determination of an
unsafe condition. We did not change this AD based on this comment.
Comment. Danbury Holdings commented that the FAA has not
substantiated that the affected ECi cylinder assemblies have separated
at 32 times the rate of the OEM cylinders. Danbury Holdings stated that
the FAA had not provided any supporting documentation to substantiate
the FAA's estimate that the OEM has produced approximately 4 times as
many cylinders as ECi did over the same period of time. Danbury
Holdings further commented that that the FAA ignores separations of
other cylinder manufacturers.
Response. We disagree. We uploaded supporting information,
including service history, to Docket No. FAA-2012-0002 (see ADDRESSES
section of this final rule for information on locating the docket). We
determined the comparative failure rate of affected ECi cylinders to
OEM cylinders through comparing the service history of ECi cylinders to
the OEM's since ECi received their PMA. That service history shows that
the affected ECi cylinders have had approximately eight times as many
separations over the same period of time as that of the OEM cylinders;
since ECi received its PMA. We compared the affected ECi cylinder
production rates from ECi, to that of the OEM since ECi received its
PMA. From ECi's and the OEM's production information, we concluded that
the OEM has produced approximately four times as many cylinders as ECi
since ECi received its PMA. This yields a normalized failure rate that
is approximately 8 (comparative cylinder failure rate) x 4 (comparative
production rate), which showed an overall failure rate 32 times higher
for ECi cylinders.
Since we first published that rate information, we subsequently
learned of more failures of affected ECi cylinders. Those additional
failures would, if included, increase the ECi failure rate. We did not
update the failure rate to higher than 32 times that of the OEM's
because it did not affect our decision regarding this AD. We did not
change this AD based on this comment.
The FAA has also issued ADs against other cylinder manufacturers,
including mandating early retirement of cylinders to preclude cracking
and separation. For example, ADs 2014-05-29 and 2007-04-19R1 both apply
to certain Superior Air Parts cylinder assemblies. We did not change
this AD based on this comment.
Comment. Danbury Holdings also commented that the FAA failed to
place all information in its purview into the docket and that the
agency had failed to link its analyses to verifiable data.
Response. We disagree. As previously noted, we have uploaded the
relevant documents used in the decision-making process of this AD in
Docket No. FAA-2012-0002 (see ADDRESSES section of this final rule for
information on locating the docket). Our analysis shows that the FAA's
actions are based on the data that we included in the docket. Our
analysis is therefore linked to ``verifiable data.'' We did not change
this AD based on this comment.
Comment. Danbury Holdings commented that the FAA had failed to
provide any evidence that cylinder separations have resulted in engine
failures, in-flight shutdowns, and/or loss of control of an airplane
and that the agency had included accidents that were not the direct
result of a cylinder separation.
Response. We disagree. A cylinder separating from its engine is an
engine failure. We did not change this AD based on this comment.
Comment. RAM Aircraft commented that it assumes that the failures
of ECi cylinder assemblies shown in the supporting document titled
``ECi AD--Additional Failures Reported by RAM Aircraft'' are based on
letters RAM Aircraft sent to the FAA in 2013. RAM Aircraft, therefore,
commented that this is not new information since the issuance of the
January 8, 2015, SNPRM. Also, of the 38 photographs of damaged cylinder
assemblies, RAM Aircraft noted that only 23 failures actually represent
ECi cylinder assemblies.
Response. We partially agree. First, we agree that the failed
cylinder assemblies identified in the supporting document ``ECi AD--Additional
Failures Reported by RAM Aircraft'' do not represent new information
since the issuance of the January 8, 2015, SNPRM. These failures are
not represented in the SDR database but are consistent with our view
that failures of these cylinder assemblies are under-reported.
Second, we agree that some of the cylinder photographs uploaded to
the docket are not cylinder assemblies affected by this AD. The FAA
sent a letter to RAM Aircraft specifically requesting any information
that RAM Aircraft had relative to failures of ECi cylinder assemblies,
P/N AEC 631397, after we learned of possible failures that had not been
reported as required by 14 CFR 21.3. RAM Aircraft responded to this
request with the photographs and data that we uploaded into Docket FAA-
2012-0002 (see ADDRESSES section of this final rule for information on
locating the docket). These photographs did not have any effect on our
decision to issue this AD. We did not change this AD based on this
comment.

Request To Describe FAA's Validation Process

Comment. Danbury Holdings requested that the FAA provide a
description of the validation process that was used for each of the
cylinder separations that the FAA used to substantiate the need for
this AD.
Response. We interpret this comment as a request for identification
of how we found out about the failures of ECi cylinder assemblies. We
found out about the ECi cylinder assembly failures from the FAA SDR
database and warranty information at ECi and RAM Aircraft, and failure
reports from operators. Many of the operator SDR reports contained
detailed information describing the nature and specific location of the
separation. The findings of ECi Technical Report 1102-13 agreed with
the original failure reports. We did not change this AD based on this
comment.

C. Comments to the FAA's Rulemaking Processes

Request To Follow the APA

Comment. IPL Group, RAM Aircraft, and Danbury Holdings commented
that the FAA had failed to follow the requirements of the APA when it
dispositioned previous comments to the August 12, 2013, NPRM, and the
January 8, 2015, SNPRM. IPL Group indicated that the FAA had, for
example, summarily discounted previous comments, failed to conduct
appropriate investigations of the failed cylinder assemblies, and
mischaracterized hazard levels in the proposed ADs.
RAM Aircraft also commented that its previous comments were
dispositioned in general categories in the January 8, 2015, SNPRM. RAM
Aircraft, however, does not believe that the specifics of its comments
were adequately or properly responded to, as required by the APA.
Response. We disagree. The commenters failed to provide any
examples of where we failed to comply with the APA in our handling of
comments to the August 12, 2013, NPRM, and by extension, the January 8,
2015, and August 25, 2015, SNPRMs. We have in our responses to the NPRM
and the SNPRMs, and herein in this final rule, fully responded to all
comments, including those comments concerning our investigation of the
unsafe condition, hazard levels, and conclusions.
We carefully considered all comments we received. In our January 8,
2015, SNPRM and August 28, 2015, SNPRM, we responded to several hundred
comments that we had received. Many were substantively the same and,
therefore, as previously discussed we grouped them into several
categories and answered the comments by category. The commenters have
not indicated what, if anything, is improper about doing so nor how
doing as we did might have violated the requirements of the APA. In
this final rule, we responded to all remaining comments. We again used
categories to group and answer comments that were similar if not
identical. As to improperly recognizing affected ECi cylinder
assemblies, we based our applicability of this AD on the reports of
failure provided by ECi, the manufacturer, the reports required by 14
CFR that form the basis for the SDR, and the reports of the commenters
themselves. We did not change this AD based on this comment.

Request To Withdraw the SNPRMs Because They Are Arbitrary and
Capricious

Comment. Danbury Holdings and ARSA referred to the proposed rule as
``arbitrary and capricious'' because it does not apply equally to
cylinder assemblies manufactured by the OEM. Danbury Holdings observed
that the OEM's cylinders also separate and that the FAA has singled out
ECi with this AD action.
Response. We disagree. The FAA is not mandating similar corrective
actions against the OEM's cylinders because OEM service history data is
different. Our review of OEM service history indicates that OEM
cylinder assembly failures, unlike ECi cylinder assembly failures, are
not traceable to any specific design or manufacturing anomaly. In
contrast, the ECi PMA cylinder separations are traceable to design
deficiencies, which ECi itself identified in ECi Technical Report 1102-
13. We did not find the ECi cylinder assembly design deficiencies in
cylinder assemblies produced by any other manufacturer. Further, ECi's
failure rate is some 32 times greater than the OEM's. We did not change
this AD based on this comment.
Comment. ARSA also indicated the rule is arbitrary and capricious
because the FAA has failed to ``examine the relevant data and
articulate a satisfactory explanation for its action including a
`rational connection between the facts found and the choice made.''
Further, ARSA cites the APA as requiring federal agencies to allow
meaningful public participation in the rulemaking process and provide a
``statement of basis and purpose'' justifying a rule's issuance.'' ARSA
notes the obligation of the FAA to demonstrate a sound factual basis
for the issuance of a rule by specifically disclosing to interested
parties the material upon which a prospective rule would be fashioned.
Response. We disagree. Beyond its generalized allegation, the
commenter did not identify any examples of agency shortcoming. We
examined the relevant data, including the failure rate of the ECi
assemblies, the ECi cylinder assembly design deficiencies, and the
consequences to the engine and airplane when an ECi cylinder assembly
failed. We reviewed and applied the applicable FAA Orders and policies.
The agency therefore, has articulated a satisfactory explanation
for its action including a ``rational connection between the facts
found and the choice made.''
We provided the public several opportunities to participate in this
rule making; through extending the comment period and the two
supplemental notices with their comment periods. For example, we first
published an NPRM on August 12, 2013 (78 FR 48828), then published an
extension of the comment period on September 26, 2013 (78 FR 59293) to
allow the public additional time to comment on the proposed rule. We
then issued a notice of availability of an initial regulatory
flexibility analysis on March 12, 2014 (79 FR 13924). We reviewed the
over 500 comments to the proposed rule that we had received, determined
that we needed to review how we proposed to address the unsafe
condition, formed a team to review the technical basis of the
proposed rule, the numerous public comments, and the additional failure
information provided by commenters to the NPRM. Through this team we
confirmed that an AD is needed to correct the unsafe condition
represented by the subject cylinder assemblies installed in aircraft
engines, but that we could do so through a lengthier compliance
interval. We published that revised compliance interval in our January
8, 2015, SNPRM.
After publication of the January 8, 2015, SNRPM, we issued the
August 28, 2015, SNPRM to allow us to explain the rationale for this AD
action. We also added several documents to Docket No. FAA-2012-0002
(see ADDRESSES section of this final rule for information on locating
the docket), including the risk analyses by our CSTA for Aircraft
Safety Analysis, and one using SARA methods, and various technical
documents that list failures of ECi cylinder assemblies. For each of
the documents we published, we allowed the public an opportunity to
provide comments. We did not change this AD based on this comment.
Comment. ARSA also commented that presentation of relevant comments
is further stymied by the agency's conclusory and unsupported responses
to the NPRM submissions. ARSA commented that the agency stated that it
was irrelevant that the root cause of the cylinder failures is unknown
and that it ``disagreed'' that pilot error was a factor.
Response. We disagree. The purpose of this AD is to remove an
unsafe condition in aircraft engines, not to identify root cause of
cylinder failure(s). This AD resolves the unsafe condition by removing
the affected cylinder assemblies from service in the engine models
listed in this AD. We did not change this AD based on this comment.
Comment. Danbury Holdings also commented that the FAA had not
provided substantiation for a change in the design requirement that
ensures safe operation with one inoperative cylinder.
Response. The comment is not germane to this AD. We direct the
commenter to the regulations relevant to design requirements, as found
in 14 CFR. We did not withdraw the August 28, 2015, SNPRM.
Comment. Danbury Holdings commented that the FAA has admitted that
the SDR database is problematic and that the FAA picked and chose data
to fit a conclusion.
Response. We disagree. The SDR database reflects input received
from field reporting. The SDR database may not reflect all service
difficulty problems with affected ECI cylinder assemblies, but what
information it contains indicates the need for this AD. Moreover, the
SDR database is only one tool in our decision-making process. We did
not change this AD based on this comment.
Comment. Several commenters commented that the FAA should withdraw
the January 8, 2015, SNPRM because it unfairly targets ECi.
Response. We disagree. This AD does not ``target'' ECi, the PMA
manufacturer of the affected cylinder assemblies. The AD resolves an
unsafe condition in a product. We did not change this AD based on this
comment.

Request To Substantiate That This AD Does Not Affect Airplanes Operated
by Federal or State Agencies

Comment. Danbury Holdings commented that the FAA had not provided
documentation to substantiate that no affected airplanes are operated
by federal or state agencies.
Response. The comment is not relevant to whether this AD is
necessary to resolve the unsafe condition presented by the engine with
the affected ECi cylinders installed. We did not change this AD based
on this comment.

Request To Substantiate That Airplanes Operating in Alaska Are Not
Affected

Comment. Danbury Holdings stated that the FAA had not provided
documentation that substantiates that remote locations of Alaska are
not served by airplanes affected by this AD.
Response. The comment is not relevant to the technical basis for
this AD. Further we state that this AD will not affect intrastate
aviation in Alaska to the extent that it justifies making a regulatory
distinction. We did not change this AD based on this comment.

Request To Send Proposed Rule to Office of Information and Regulatory
Affairs (OIRA) and Small Business Administration (SBA)

Comment. Danbury Aerospace commented that per Executive Order
(E.O.) No. 13272, the FAA should provide the draft rule to the OIRA in
the Office of Management and Budget (OMB) under E.O. No. 12866 and to
the SBA's Chief Counsel for Advocacy.
Response. We partially agree. We do not agree that this rule meets
the criteria of a significant regulatory action under E.O. 12866.
Therefore, we did not provide the draft rule to the OMB. We agree that
the rule has a significant effect on a substantial number of small
entities. We, therefore, provided a copy of the rule to the SBA's Chief
Counsel for Advocacy for comment. We received no comments from the SBA.

D. Comments to the Cost of This AD

Request To Revise and Provide Supporting Data for Number of Affected
Cylinder Assemblies and Engines

Comment. Danbury Aerospace and RAM Aircraft indicated that the FAA
has under-estimated the numbers of airplanes and engines affected and
up to 11,000 aircraft may be affected based on the aircraft registry,
or otherwise hasn't provided the data it used to determine the affected
population of engines and cylinders.
Response. We disagree in part. We do not agree that 11,000 aircraft
may be affected by this AD, or that we haven't provided the data used
to determine the affected populations. Not all aircraft and engines on
the aircraft registry use the affected ECi cylinder assemblies.
Further, the commenter hasn't provided any factual basis for its
assumption that all aircraft on the aircraft registry use ECi cylinder
assemblies.
We agree that we could better estimate the number of engines
affected by this AD. We again reviewed our estimate. We now estimate
that approximately 6,200 engines are affected by this AD. That number
is based on our initial estimate of approximately 43,000 affected
cylinder assemblies produced by ECi from 2002 to 2011. This number is
supported by AEC Technical Report 1102-13, dated April 30, 2011. We
then reduced 43,000 by our estimated number of cylinder assemblies that
would have been removed from service.
Our review indicates that approximately 6,000 of the 43,000
cylinder assemblies would have been retired from service by the time of
the publication of this AD. Therefore, we estimate 37,000 cylinder
assemblies may be in service, as of June 1, 2016. We divided this
number by 6 cylinders per engine to give us an estimated 6,167 engines
in service. To increase the conservatism of our cost estimate, we
rounded this figure to 6,200 engines. We revised our cost estimate to
reflect these updated calculations.

Request To Revise the Number of Labor Hours to Perform This AD

Comment. A few commenters, including IPL Group, indicated that the
number of hours to replace 6 cylinders would be greater than the 18
hours that we estimated in our costs of compliance.
Response. We agree. In the August 12, 2013, NPRM, and the January
8, 2015, and August 28, 2015, SNPRM, we
estimated 18 work hours. Although the commenters did not provide data
to support increasing the number of work hours, we held discussions
with manufacturers regarding the number of hours they would allow to
perform this work. Based on these more recent discussions, we revised
our estimate for the number of work hours to replace 6 cylinder
assemblies to 32 hours.

Request To Revise Cost of Replacing a Cylinder Assembly in This AD

Comment. Danbury Aerospace, Danbury Holdings, RAM Aircraft, and IPL
Group commented that the cost of a cylinder assembly, as calculated by
the FAA, does not accurately represent replacement costs. The
commenters indicated that the FAA's use of ``pro-rated cost'' allows a
vast underestimation of actual expenses that would be incurred by
owners. The agency must at least provide sound reasoning and facts
supporting the assertion that the pro-rated cost ``more accurately
reflects'' replacement cost. IPL Group further commented that a ``pro-
rated value'' is inconsistent with FAA policy and the Regulatory
Flexibility Act.
Response. We disagree in part. Industry, including ECi, uses pro-
rated cost in its cost estimates. For example, ECi, in its MSB 05-8,
Revision No. 1, dated December 29, 2005, used a similar time in service
based pro-rated cost calculation to determine the discounted cost to
operators for replacement cylinders, instead of providing the cylinders
to the operators at no cost. Further, we typically use pro-rated cost
for larger, turbofan engines when life-limited parts are involved.
Operators of those engines are typically airlines and other large
operators. Pro rata estimating therefore, is an acceptable method of
estimating cost.
We agree however, that engines with affected ECi cylinders
installed may be installed on airplanes owned by individual operators
in the general aviation community, who are less familiar with the
concept of pro-rated costs to ADs. In consequence, we revised our
estimate to use the full replacement cost of each cylinder assembly
even though this will likely result in an over-estimate of the total
cost of this AD. We, therefore, used the replacement cost of 6 cylinder
assemblies in this final rule. This resulted in an increase from $4,202
in the SNPRMs to $11,520 in this final rule.

Request To Include Additional Costs in the Overall Cost Estimate

Comment. IPL Group and Danbury Aerospace requested that we add
additional costs to our overall cost estimate. IPL Group indicated that
the FAA should include costs for loss of use of the aircraft, test
flight, and break-in expenses. Danbury Aerospace commented that we
should account for loss of overhauled assemblies as replacement items
and new costs associated strictly with their replacement.
Response. We disagree. In constructing our cost estimate, we
followed the guidance of the FAA's Airworthiness Directives Manual,
FAA-IR-M-8040.1C, dated May 17, 2010, which states ``Do not state any
costs beyond initial work-hours and parts costs. . . .'' The additional
costs cited by the commenters are not appropriate to our cost
estimates. We did not change this AD based on this comment.

Request To Withdraw the SNPRMs Because of Excessive Overall Cost

Comment. Several commenters commented that the FAA should withdraw
the January 8, 2015, SNPRM and the August 28, 2015, SNPRMs because the
FAA has underestimated the cost of compliance of this AD. These
commenters represented that the true cost is too high and that the FAA
has ignored the broader impact of this AD on industry. Most commenters
failed to provide any data to support these claims, however, IPL Group
provided some calculations to show that the total cost of this AD
should be somewhere between $168,666,625 and $320,360,156.
Response. We disagree. We considered the impact that this AD would
have on operators. As explained in response to the comments above, we
increased our estimates of inspection costs, labor costs, and
replacement costs of the cylinder assemblies. Although we increased our
cost estimate, we still conclude that the unsafe condition represented
by the affected cylinder assemblies requires an AD. We did not withdraw
the SNPRMs based on this comment.

Request To Substantiate Record-Keeping and Time Estimates

Comment. Danbury Holdings also stated that the FAA had not provided
documentation to substantiate its estimated record keeping cost and
time estimates.
Response. We agree in part. We interpret this comment as a
reference to both time spent on checking log books and reporting
requirements. We withdrew our reporting requirement when we published
the January 8, 2015, SNPRM, so we have no need to account for that
cost. We added an inspection cost in this final rule for the time
operators spend determining if they own an ECi cylinder assembly
affected by this AD. The Costs of Compliance section now states ``We
estimate 0.5 hours will be needed to check log books to determine if an
engine is affected by this AD.''

E. Administrative Comments

Request To Clarify Address

Comment. The Continental Motors Group commented that the business
at the address and telephone number listed in the August 28, 2015,
SNPRM (9503 Middlex Drive, San Antonio, Texas 78217, Phone 210-820-
8101) is now that of Continental Motors Inc., San Antonio. Continental
Motors Group also indicated that the associated company Web site
(http://www.eci.aero/pages/tech_svcpubs.aspx) listed in the August 28,
2015, SNPRM is not functional at this time.
Response. We agree. We updated the address and Web site information
listed in the ADDRESSES and ``Related Information'' sections of this
AD.

Request To Provide Names of Those Involved in the AD Process

Comment. Danbury Aerospace and Danbury Holdings commented that the
FAA should provide the names and technical positions of each of the
members of the multi-disciplinary/multi-directorate team that were
involved in the review of this service difficulty problem, along with
the dates, locations, and minutes for any meetings that were held.
Response. We disagree. The names and positions of personnel
associated with reviewing this AD are not necessary to the public's
participation in the development of this AD. We did not change this AD
based on this comment.

F. Support for the SNPRM

Comment. The NTSB commented that it believes that the August 28,
2015, SNPRM will satisfy the intent of NTSB Safety Recommendation A-12-
7. An individual commenter indicated that he had reviewed the SDR
database and determined that the separation rate of ECi cylinder
assemblies is approximately 10 times the rate of OEM cylinder
assemblies.
Response. We note the comment.

Conclusion

We reviewed the relevant data, considered the comments received,
and determined that air safety and the public interest require adopting
this AD as proposed.

Costs of Compliance

We estimate that this AD affects about 6,200 CMI model IO-520,
TSIO-520, IO-550, and IOF-550 reciprocating engines and all other CMI
engine models approved for the use of CMI models -520 and -550 cylinder
assemblies (such as the CMI model -470 when modified by STC), installed
on airplanes of U.S. registry. The average labor rate is $85 per hour.
We estimate 0.5 hours will be needed to check log books to determine if
an engine is affected by this AD. We estimate that about 32 hours will
be required to replace all six cylinder assemblies of an engine during
overhaul. We estimate the cost of replacement of six cylinder
assemblies to be, on average, about $11,520 per engine. Based on these
figures, we estimate the total cost of this AD to U.S. operators to
change all ECi cylinder assemblies to be $88,551,500.

Authority for This Rulemaking

Title 49 of the United States Code specifies the FAA's authority to
issue rules on aviation safety. Subtitle I, section 106, describes the
authority of the FAA Administrator. Subtitle VII: Aviation Programs,
describes in more detail the scope of the Agency's authority.
We are issuing this rulemaking under the authority described in
Subtitle VII, Part A, Subpart III, Section 44701: ``General
requirements.'' Under that section, Congress charges the FAA with
promoting safe flight of civil aircraft in air commerce by prescribing
regulations for practices, methods, and procedures the Administrator
finds necessary for safety in air commerce. This regulation is within
the scope of that authority because it addresses an unsafe condition
that is likely to exist or develop on products identified in this
rulemaking action.

Regulatory Flexibility Determination

The Regulatory Flexibility Act of 1980 (RFA) establishes ``as a
principle of regulatory issuance that agencies shall endeavor,
consistent with the objective of the rule and of applicable statutes,
to fit regulatory and informational requirements to the scale of the
business, organizations, and governmental jurisdictions subject to
regulation.'' To achieve that principle, the RFA requires agencies to
solicit and consider flexible regulatory proposals and to explain the
rationale for their actions. The RFA covers a wide range of small
entities, including small businesses, not-for-profit organizations, and
small governmental jurisdictions.
Agencies must perform a review to determine whether a proposed or
final rule will have a significant economic impact on a substantial
number of small entities. If the agency determines that it will, the
agency must prepare a regulatory flexibility analysis as described in
the Act. The FAA determined that this rule will have a significant
economic impact on a substantial number of small entities and,
accordingly, as required by Section 603(a) of the RFA, the FAA prepared
and published an initial regulatory flexibility analysis (IRFA) (79 FR
13924, March 12, 2014) as part of the NPRM (79 FR 48828, August 12,
2013) and initial SNPRM (80 FR 1008, January 8, 2015) for this rule.
For the second SNPRM, the FAA inadvertently stated that there would be
no significant impact on a substantial number of entities. We also
omitted the IRFA from the second SNPRM because we thought republication
unnecessary as costs had not changed and the IRFA had already been
published in the first SNPRM. In addition to the IRFA, Section 604 of
the RFA also requires an agency to publish a final regulatory
flexibility analysis (FRFA) in the Federal Register when issuing a
final rule.
With this FRFA we correct our misstatement in the second SNPRM and
restate our previous conclusions for the NPRM and in the first SNPRM
that the rule will have a significant impact on a substantial number of
small entities. Accordingly, in the following section we undertake the
regulatory flexibility analysis.

Final Regulatory Flexibility Analysis

Under Section 604(a) of the RFA, the Final analysis must address:
(1) Statement of the need for, and objectives of, the rule.
This final rule AD was prompted by failure reports of multiple
cylinder head-to-barrel separations and cracked and leaking aluminum
cylinder heads. This AD will apply to certain CMI San Antonio
replacement PMA cylinder assemblies marketed by ECi, used on the CMI
model -520 and -550 reciprocating engines, and all other engine models
approved for the use of CMI model -520 and -550 cylinder assemblies
such as the CMI model -470 when modified by STC.
(2) Statement of the significant issues raised by the public
comments in response to the initial regulatory flexibility analysis, a
statement of the assessment of the agency of such issues, and a
statement of any changes made in the proposed rule as a result of such
comments.
Danbury Holdings commented that the FAA had not provided the raw
data that was used in the IRFA. We note that the provision of raw data
is not required by the FAA's rulemaking procedures or orders.
In response to comments about problems with the repetitive
compression/soap test proposed by the NPRM, the FAA agrees that these
tests do not always reliably detect a cracked cylinder of this failure
mode and therefore the costs associated with such tests outweigh the
safety benefits. In the January 8, 2015 SNPRM the FAA removed the
requirement for repetitive compression/soap inspection tests.
The FAA received comments questioning the reduction of the
estimated number of smaller air service businesses (in addition to the
estimated 609 small part 135 operators) that would be affected by the
rule, from 5,000 in the IRFA to 2,000 in the January 8, 2015, SNPRM. We
note that in both cases the FAA stated that a substantial number of
small entities would be affected. Given the lack of available data, the
FAA is unable to make an accurate estimate of the number of smaller air
service businesses that will be affected by this rule, but acknowledges
that this number is substantial. In addition to the 609 small part 135
operators, we therefore estimate in this final rule that the number of
smaller air service businesses affected is substantial.
After publication of the NPRM and after publication of each of the
two SNPRMs, we also received comments from small businesses concerning
understated compliance costs. Some commenters stated that the labor
rate and the hours required to replace an affected engine's cylinders
are underestimated. We agree with this comment in part and have
increased our estimate of the labor hours required to replace an
affected engine's six cylinder assemblies from 18 to 32 hours, with a
corresponding labor cost increase from $1,530 to $2,720.
In response to comments we have also increased our cost of
materials estimate from a loss-of-service estimate of $4,202 to the
full cost to replace all six cylinders, which has increased to $11,520.
Our estimate of the total cost to replace all six cylinders has
therefore increased from $5,732 to $14,240.
After publication of the August 28, 2015, SNPRM, we received
negative comments concerning the inadvertent change from our original
determination of a significant economic impact on a
substantial number of small entities in the IRFA (and the January 8,
2015, SNPRM) to a determination of no significant impact on a
substantial number of small entities. As noted in the introductory
section, we are correcting this oversight in this FRFA.
(3) The response of the agency to any comments filed by the Chief
Counsel for Advocacy of the Small Business Administration in response
to the proposed rule, and a detailed statement of any change made to
the proposed rule in the final rule as a result of the comments.
The SBA did not submit comments.
(4) Description and an estimated number of small entities to which
the final rule will apply.
Of the 610 part 135 operators we found to be affected by this rule,
we identified 609 that meet the Small Business Administration (SBA)
definition of a small entity (entities with 1,500 or fewer employees)
that will be affected by this final rule. On this basis alone, we
conclude that the final rule will affect a substantial number of small
entities. In addition, we estimate that a substantial, but undetermined
number of smaller air services businesses will be affected by this
final rule. The FAA is unaware of the assets or financial resources of
these businesses. The affected part 135 and smaller air services fly
fixed wing aircraft; employ less than 1,500 employees; and conduct a
variety of air services such as fly passengers and cargo for hire.
(5) Description of the record keeping and other compliance
requirements of the final rule.

Record Keeping Requirement

The FAA estimates 0.5 hours will be needed to check log books to
determine if an engine is affected by this AD. At a wage rate of $85
per hour, the estimated cost will be $42.50 per engine. As the affected
small part 135 operators have between one and 88 affected airplanes,
the costs of this requirement range from $42.50 to $3740 per part 135
operator.

Compliance Requirement To Replace Cylinder Assemblies of Affected
Engines

This AD applies to certain CMI model IO-520, TSIO-520, IO-550, and
IOF-550 reciprocating engines and all other engine models approved for
the use of CMI models -520 and -550 cylinder assemblies (such as the
CMI model -470 when modified by STC), installed on airplanes of U.S.
registry. For the affected engines the AD requires replacement of the
cylinder assemblies at reduced times-in-service.
As noted above our estimate of the total cost to replace all six
cylinders has increased from $5,732 to $14,240. As the number of
airplanes held by affected small part 135 operators ranges from one to
88, the costs of required cylinder assembly replacement per operator
range from about $14.2 thousand to about $1.3 million.
To determine whether compliance costs will have a significant
economic impact, we measured the cost of replacing cylinder assemblies
of affected engines relative to the value of the affected airplanes
held by the small part 135 operators. The estimated asset value of the
affected airplanes held by the small part 135 operators ranges from
$22,000 to $19.6 million. We find that the cost of replacing cylinder
assemblies relative to affected airplane asset value is greater than 5
percent for 468 of the 609 affected small part 135 operators.\2\ We
therefore conclude that the final rule will have a significant economic
impact on a substantial number of small entities.
---------------------------------------------------------------------------

\2\ This assessment does not take into account record keeping
requirement costs. These costs, however, are minor and do not affect
our assessment of the number of small part 121 operators
significantly impacted by the final rule.
---------------------------------------------------------------------------

(6) Steps the agency has taken to minimize the significant economic
impact on small entities consistent with the stated objectives of
applicable statutes, including a statement of the factual, policy, and
legal reasons for selecting the alternative adopted in the final rule
and why each one of the other significant alternatives to the rule
considered by the agency which affect the impact on small entities was
rejected.
In response to comments about problems with repetitive compression/
soap test, the FAA agrees that these tests do not always reliably
detect a cracked cylinder of this failure mode and the costs associated
with such tests outweigh the safety benefit. The FAA removed that
requirement for repetitive compression/soap inspection tests. We also
considered these following alternatives:
(a) Do nothing--This option is not acceptable due to the number of
failures of ECi cylinder head assemblies and the consequences of the
failures.
(b) Periodic inspections only (no forced removals)--Though the NTSB
recommended this option in its comments to the NPRM (August 12, 2013,
78 FR 48828), the service history has shown that such inspections may
not reliably detect existing cracks and the rate of crack growth to
separation is unknown and variable. The NTSB also submitted a later
comment, in response to the August 28, 2015, SNPRM, that the revised
rule as adopted in this final rule, meets the intent of its Safety
Recommendations A-12-7.
(c) Forced removal with periodic inspections--Periodic inspections
may not reliably detect cracks and even with removal the rate of crack
growth to separation is unknown and variable. Forced removal is the
only remaining option.

Regulatory Findings

This AD will not have federalism implications under Executive Order
13132. This AD will not have a substantial direct effect on the States,
on the relationship between the national government and the States, or
on the distribution of power and responsibilities among the various
levels of government.
For the reasons discussed above, I certify that this AD:
(1) Is not a ``significant regulatory action'' under Executive
Order 12866,
(2) Is not a ``significant rule'' under DOT Regulatory Policies and
Procedures (44 FR 11034, February 26, 1979),
(3) Will not affect intrastate aviation in Alaska to the extent
that it justifies making a regulatory distinction, and
(4) Will have a significant economic impact, positive or negative,
on a substantial number of small entities under the criteria of the
Regulatory Flexibility Act.

List of Subjects in 14 CFR Part 39

Air transportation, Aircraft, Aviation safety, Incorporation by
reference, Safety.

Adoption of the Amendment

Accordingly, under the authority delegated to me by the
Administrator, the FAA amends 14 CFR part 39 as follows:

PART 39--AIRWORTHINESS DIRECTIVES

1. The authority citation for part 39 continues to read as follows:

Authority: 49 U.S.C. 106(g), 40113, 44701.

Sec. 39.13 [Amended]

2. The FAA amends Sec. 39.13 by adding the following new airworthiness
directive (AD):