Академический Документы
Профессиональный Документы
Культура Документы
Masters Thesis
By
ir. Anthony K. Muchiri
Supervisors:
Prof. ir Klaas Smit
ing. Willem van Steenis
Colophon
Final Report
Board Of Examiners:
Prof. Ir. K. Smit - TU Delft (Coach)
Ir. B. Agusdinata - TU Delft
ing. W. van Steenis - Transavia Airlines (Supervisor)
Ir. R. Hol - Transavia Airlines
Preface
The report has been subdivided into five chapters, and it can be read as follows:
- For readers interested in the development of a new maintenance planning
strategy, Chapters 3 and 4.4 are recommended
- For readers interested in the analysis of the current situation and the
proposed situation, it is recommended that they read Chapter 4.4
- For readers interested in the conclusions and recommendations, they should
look at Chapter 5
- For readers interested in the development of the whole research, all chapters
and Appendix documents are recommended. A summary highlighting all
major items from the report has also been included
I would also like to take this opportunity to thank my professor, Prof. Ir. K. Smit of
the TU Delft, and my supervisor, Mr. W. van Steenis, Maintenance Manager
Transavia Airlines, for their coaching and encouragement. My thanks also go to all
staffers of the MP&S department for their assistance and for the good time they gave
me at the department, the head of Engineering, Mr. R. Hol, all maintenance engineers
- and especially the maintenance programme engineers, and all members of the
technical department for their corporation.
Anthony Muchiri
iii
Executive Summary
Executive Summary
With the increasing need to reduce maintenance costs and increase aircraft
availability, the need to simplify the way maintenance is planned and executed has
become a major issue in the aircraft industry. Aircraft manufacturers continue to
develop aircraft with a low maintenance demand, while airlines strive to keep their
maintenance costs as low as possible.
The Boeing 737 Next Generation (737 NG) is an example of such an aircraft,
developed to demand less maintenance, as compared to previous versions of the
Boeing 737 series. This 737 NG aircraft has a Maintenance Planning Data (MPD)
document that is based on the Maintenance Steering Group (MSG)-3 philosophy.
This is a task-based maintenance philosophy that looks at maintenance more at a
task-level, as compared to previous philosophies, which were more focused on
maintenance processes. Transavia airlines currently possesses 13 737 NG aircraft, but
has a maintenance program that strongly borrows from other aircraft models in its
fleet (namely the B 757-200 and the B 737-300).
Clustering is the process of grouping maintenance tasks together into packages that
can be planned in for execution. Clustering can be done by following two
v
Maintenance Planning Optimisation: B737 NG
Line?
Base?
Line Maintenance Base Maintenance
Tasks Tasks
Transavia L/B
Classification Criteria
Set-up properties
Maintenance Intervals
(CT/FH/FC)
Maintenance Maintenance
Task Packages Checks
Figure 0. 1 Creation of Line Maintenance Task Packages and Base Maintenance Checks
The Transavia Line/Base classification criteria classifies all tasks, for which Transavia
is equipped to perform, has the certified personnel and materials needed, as line
maintenance tasks. All other tasks are base maintenance tasks.
The next step in the Bottom-Up approach is to determine when the Maintenance
Items (Maintenance Task Packages or Maintenance Checks) are due for performance.
This follows from combining the aircraft utilisation (flight hours and flight cycles)
and calendar with the maintenance interval limits of the various Maintenance Items.
vi
Executive Summary
The clustering process is done using a computer model, developed in Visual Basic
and MS Excel. The model combines maintenance item intervals with simulated
aircraft utilisation scenarios (high, average and low utilisation) and maintenance
scenarios (such as low maintenance frequencies). From these, the Maintenance
Demand (in number of visits and maintenance man-hours) is calculated. Further, the
model also calculates losses following from maintenance performed before the
interval limits are reached.
Line Maintenance clusters are evaluated for frequencies ranging between 4-6 weeks,
for a period of 6 years (starting from 2001, where an assumption of aircraft
introduction is made). Frequencies above 6-weeks would lead to the escalation of
maintenance interval limits. It is observed that line maintenance is strongly
utilisation dependent (in terms of flight hours and flight cycles). The results of the
evaluation are tabulated in Table 0. 1 below. It follows that the maintenance demand
decreases with a decreasing maintenance frequency, as does the de-escalation.
Table 0. 1 Maintenance man-hour demand per aircraft by a varying maintenance frequency
Base maintenance clusters are evaluated for an 18-month and a 24-month interval,
starting from April 2001, where an assumption of aircraft introduction is made.
Larger intervals would result in the escalation of maintenance interval limits. The
results of the evaluation are tabulated below.
It follows from Table 0. 2 that base man-hour demand decreases significantly when
the maintenance interval is raised to 24 months. The de-escalation does, however,
increase.
Further analysis reveals that by performing the first base maintenance visit a month
before it is due (initial de-escalation), more gains can be made in reducing the base
maintenance man-hour demand. The lowest man-hour demand follows from a 30-
day initial de-escalation. This is shown on Table 0. 3 below.
vii
Maintenance Planning Optimisation: B737 NG
Table 0. 3Base maintenance man-hour demand per aircraft: Before/After initial de-
escalation
In order to maximise the maintenance advantages that come along with the B737 NG,
the following is recommended:
viii
Samenvatting
Samenvatting
Onderhoudsplanning en inplanningsachtergrond
ix
Maintenance Planning Optimisation: B737 NG
Definite en ontwikkeling van Onderhoudsclusters
Line?
Base?
Line Maintenance Base Maintenance
Tasks Tasks
Transavia L/B
Classification Criteria
Set-up properties
Maintenance Intervals
(CT/FH/FC)
Maintenance Maintenance
Task Packages Checks
Figuur 0.1 Ontwikkelen van Onderhoudstaken-Paketten en Onderhoudsinspecties
x
Samenvatting
De samenstelling en evaluatie van onderhoudsclusters
De tabel hierboven laat zien dat de onderhoudsbehoefte vermindert met een lagere
frequentie, zo ook de de-escalatie van onderhoudslimieten.
xi
Maintenance Planning Optimisation: B737 NG
Tabel 0.2 toont aan dat de man-uren behoefte voor groot-onderhoud vermindert,
wanneer het onderhoudsinterval verhoogt wordt naar 24 maanden. Tegelijkertijd
neemt de de-escalatie toe. Verdere analyse wijst uit dat als de eerste groot-
onderhoudsbeurt uiterst een maand eerder wordt uitgevoerd (initile de-escalatie),
meer winst behaald kan worden in het reduceren van de man-uren behoefte voor
groot-onderhoud. De laagste behoefte aan man-uren volgt uit een initile de-escalatie
van 30 dagen. Zie ook Tabel 0.3 hieronder.
Tabel 0.3 Manuren behoefte per vliegtug: vr en na initile de-escalatie.
Conclusies en aanbevelingen
xii
Acronyms
Acronyms
xiii
Definition of Terms
Definition of terms
Maintenance Task Package: A group of maintenance tasks that share the same set-
up activities, and have a common maintenance interval
xv
Outline
Outline
Colophon .................................................................................................................................. i
Preface ..................................................................................................................................... iii
Executive Summary ............................................................................................................... v
Samenvatting.......................................................................................................................... ix
Acronyms.............................................................................................................................. xiii
Definition of terms................................................................................................................ xv
Outline.................................................................................................................................. xvii
1 Assignment Analysis ...................................................................................................... 1
1.1 Problem Background.............................................................................................. 1
1.2 Problem formulation .............................................................................................. 2
1.3 Objectives ................................................................................................................. 3
1.4 Research environment ........................................................................................... 6
1.5 Approach Strategy .................................................................................................. 6
1.5.1 Assignment analysis....................................................................................... 7
1.5.2 Background research ...................................................................................... 7
1.5.3 Maintenance program analysis: 737NG ...................................................... 8
1.5.4 Maintenance planning definition and development ................................. 9
1.5.5 Report phase .................................................................................................... 9
1.6 Requirements, constraints and assumptions ...................................................... 9
1.6.1 Requirements .................................................................................................. 9
1.6.2 Constraints ....................................................................................................... 9
1.6.3 Assumptions...................................................................................................10
2 Maintenance Planning and Scheduling ......................................................................11
2.1 Maintenance Program (MP) Development ........................................................11
2.1.1 MRIs based MPD items:................................................................................12
2.1.2 MRIs based on non-MPD items ...................................................................12
2.1.3 Maintenance Checks......................................................................................13
2.1.4 Maintenance Task Cards and task card control ........................................14
2.1.5 The B737-300 MP vs. the B737 NG MP .......................................................15
2.2 Maintenance Planning and scheduling ..............................................................17
2.2.1 Long term planning .......................................................................................18
2.2.2 Medium term planning .................................................................................18
2.2.3 Short term planning ......................................................................................18
2.3 Line Maintenance ..................................................................................................19
2.3.1 Scheduling of hangar visits ..........................................................................19
2.4 Base Maintenance ..................................................................................................22
2.4.1 Scheduling of Base Maintenance .................................................................22
2.5 The Planning department vs. the Sales department .........................................25
2.5.1 Commercial Planning....................................................................................25
2.6 Maintenance planning and the cost of ownership ............................................26
2.7 Conclusions: ...........................................................................................................27
3 Maintenance Clusters Definition and Development ................................................29
3.1 Approach ................................................................................................................29
3.2 Bottom-Up approach .............................................................................................30
3.2.1 Maintenance Checks......................................................................................30
3.2.2 Maintenance Task Packages .........................................................................30
xvii
Maintenance Planning Optimisation: B737 NG
3.2.3 Aircraft utilisation: Determining due dates and due items .....................32
3.2.4 Maintenance Scheduling ..............................................................................35
3.3 Top-down approach ..............................................................................................38
3.3.1 Aircraft utilisation requirement...................................................................38
3.3.2 Maintenance scheduling ...............................................................................41
3.4 Conclusions ............................................................................................................43
4 Cluster formation and evaluation ...............................................................................45
4.1 Approach ................................................................................................................45
4.2 Maintenance Item Clustering ...............................................................................46
4.2.1 Clustering method selection ........................................................................46
4.2.2 Calculation and simulation software ..........................................................47
4.3 Simulation process description ............................................................................48
4.3.1 Simulation Components: ..............................................................................48
4.3.2 Maintenance Item Allocation Model (MIAM): Model design .................51
4.3.3 Clustering: Line Maintenance Transavia Hangar ..................................53
4.3.4 Clustering: Base Maintenance MRO Station ...........................................53
4.3.5 Maintenance item Interval de-escalation ...................................................54
4.3.6 Model Validation and verification ..............................................................55
4.4 MIAM Results ........................................................................................................56
4.4.1 Base Maintenance ..........................................................................................56
4.4.2 Base Maintenance Optimisation ..................................................................62
4.4.3 Line Maintenance ..........................................................................................67
4.4.4 Line Maintenance Optimisation ..................................................................75
4.5 Summary .................................................................................................................80
5 Conclusions and Recommendations ...........................................................................81
5.1 Conclusions ............................................................................................................81
5.2 Recommendations .................................................................................................83
References ...............................................................................................................................85
A. Assignment .................................................................................................................87
B. Transavia Airlines......................................................................................................89
B.1. History.....................................................................................................................89
B.2. The fleet ...................................................................................................................89
B.3. Maintenance Facilities ...........................................................................................90
B.4. Organisation ...........................................................................................................90
B.5. Operational area ....................................................................................................93
C. The MSG-3 Philosophy .............................................................................................97
C.1. MSG Development: ...............................................................................................97
C.1.1. MSG-2 (Process oriented) .....................................................................97
C.1.2. RCM (Reliability Centred Maintenance) ............................................98
C.1.3. MSG-3 (Task oriented maintenance)...................................................98
C.2. Maintenance program objectives (MSG-3) .........................................................98
C.3. MSG-3 Logic flow analysis ...................................................................................99
D. Maintenance Program Packaging Methods .........................................................101
D.1. Maintenance program packaging methods .....................................................101
D.2. Block Packaging ...................................................................................................101
D.3. Equalised/Phased or progressive packaging: .................................................102
E. Line and Base Maintenance Redefinition .............................................................105
E.1. Cause .....................................................................................................................105
E.2. Aim ........................................................................................................................105
E.3. Approach ..............................................................................................................105
E.3.1. Engineering...........................................................................................106
xviii
Outline
xix
Introduction
1 Assignment Analysis
In order to sustain the inherent reliability1 of an aircraft, the aircraft manufacturer develops a
document referred to as Maintenance Planning Data (MPD). This is a generic document as it
contains all Maintenance Review Board (MRB) requirements, plus Boeing requirements that
result from service experience. It also includes aircraft information such as zonal diagrams,
ground rules, and recommended check intervals. The document provides maintenance
planning information necessary for an operator to develop a customised scheduled
maintenance program [Boeing CAS, 2000].
On top of tasks stipulated in the MPD, airlines also have additional tasks that follow from
Airline specific task cards. These tasks are meant to satisfy the operators specific needs for
aircraft utilisation. Such tasks may be scheduled in as routine maintenance. Other tasks
follow from incidents and accidents (Repair Instructions), Deferred Defects (non-routine
maintenance tasks), Maintenance Instructions and Aircraft Modifications (from Service
letters, Service bulletins and Airworthiness Directives). These are tasks that are performed
when opportunity arises, but may also have specified time limits.
Following the stipulated maintenance tasks and intervals from the MPD, an operator
develops a Maintenance Program (referred to as the Operators Maintenance Program
OMP- by Transavia). The maintenance program has to be approved by the local Aviation
Authority. This document reflects how an operator intends to schedule, package and
perform maintenance (maintenance checks). The operator arrives at an OMP by reviewing
the aircraft anticipated utilisation, fleet size, seasonal constraints, and maintenance
capabilities, operational requirements, dispatch reliability, amongst other factors, and
combining these with aircraft operational requirements.
The band of possible task packages2 for a given aircraft can range from:
1. A program consisting of a large number of progressively performed small packages,
resulting from Task-Based Maintenance (Equalised/Progressive check)
2. A program which bundles most scheduled tasks in relatively few large checks
performed at higher intervals (Block check).
Traditionally, the MPD documents recommended the use of Block checks (letter checks A, B,
C, D), by pre-determining the packages for each letter check. However, with the
advancements in aircraft maintenance analysis brought about by the task-oriented MSG-3
and its revisions (rev.1 en rev.2)3, the Boeing Company is tending to opt for non-packaging
of maintenance tasks in the MPD, leaving this open for the operator. This has been the case
with the Boeing 737 Next Generation (henceforth referred to as 737NG) MPD document. The
main reason for this is that the operator gets more freedom to plan maintenance based on
aircraft scheduling and utilisation, hence having more control over direct maintenance costs.
By reviewing an operators anticipated utilisation, environmental considerations, fleet size,
seasonal constraints, and other factors, it is possible to package an aircrafts maintenance
program taking full advantage of the allowable utilisation parameters as specified in the
MPD document.
1
Maintenance Planning Optimisation: B737 NG
The planning and execution of maintenance tasks can be boldly classified into two
categories:
a. Line maintenance: This can generally be defined as maintenance activities with a pre-
flight, daily, weekly, and periodic (A Check, B Check) frequency (JAR-145 Section 2,
paragraph 3.2.2). It mainly consists of visual inspections, lubrication, and general
servicing of aircraft and does not require the opening of panels and access doors for
detailed inspections.
Line maintenance is mainly characterised by short downtime periods of twenty-four
hours or less and small maintenance packages, quantified in terms of man-hours.
b. Heavy (Base) maintenance: This can generally be defined as maintenance activities falling
under C Checks and D checks (JAR-145 Section 2 paragraph 3.2.3), requiring panel and
access doors opening and disassembly for deeper inspection and eventually repair.
Base maintenance is mainly characterised by long downtime periods and large
maintenance packages, quantified in terms of man-hours
450.0
High S ea son
400.0
Low S e ason Low S e ason
Flight hours per aircraft
350.0
300.0
250.0
200.0
150.0
Ja n F eb M ar A pr May Jun Jul A ug S ep O ct Nov De c
2
Introduction
Maximum aircraft availability is desired during the summer (high) season (May - October)
when fleet utilisation is high. During the winter (low) season (November - April), the aircraft
utilisation demand is low. Maintenance tasks requiring significant downtime (mainly base
maintenance tasks) are, therefore, performed during the winter season. This often results in
the de-escalation of task-intervals for items performed much in advance (e.g. by performing
base maintenance at a frequency of 18 months, all maintenance tasks that will be due
between 19-35 months will have to be performed then, leading to an interval de-escalation of
between 19-35 months, depending on the task interval). There is, therefore, a need to
determine when, and how often, maintenance should be performed, in order to reduce task de-
escalation and consequently, maintenance costs.
In an effort to reduce maintenance costs arising from work contracted out and task-interval
de-escalation, line and base maintenance has been redefined within Transavia [Stam, 2001].
The new definitions for these terms are as follows:
a. 1. Line Maintenance: defined as all Pre-flight checks, Daily Checks, Service checks, and
all MRIs classified upon their demand for certified staff, required
tools/equipment/materials and facility demands, for which Transavia can provide,
i.e. Transavia meets these demands.
a. 2. Base Maintenance: defined as all MRIs classified upon their demand for certified
staff, required tools/equipment/materials and facility demands, for which Transavia
cannot provide, i.e. Transavia does not meet these demands.
(See also Appendix E: Line and Base Maintenance Redefinition)
These two definitions, which differ from those given in section 1.1, imply that each
individual task has to be analysed before being classified into either line or base
maintenance. They also imply that the line maintenance tasks (work performed in-house)
will significantly increase (coupled with a decrease in Base Maintenance tasks), hence
demanding a new approach in the planning and scheduling of maintenance in-house and
maintenance work contracted out. There is, therefore, the need to determine the impacts of the
increase in line maintenance tasks, and how this should be dealt with within the Transavia, with as
end-result the reduction of maintenance costs.
Finally, there is also the need to define exact work packages for both line and base maintenance in
order to facilitate easier financial budgeting. This financial budgeting is done once a year, and it
creates a capital reserve for base maintenance, as well as line maintenance. The base
maintenance reserve is based on the anticipated work to be issued out, and the current/
projected rates of base maintenance.
The definition of exact work packages would also help monitor and evaluate the impact of
line and base maintenance redefinition and hence facilitate further optimisation of the
maintenance program and maintenance planning.
1.3 Objectives
The objectives of the above-stated assignment can be achieved by answering the following
major questions, which will serve as the main topics:
3
Maintenance Planning Optimisation: B737 NG
A. What are the most significant factors determining maintenance task clustering and
planning optimisation within the Transavia organisation?
B. What is the most optimum way of clustering maintenance tasks to be performed by the
maintenance department at Transavia, and by a base maintenance station?
C. What impact do such clusters have on the cost of ownership4 of the aircraft (with regard
to maintenance)?
These questions have in turn been split into sub-questions for a deeper analysis of the
assignment. These sub-questions will also serve as the sub-topics for the research.
The assignment will be carried out in four phases (see the next chapter), in the course of
which the following sub-questions will be answered.
Ad A. The following sub-questions will be used to collect facts on the current situation
within the organisation, and to establish a base on which an eventual clustering will
be implemented.
A.3 How do the maintenance costs of the old and new fleet compare (-300, -800 Cost breakdown)?
As mentioned in section 1.2 above, the B737-300 MPD differs in many aspects from that
of the 737 NG. The aircraft operator (Transavia) expects a reduction in maintenance
costs through the maximisation of improved maintenance characteristics of the 737
NG. It is, therefore, necessary to establish whether this is currently the case.
A.4 What are the boundary conditions for operations and maintenance?
The research results are expected to be as realistic as possible. Therefore, various
factors that may limit the implementation of results should be established early in
advance in order to realise the objectives of the research.
Ad B. The following sub-questions will be used to guide the formation of maintenance task
packages.
B.1 What are the significant differences between the old and the new operators maintenance
program?
The maintenance planning is based on an approved maintenance program. The
maintenance program identifies all maintenance tasks that have to be performed, and
how often they have to be performed.
4
Introduction
This sub-question should lead to the determination of the key differences between the
traditional maintenance program structure and the new maintenance program
structure for the 737NG.
B.4 What are the lessons learnt from the past on scheduled and unscheduled maintenance?
This is to establish whether the maintenance execution department employs a certain
policy as pertains to performing unscheduled maintenance in combination with
routine maintenance. This sub-question should also determine how Deferred Defects
are dealt with within the organisation.
B.6 What conclusions can be drawn from the answers to the questions above on fields that need to
be optimised?
Information gathered through sub-questions B.1 B.5 should form a base on which
maintenance clusters should be built on to satisfy the needs of Transavias technical
department
Ad C.After the formation of clusters, an evaluation will have to be done in order to establish
their effectiveness within the organisation. Sub-questions to be used in the evaluation
are:
C.1 How do the new task packages and tasks compare/differ from the previous task packages?
It will be necessary to establish the main similarities and differences between the
current and the new maintenance packages. This will be the first step in the evaluation
of maintenance packages resulting created from section B.7 above.
C.2 Is it possible to determine what impact the new work schedule will have on the maintenance
department?
The second step in the evaluation and optimisation of maintenance clusters is
establishing what impact such clusters have on the planning and execution of
maintenance, both on the work floor and at the planning department
5
Maintenance Planning Optimisation: B737 NG
The research is to be carried out at the Maintenance Planning and Support (MP&S)
department of the Transavia airlines (Appendix B).
6
Introduction
The following diagram (Figure 1.2) illustrates the relationship between elements of the
phases listed above and the documents resulting there from:
R ese arch P h ase C on te n ts R e su ltin g D ocu m e n ts
C h ap te r 1:
A ssig n m e n t an alysis P H A SE 1 A ssig n m e n t A n alysis
M ain te n an ce P lan n in g C h ap te r 3:
d e fin ition an d d e v e lop m en t M ain te n an ce C lu ste rs d e fin ition an d d e ve lop m e n t
P H A SE 3 C h ap te r 4:
M ain te n an ce C lu ste r form ation , e valu ation an d
M ain te n an ce P lan n in g
op tim ization
O p tim isation
C h ap te r 5:
R e port ph ase P H A SE 4 C on clu sion s an d re com m e n d ation s:
O ptim isation proposal
Assignment Analysis
Phase 1
Background
Literature Study
Research
Maintenance
Program and Planning
Analysis Phase 2
Maintenance
Optimization
Planning definition
Proposal
And Development
Phase 4
Maintenance
Planning Evaluation
And Optimization
Phase 3
This phase will be concluded with an Assignment analysis report (this document).
7
Maintenance Planning Optimisation: B737 NG
This first part of phase 2 will assess the current situation within the Engineering and
maintenance department of Transavia, as pertains to the maintenance of its B737 fleet.
Information gathered from this phase will be used to form a technical and economic base for
the research. The following is a description of the approach for this phase:
c. Literature Study
Literature references may provide an insight on various elements affecting
maintenance planning and maintenance processes, and also proposals for solutions
on similar problems.
During this second part of phase 2, a deeper look will be taken into the MPD and the
Operators Maintenance Manual (OMP) of the 737NG. The following is a description of the
contents of this phase
It may also be important to establish what elements of the Next Generation MPD make the
maintenance planning and execution on the aircraft better than that of previous models, as
stated by the aircraft manufacturer (Boeing).
Owing to the fact that the MPD also proposes a maintenance-planning schedule, it is worth
analysing this schedule in order to establish whether it is also applicable for the Transavia
fleet. Ideas on an eventual maintenance-planning schedule may also be borrowed from this
proposal.
8
Introduction
Information gathered from the previous phases will be put together in this phase, and this
should result in new maintenance clusters. The intervals in which different clusters fall will
be pre-determined before clustering is done. These clusters will then be evaluated with the
help of a maintenance allocation model (see Chapter 4), and optimised where necessary.
This will be done as follows:
a. Listing all maintenance tasks with all relevant features established in section 1.5.3a
above
b. Developing a clustering strategy and forming clusters based on facts gathered from
the previous stages of the research
c. Evaluating the effectiveness of maintenance clusters created in (b.) above, and if
necessary, optimising these. This should then result in the identification of the most
effective way to cluster task packages and perform maintenance.
The above strategy is preferred, owing to the fact that it has a logical sequence as pertains to
achieving the desired results of the research.
Phase 3 will be concluded with a Maintenance planning definition and development report.
1.6.1 Requirements
- The research should be completed within the stipulated period of nine months
- The results of the research should be realistic and should fit within the framework of
Transavia airlines. A look should also be taken on how the results fit into the KLM
737 Business Unit
1.6.2 Constraints
a. Transavia flight operations are seasonal in nature
Implications:
Large variations in the number of flight hours and cycles between summer and
winter
Maximum availability is required for the summer season, thus maintenance activities
with long downtime periods should not fall in this season
b. The airline frequently leases aircraft in and out
Implications:
Variations in fleet capacity, as requested by the flight operations department, and
consequently the (non-) availability of aircraft.
Introduction of different maintenance programs for aircraft leased in under foreign
registration.
c. Aviation authorities regulate maintenance programs
Implications:
9
Maintenance Planning Optimisation: B737 NG
1.6.3 Assumptions
The term B737 NG is used to define both the B737-800 and the B737-700. Presently, the
800 version totals 12 aircraft and the 700 totals 1 aircraft.
Unless otherwise stated, maintenance programs, maintenance strategies and
maintenance tasks referred to in the report are 737 NG specific
10
Chapter 2
This chapter will look into the background of maintenance planning and scheduling
within Transavia.
The process of establishing a maintenance planning and scheduling at Transavia
follows the sequence illustrated below:
Maintenance
Maintenance Maintenance
Program (MP)
Planning Scheduling
Development
(2) (3)
(1)
The initial source for the development of the OMP is the manufacturers
Maintenance Planning Data (MPD) document. The MPD covers all manufacturer
recommended tasks, the MSIs (developed following the MSG-3 philosophy). Within
Transavia, the MP is documented in the Operators Maintenance Program (OMP).
This document is an approved maintenance program (see Figure 2.2) meant to guide
maintenance in a JAR-145 maintenance organisation (See also Appendix G.2).
The Maintenance Programs section within the Engineering Department takes the
responsibility for the development and amendment of the Transavia Maintenance
Program (as illustrated in Figure 2.2 below). Each Maintenance Significant Item (MSI)
or Structural Significant Item (SSI) from the MPD is translated into a Maintenance
Requirement Item (MRI). The MRIs are limited by calendar days and/or flight hours
and/or flight cycles.
Principally, there is no difference between an MSI and a MRI. However, MRIs also
encompass SSIs, and they do not include items listed in the MPD, not applicable to
the Transavia fleet. Section 2.1.3 will go deeper into MSI/MRI features.
All MRIs are evaluated and classified as base maintenance MRIs or line maintenance
MRIs. The evaluation process has already been mentioned in section 1.2, and is fully
described in Appendix E.3.
11
Maintenance Planning Optimisation: B737 NG
12
Chapter 2
MSG-3 rev.2
Airline/Manufacturer Manufacturer B737 NG
Development Airlines Maintenance
Document Requirement
(ATA) Review Documents
Service
Letters EOs/MIs
(SLs)
Boeing MRIs Tasks,Task-
Service Intervals
Bulletins EOs/MIs
(SBs) Transavias B737 NG
Maintenance Transavia
Airworthiness HV MRIs Program Job Cards
Directives EOs/MIs Tasks,Task- (OMP) 1.1
(ADs) Intervals
Transavias
Maintenance National Aviation
Instructions
Transavia Task Authority (IVW/DL) Source: Boeing 737-NG MPD (derived)
(MIs)
Cards Approval
Property Description
MRI number A MPD Item Number: see also Figure 2.3)
Zone Section of the plane where task is performed
Access: Access panel or door required to be opened when performing the task
Interval Expressed in terms of frequency and usage parameters, such as flight hours,
flight cycles and calendar time (in days)
13
Maintenance Planning Optimisation: B737 NG
HV interval A conversion of the MSI interval into an acceptable Transavia interval, e.g., a
MSI interval of 2000CYC (cycles) becomes 38C2000L or 38C2000B, depending
on whether the task is a line or a base maintenance task. Other codes are
38D_ (Day), 38D_* (Day/Cycle or Day/Hour), 38H_ (Hour) and 38C_*
(Cycle/Day)
Applicability Applicability of the task to aircraft model and engine model specified
Zonal title/ task This is a general description of the task to be performed. It also provides
description additional explanation on access and interval for the specified task
HV Task Classification Transavias classification into either Line (L) or Base (B) Maintenance
Considering that there are over 1500 maintenance tasks, it would not be reasonable
to deal with each task separately. The B737 NG OMP deals with this problem by
grouping maintenance tasks into checks, solely on their HV intervals and Transavia
classification properties. This strategy follows from the OMPs of other aircraft types
in the Transavia fleet, namely the B737-300 and the B757-200. A complete overview
of these checks is included in Appendix I.
It should be observed at this point that the Transavia maintenance checks do not
portray any properties of MSG-3 philosophy. Features such as the MSG-3 category
and the MSG-3 classification play little to no role in the checks.
For all MPD tasks, the aircraft manufacturer (Boeing) prepares task cards (see Figure
2.2 above). Tasks cards contain a thorough description of the task, a zonal diagram
illustrating the position or component to be maintained, and an approximation of the
amount of Man-hours and the elapsed time (time needed to complete the task for a
given amount of man-hours) needed to complete the task. These task cards are
revised at a constant basis, either by Transavia, or by Boeing, in order to incorporate
improvements and corrections. Transavia also prepares task cards for all its
operational-specific maintenance tasks.
14
Chapter 2
MRI Number
27 - 010 01 - 01
Task Card Sequence Number
ATA Chapter
Before the introduction of the 737NG aircraft (the subject of this research), Transavia
was, and still is, operating the B 737-300 (classics).
The B733 (classic) maintenance program differs from that of the next generation on a
number of aspects. These aspects are tabulated below:
15
Maintenance Planning Optimisation: B737 NG
Maintenance Tasks Packaged into OMP Checks as: Packaged into OMP Checks as:
Transit Check Pre-flight Check
24-hr Check Ground Handling Check
(Boeing) A-Check at 250 Flight hours Tire & Oil Check
(Boeing) C-Check or Annual Visits AV Service Check
4000 Flight Hours 1AV Service Check
8400 Flight Hours 2AV (Transavia) Line Maintenance Checks
15600 Flight Hours 4AV (Transavia) Base Maintenance Checks
30000 Flight Hours 8AV
4. The Maintenance task list indicated under the NG column implies that only
the task execution frequency is stated, but not a package in which the task
should be performed (as witnessed in the previous column)
It should also be observed that Transavia has been operating the B737 classics for
over 15 years. The current B737-classics fleet average age is 10.14 years. From the
maintenance point of view, a lot of experience is gathered over the years by
maintaining the same type of aircraft. Routine and non-routine maintenance of such
a fleet is easily quantified and can be predicted to a certain extent.
However, the case above is not applicable for the next generation fleet. These aircraft
have been designed for easier and more flexible maintenance. The average fleet age
of the B737 NG is 1.5 years (see also Table B.1), and not enough experience has been
gathered to help establish the exact maintenance characteristics of this type of
aircraft. This implies that continuous adjustments on the maintenance program and
the maintenance planning are necessary, as compared to the minimal adjustments
(based on AD notes and SBs) on the maintenance of the classic fleet.
The maintenance program of the B737 NG is, therefore, under constant revision,
partly because it has to be revised every three months (following quarterly MPD
16
Chapter 2
reviews), and partly because new procedures have to be added to it. The
maintenance program section of the engineering is responsible for this.
1. Routine maintenance:
This is performed in accordance with the instructions stated in the OMP. Such
instructions and consequently utilisation limits serve as a basis for the
planning of aircraft maintenance.
2. Non-routine maintenance
In cases of component failure, rectification may be performed immediately or
may be deferred, depending on the severity. Operational critical items and
safety related items (listed in Minimum Equipment List MEL) would
require immediate corrective action, while non-safety and non-critical items
will be put on a deferred defect sheet (DDS) for rectification when
opportunity arises.
Maintenance planning is done in three phases, namely Long-term, Medium term and
Short-term.
Maintenance
Planning
17
Maintenance Planning Optimisation: B737 NG
Task packages
The Transavia OMP groups maintenance tasks into inspections/checks, based on
their maintenance intervals. A complete list of all packages is listed in Appendix I.1
and I.2. Maintenance planning then packages these tasks in preparation for a hangar
18
Chapter 2
or base maintenance visit. The packages are segmented due to the inequality of
maintenance windows.
Tyre and Oil Check: Analytical check performed at the end of each day to visual
certify operational condition and for trend monitoring purposes
(on-condition maintenance)
Service Check: Visual and functional check performed every two days to
ensure that all operational and safety standards are met
Ground Handling Checks to ensure that aircraft is configured for gate arrival and
Checks: gate departure
Hangar visit intervals are determined by the maintenance package with the lowest
interval (either in days, hours or cycles). This interval is currently 5 weeks for the
aircraft in question. Routine maintenance performed during such a visit includes all
checks due at the moment of the visit, and all checks due between that moment and
the next visit.
Hangar planning5:
i. Hangar maintenance is carried out on weekdays, between Monday evening and
Thursday night;
D = DAY 07:00 till 15:00
D A N D A N D A N D A N D A N D A N D A N
5Based on a fixed agreement between MP&S, Aircraft Maintenance (V.O.), and Hangar space
managers. See also Appendix H.
19
Maintenance Planning Optimisation: B737 NG
ii. Each day (24 hours) is sub-divided into three eight (8)-hour segments, also referred
to as shifts: a day shift, evening shift and night shift;
iii. Each week has a total of eleven (11) shifts available for hangar maintenance;
iv. Each eight-hour shift has six effective maintenance hours. The extra two hours are
allocated for rest for the maintenance crew.
v. Each shift comprises of an average of 12 technicians (Team), headed by 2 Team Co-
ordinators (TC). Each team comprises of mechanical as well as avionics technicians.
However, their proportion may vary from team to team. The proportions may also
be altered in order to provide enough skills for a given work package;
vi. Each shift can therefore produce an average of 72 man-hours;
vii. Hangar planning rules are such that only two aircraft may be maintained
simultaneously. The distribution of manpower in the hangar, the sequence with
which maintenance tasks are carried out, and eventual swaps of technicians
(platform/hangar) is done on the work floor, and is controlled by Maintenance
Controllers (MC);
viii. Planning for hangar maintenance is based on, respectively, manpower availability,
hangar space availability and aircraft availability;
The available hangar shifts can be used distributed over any of the three types of
aircraft operated by Transavia, depending on the demand for maintenance. This
report will, however, concentrate on the maintenance demand of the B737 NG
aircraft.
Hangar slots do not necessarily have to run concurrently with hangar shifts, even
though this would be more preferable. The following is an example of why this is the
case:
Table 2.4 Aircraft Flight Schedule: Singe Mx slot (Example: HZE, Week 27, 2000)
(*: Points vi, vii and viii above answer questions that may arise from hangar capacity
and manpower utilisation)
As mentioned earlier in this section, this aircraft is slotted in for a hangar visit once
every five weeks. The exact demand for maintenance (following from maintenance
due items at time of maintenance) determines the total amount of man hours
required and the total amount of downtime for a maintenance visit.
Table 2.5 below tabulates the line maintenance man-hours, and the frequency with
which hangar slots are allocated to each B737 NG aircraft annually. The man-hours
allocated are based on the assumption that all routine (75%) and non-routine (25%)
(Normal wear and tear) maintenance is fully provided for (see also Table 2.6 on the
realisation).
Week Days Slots Max. Available Man-hours
1 7 - -
5 35 1 < 55
10 70 1 < 55
15 105 2 < 100
20 140 1 < 55
20
Chapter 2
25 175 1 < 55
30 210 4 < 300
35 245 1 < 55
40 480 1 < 55
45 315 2 < 100
50 350 1 < 55
52 364 - -
Totals (364) 14 < 885
The realisation of the planning above, however, portrays a totally different picture,
as tabulated in Table 2.6 below. This table shows the actual number of slots allocated
to three 737NG aircraft (registrations PH-HZA, PH-HZB and PH-HZF) over the year
2000. These aircraft have been selected because they operated for Transavia
throughout the whole year (not leased out or introduced into the fleet in 2000).
Calendar week HZA HZB HZF Calendar Week HZA HZB HZF
1 2 27 1 1
2 28 2
3 1 1 29 1(BMx)
4 2 30
5 31 1
6 2 2 32 2 1
7 33 2 1
8 1 34
9 35
10 1 36 1 1
11 2 37 4
12 1 1 38 1
13 39
14 40 1
15 1 2 41 1
16 2 42 2
17 43
18 44
19 3 3 45 1
20 1 46 1 4 2
21 47 BMx
22 48
23 1 49 2 1
24 1 1 50
25 51 1
26 1(BMx) 52
Slots Routine* 24 24 19
Grey region = High Season Slots non-routine** 16 4 15
BMx = Base Mainte Total Slots 40 28 34
nance
* = Including non-routine from normal wear and tear
** = See also Table 2.7
Table 2.6 Realisation Hangar maintenance planning (Source: Maintenance Schedule, 2000 MP&S)
The main reason for the differences between the planning and the realisation result
from:
21
Maintenance Planning Optimisation: B737 NG
Table 2.7 below quantifies how the non-routine maintenance activities above are
distributed per aircraft tail number:
Maintenance checks classified as Base Maintenance tasks are always contracted out
(see section 2.1 and Appendix E).
In order to provide a good base for Base Maintenance (technical and financial)
planning, Transavia airlines enters into long-term contracts with base maintenance
organisations. Currently, KLM Maintenance and Engineering performs all base
maintenance for the Transavia 737 NG fleet.
22
Chapter 2
Unlike the 737-Classics, the Next Generation aircraft do not have a well-defined base
maintenance. Base maintenance checks are therefore packaged based on the number
of due base maintenance checks. Further, the base maintenance interval of 18 months
is not based on a recommended interval, but on a maintenance tradition adapted
from B757-200 maintenance pattern. This aircraft also forms a part of the Transavia
fleet.
At the start of this research, 6 aircraft had had their first base maintenance visit. The
following figure illustrates the composition of the various visits.
1800
1600
1400
1200
KLM Manhours
1000
800
Additional work
600 Non-routine
Routine
400
200
0
HZA (Nov-99) HZB (Jan-00) HZC (Feb-00) HZD (Nov-00) HZE (Nov-00) HZF (Nov-00)
Base maintenance visit per tail number
At the beginning of this research, the three oldest aircraft in the fleet had had their
second base maintenance visit. The following is an overview of the total man-hours
spent on these aircraft:
23
Maintenance Planning Optimisation: B737 NG
3000
2500
2000
KLM Manhours
1500
1000
Additional work
Non-routine
Routine
500
0
HZA (Mar-01) HZB (Mar-01) HZC (Mar-01)
Base maintenance visit per tail number
Figure 2.7: Second Base Maintenance Visit: Overview (Source: HV Project Management)
In order to give an illustration of what the routine, Non-routine and additional work
entails, the base maintenance packages for the HZA (first and second) are
summarised in table x below.
First Base Mx: HZA [Nov 1999] KLM Second Base Mx [Mar 2001] KLM
Routine Maintenance: Routine Maintenance:
Checks (Boeing MHrs: 134.21) 268.7 Checks (Boeing MHrs: 153.31) 312.8
Non Routines: Non Routines:
Normal wear and tear 516 Normal wear and tear 206.2
Shop work 69 Shop work 5.6
Table 2.8 Overview of Base maintenance visit components (Source: HV Project Management)
Note: Docking, additional KLM tasks and Open and Close fall under routine
maintenance.
From Figure 2.6, Figure 2.7, and from Table 2.2 above, the following is observed:
1. The base maintenance station (KLM) uses a factor 2 to calculate man-hours
needed for routine maintenance, without considering eventual non-routines.
Including eventual non-routines, a factor 3.6 is applied on the original Boeing
man-hours.
2. The routine maintenance work from both the first and the second base
maintenance visit is approximately equal.
24
Chapter 2
The following diagram (Figure 2.8) illustrates how these two departments interrelate
Aircraft
factors
Aircraft for
Objective: Optimum utilisation of aircraft Availability
production Safety
factors
The CPV department (which falls under the Sales and Marketing) is charged with the
sale of aircraft seats, hence directly determining the demand for aircraft capacity. In
order to ease the task of maintenance forecasting, CPV prepares an aircraft demand
forecast based on information received from contractual sales and fleet capacity
planning departments.
25
Maintenance Planning Optimisation: B737 NG
- The CPV long-term planning is seasonal and follows the utilisation pattern
illustrated in section 1.2, while short term planning ranges from between a month
to a day before the actual execution of a planned flight.
- Short term planning from the CPV is basically sales driven, either in advance or
ad hoc, but it can also be influenced by aircraft availability, i.e. aircraft
serviceability. Aircraft maintenance only has an impact on availability and the
technical dispatch reliability.
Principally, MP&S prepares a long term planning for both line and base
maintenance. Owing to the fact that the Transavia fleet is fully interchangeable, with
variations only capacity-wise, CPV is able to plan in serviceable aircraft for a whole
season in advance, and allocate maintenance slots to each aircraft, by combining
aircraft demand for operation and maintenance demand (from Table 2.5).
The relationship between the maintenance planning and scheduling, and the cost of
ownership (Figure 2.9 below) may, therefore, be visualised as follows:
Fuel/Oil
Depreciation/Rental
The amount of labour (man-hours) and materials spend on (routine and non-routine)
maintenance directly determines the direct maintenance costs.
26
Chapter 2
As pertains to costs arising from aircraft downtime, Figure 2.10 below can be used to
elaborate this.
In Serviceable aircraft
Waiting for
Non-routine
repair
4 maintenance Unplanned
Unavailability 5
resources (mainly to
(lost production) Unavailability
Breakdown: un-serviceable aircraft
Down-time
Maintenance resources Maintenance
In routine (Manpower, spares, Tools) plan
maintenance
3 Feedback
Unavailability for control
(lost production) Planned resources
(mainly routine)
Out of
In operation
1 2 operation Source (derived): Kelly, 1984
Ground-time
Up-time
In block 3 above, the aircraft is withdrawn from operation in order to provide for
maintenance downtime. This un-availability may be considered as a loss in terms of
possible seats that could have been available for sale if the aircraft was in operation.
In blocks 4 and 5, the aircraft is un-available due to technical reasons, and this leads
to inevitable losses in terms of production (seats, flight hours). Further, additional
maintenance resources have to be provided for in order to make the aircraft
serviceable again.
On the other hand, if maintenance was to be performed in block 2, there will be no
losses experienced in terms of aircraft un-availability, owing to the fact that the
aircraft is not needed in that block. No maintenance is performed in block 1.
Indirect maintenance costs can vary greatly because these mainly depend on the
maintenance organisation. On the other hand, direct maintenance costs of an aircraft
can be viewed as more systematic and controllable costs.
2.7 Conclusions:
The introduction of the B737 NG in the Transavia fleet has created the need to review
the way in which the organisation plans its maintenance. The exclusion of the letter
checks from the MPD document was done in order to give airlines more certainty in
the planning and execution of maintenance. However, Transavia still tends to utilise
the letter check system to plan maintenance.
27
Maintenance Planning Optimisation: B737 NG
There is, therefore, a need to radically change the mentality and the traditions of
maintenance program development and maintenance planning and scheduling in
order to utilise fully the advantages of Next Generation aircraft, which should lead to
a reduction in the amount of maintenance performed on the fleet
In short:
1. There is a need to determine the composition and the contents of maintenance
packages for both line and base maintenance (Main question 1.3.B)
2. There is a need to determine optimum intervals for both line and base
maintenance scheduling (Main question 1.3.B)
3. There is a need to quantify the gains made from (1) and (2) above, in terms of
cost reduction and improved maintenance effectiveness (Main question
1.3.C)
28
Chapter 3
This chapter will establish a strategy through which maintenance tasks should be
grouped together and scheduled for performance. The grouping up of maintenance
activities for execution at a specified date will be referred to as clustering, while the
resulting groups of maintenance tasks will be referred to as clusters.
The first section 3.1 will discuss clustering strategies, followed by an application of
each strategy to the Transavia situation in sections 3.2 and 3.3. This will then be
followed by conclusions in section 3.4.
3.1 Approach
In order to cluster tasks together and schedule then for execution, there is a need to
identify which tasks (1) can be grouped together and how (2) the tasks have to be
grouped. Thereafter, it is necessary to establish when (3) these clusters will be
performed.
These three questions can be answered using two approaches, as suggested by File
(2000). These approaches are:
Maintenance Scheduling
Maintenance Tasks
Maintenance Schedule
29
Maintenance Planning Optimisation: B737 NG
As was mentioned earlier, the grouping together of maintenance tasks into checks
only looks at the HV (Transavia) Interval, and the line/base maintenance
classification. This is also the case for the rest of the aircraft in the Transavia fleet.
Given that the 737NG MPD is a MSG-3 based document, it can be said that there is
very little application of the MSG-3 Task-Based Maintenance philosophy (Appendix
C.1.3) in the MPD. The result of this is that there are very few differences in the way
maintenance is planned and executed for the entire fleet. These differences are the A-
Checks for the B737-300 (see also Table 2.2), and the Line Maintenance phases for the
B757.
Consequently, very few gains can be made in terms of hangar maintenance for the
Next Generation fleet.
Apart from grouping tasks together by only using the task interval property, other
properties can be considered (see also Table 2.1).
Dijkhuizen [1998] suggests the use of Set-Up properties for the packaging of
maintenance tasks, especially those performed frequently (Line Maintenance). He
defines Set-Up as: where one or more tasks require a fixed condition/procedure/cost,
irrespective of how many components are being maintained.
By considering tasks that share the same set-up activities (e.g. the same access area or
access panels, aircraft preparation), tasks can be clustered together. But by
30
Chapter 3
considering that such tasks may bear different maintenance intervals, the
maintenance interval should also be considered.
Line?
Base?
Line Maintenance Base Maintenance
Tasks Tasks
Transavia L/B
Classification Criteria
Set-up properties
Maintenance Intervals
(CT/FH/FC)
Maintenance Maintenance
Task Packages Checks
The following is an example of the results of the process illustrated on Figure 3.1
above.
Task MPD Interval* HV Check Boeing MSG-3 MSG-3 Skill ATA 100 (MRIs)
Package Notation Man-hours Category Class. Chapter
C4000A* 4000CYC or 540 DAYS 38C4000* 1.85 5DT, 2GV 5, 6 M 20. 32 7
C4000B* 4000CYC or 540 DAYS 38C4000* 1.08 GV 0 M 52 10
C4000C* 4000CYC or 540 DAYS 38C4000* 2.49 GV 0 M 53 12
C4000D* 4000CYC or 540 DAYS 38C4000* 4.16 GV 0 M 53 9
C4000E* 4000CYC or 540 DAYS 38C4000* 0.59 GV 0 M 53 3
C4000F* 4000CYC or 540 DAYS 38C4000* 0.54 GV 0 M 53, 54 4
C4000G* 4000CYC or 540 DAYS 38C4000* 3.96 GV 0 M 55 23
C4000H* 4000CYC or 540 DAYS 38C4000* 11.21 GV 0 M 57 54
8 packages 25.88 122
Table 3.2 Sample Maintenance Task Packages resulting from the Figure 3.1 process
Notes:
31
Maintenance Planning Optimisation: B737 NG
*: Which Ever Comes First
DT: Detailed inspection
GV: General Visual
M: Mechanical Skills
MSG-3 Class. /Cat: (See Table 2.1)
A complete list of the Maintenance Task Packages that follow from the classification
tree above is included in Appendix I.3. This list consists of a total of 121 Maintenance
Task Packages, as compared to the previous 51 Line Maintenance checks.
It is evident from the example above that from a large Transavia Check (122
maintenance tasks, 26 Boeing man hours - Table 3.1), 8 manageable Task Packages
can be created (Table 3.2). A deeper analysis into the exact nature of the task
packages reveals that the C4000G* and C4000H* contain tasks requiring significant
ground time, as compared to the other 6. Previously, the C4000* check was
considered as a time consuming check, requiring a lot of man-hours (i.e. 93 man-
hours base maintenance or 51.76 man-hours line maintenance).
From the example above, it is also clear that other Maintenance Task Package
properties are revealed, than was previously the case. These included:
1. The ATA chapters of the tasks being performed
2. The skills required per task package (M)
3. The Type of check (General Visual, Detailed)
The three properties listed above are of importance for the MP&S department, but
their relevance will not be dealt with in this report.
In the following sections, Maintenance Task Packages and Maintenance Checks will
be referred to as maintenance items. This will be in cases where both line and base
maintenance are dealt with together.
Aircraft utilisation can be determined using data gathered from two sources, namely:
1. METALS Utilisation: Data used to predict maintenance due dates and to
monitor component ages in the METALS system. The Data is stored in the
METALS system
2. Actual B737 NG utilisation Records: Data gathered by the records department
for Reliability analysis, and for updating of the data in METALS. Such Data is
issued through Transavias Boeing 737-700/800 Monthly and Quarterly Fleet
Reports
32
Chapter 3
METALS:
METALS utilises average fleet utilisation data from previous years in order to
generate forecasts for future aircraft utilisation. Instead of dividing up a year into
logical segments, e.g. months or quarters, METALS works with unique segments
(see Table 3.3). These segments are derived from trends monitored (by METALS)
over the years of fleet data collection. The utilisation that follows is, therefore, not
specific for a given type of aircraft, for it applies to the whole fleet.
Table 3.3 Annual Aircraft utilisation METALS (as per 1 April 2001)
These values have been plotted on Figure 3.2 below.
The shortcomings of the utilisation pattern above are:
- A poor definition of the high and the low season. While section 1.2 clearly defines
the high season as the period between April 1 and October 31, the METALS
utilisation misses out on the end of the high and the beginning of the low season.
- The utilisation above works with the total Transavia fleet average. In practice, the
utilisation of different types of aircraft does differ.
- The assumption that aircraft utilisation for all months in a segment is the same
16
14
12
Flight Hours Per Day
10
8
Hours
Cycles
6
0
01/01-31/03 01/04-03/06 04/06-23/09 24/09-02-12 03-12-31/12
Period
Figure 3.2: Aircraft utilisation pattern METALS (as per 1 April 2001)
33
Maintenance Planning Optimisation: B737 NG
Using data collected by the technical records department over a period of three years
(June 1998 till June 2001) for the B737 NG fleet, the following aircraft utilisation
patron can be generated:
Average Aircraft Utilisation - Actual Utilisation
14
12
Flight Hours/Cycles per Day
10
6 Flight Hours/Day
Flight Cycles/Day
0
March
March
March
May
May
May
April
April
April
July
November
December
July
November
December
July
November
December
February
February
February
January
January
Jun-98
January
Jun-99
Jun-00
August
August
August
September
October
September
October
September
October
Calendar Months
Figure 3.3: Actual aircraft utilisation pattern B737 NG [June 1998-June 2001]
The continuous lines illustrate the average fleet utilisation for the indicated months,
while the dotted lines indicate the average utilisation over the three years.
Average annual utilisation per aircraft: 3747 flight hours, 1364 flight cycles
Average FH/FC Ratio: 2.74 FH
At this point, it can be concluded that the METALS utilisation is not representative
for the actual utilisation of the B737 NG.
34
Chapter 3
It is, therefore, preferred to use pattern that is based on Figure 3.3, rather than that
from METALS.
By combining the aircraft utilisation with the maintenance items, maintenance due
dates per item can be generated
After determining maintenance due dates and Items, the next step in the bottom-up
approach is to determine when these items can be performed (hence scheduling).
This section will mainly focus on factors influencing and limiting the scheduling of
aircraft maintenance.
Line Maintenance:
As described earlier in section 2.3.1, the scheduling of line maintenance follows
guidelines laid down in a Maintenance-Planning guide. This guide allocates eleven
slots a week to the technical department for hangar maintenance. The maintenance
planning then sub-divides these slots over the three types of aircraft in the Transavia
fleet.
For the purpose of definition only, the following values of slot allocation will be
used:
Aircraft Type Fleet Size Minimum slots Hangar Visit Aircraft Visiting Total slots per
per hangar visit frequency hangar per week aircraft type per
per aircraft (Routine Mx) week (approx)
737-300 7 2 4 weeks 1.75 3.5 (4)
757-200 4 3 5 weeks 0.8 2.4 (3)
737 - NG 13 1.5 5 weeks 2.6 3.9 (4)
Slots required Per week (Routine Mx) 9.8 (11)
Table 3.5 Minimum maintenance slot allocations for the Transavia fleet (Sources: MP&S)
This research is on the 737NG, and will therefore concentrate on the maintenance
demand for this aircraft. Table 3.5 reveals that a maximum of 4 slots are available per
week for routine maintenance of the 737NG fleet. This also translates to 15 slots a
year.
In order to determine how frequent the aircraft has to visit the hangar for
maintenance, it is necessary to look at intervals of maintenance tasks performed
during such visits. For this purpose, only the common variations in the Maintenance
Task Package intervals will be analysed:
35
Maintenance Planning Optimisation: B737 NG
Table 3.6 Line Maintenance Task Package Limits (Sources: OMP/ 737NG MPD; Table I. 3)
From Table 3.6 above, it can be seen that an aircraft will have to visit the hangar, at
least once every 60 days, or after it makes 450 flight hours or 300 flight cycles. These
are actually the maintenance intervals that determine how often the aircraft should
have a hangar visit. The limits above are, in actual fact, independent of each other,
and do not reflect the relationship between flight hours per day or flight hour per
cycle ratios. The process from which the interval limits are determined falls outside
the scope of this research.
Another observation from Table 3.6 above is illustrated in Figure 3.4 below:
Man-hour Proportions by Package Interval Limits- Line Maintenance
Limited By Calendar
Limited By Calendar
Days/Flight Hours
Days/Flight Cycles
2%
35%
36
Chapter 3
should make less than 90 hours a week in order to remain within this bracket. If the
aircraft makes 100 hours a week, 450 hours will be achieved in 4.5 weeks, at which
the first hour limit (450 hours -see Table 3.6) will have to be performed. Below this,
the maintenance will always be Calendar Day-driven. This fact will be illustrated in
section 4.4.3.
By combining the limits (in Table 3.6) above with the aircraft utilisation (see Figure
3.3 above), the actual due dates for each Maintenance Task Package can be
determined. From here, the exact aircraft maintenance demand can be established.
If it is decided to perform each Maintenance Task Package once it is due, the aircraft
will be visiting the hangar, as frequently as all checks are due. This is not practical,
owing to the to the following facts:
- The Task Packages have different weights (number of due items, total man-
hours demand and total ground time demand), and such visits would lead to
large variations in manpower demand and hangar space demand.
- Such a strategy would have a heavy impact on aircraft utilisation, an
undesired situation.
- The frequency with which the aircraft visit the hangar will exceed the number
of hangar slots allotted for each aircraft type (see Table 3.5)
The bottom-up approach can, therefore, not be utilised on its own in the creation of
optimum maintenance clusters for line maintenance. This creates the need to utilise a
second approach (the Top-down approach, section 3.3) in conjunction with the
Bottom-up approach.
Base Maintenance:
It has been stated earlier that base-maintenance is contracted out to an authorised
base-maintenance station. Base Maintenance Checks are also referred to as Heavy
Maintenance Checks owing to the nature of work they encompass, and the amount
of downtime required to perform the work.
Maintenance Check Intervals Boeing Man-hours Maintenance Check Intervals Boeing Man-hours
540 DAYS 36 8000 HOURS 4.35
720 DAYS/ 4000CYCLES 0.4 10000 HOURS 0.5
1620 DAYS/ 12000CYCLES 3.84 12000 HOURS 5
1800 DAYS 0.8 15000 HOURS 2
1800 DAYS/18000CYCLES 17.6 20000 HOURS 2.15
2160 DAYS/ 18000 CYCLES 0.5 22400 HOURS 1.2
2880 DAYS 5.1 25000 HOURS 1.6
3600 DAYS 14.2 30000 HOURS 1.3
3600 DAYS/ 36000CYCLES 49.38
4320 DAYS 4.3
4320 DAYS/ 36000CYCLES 62.65
5400 DAYS 8
7200 DAYS 0.4
Table 3.7 Transavia Base Maintenance Checks (Sources: 737 NG OMP/ 737NG MPD; Table I. 2)
37
Maintenance Planning Optimisation: B737 NG
As was the case in Line maintenance, the maximum Base-Maintenance intervals will
be determined by the 540-day check, or by the 8000-hour check or the 9000-cycle
check. Once again, these three checks are not interrelated.
In the following chapters, the 540-Day check will be considered as a Line
Maintenance Check (as agreed upon by the Engineering and the MP&S department.
One observation from the table above is that Base maintenance is essentially
Calendar day-driven (See also Figure 3.5 below).
The bottom-up approach can also not be utilised on its own in the formation of
maintenance clusters for base maintenance, as indicated by the fact stated above. The
planning hereof is done on a Top-down principle, as explained in the next section.
Ad.1
38
Chapter 3
The annual utilisation rhythm (Figure 3.3) gives an indication on aircraft demand
spread out over a year. From this, a seasonal pattern can be established, as illustrated
in Figure 3.6 below.
450
> 12 hours per day
400
350
250
High Season
Flight
200 Low Season Low Season Hours
> 4 cycles per day Flight
150 Cycles
50
0
Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec
Period [Months]
The analysis of the utilisation at an annual level provides for the creation of
maintenance slots at an early stage (long-term planning, section 2.2.1).
Further, it may also be said that the fleet size also varies with the seasons. The High
season is normally characterised with the increase in the number of aircraft through
leasing, while the low season is characterised by a decrease in the fleet size, coupled
with the leasing out of aircraft. However, the average utilisation (Figure 3.6) remains
stable, owing to the fact that is dependent on the capacity demand.
Ad.2
The weekly utilisation rhythm gives an indication on trends around aircraft
utilisation on a weekly basis, over a whole year. From this, it may be possible to
establish what days of the week are more convenient for performing maintenance,
and days in which maintenance should be kept at a minimum.
Below are two figures drawn from the utilisation of a sample aircraft from the 737NG
fleet. Figure 3.7 is drawn from the low season, while Figure 3.8 is from the high
season.
39
Maintenance Planning Optimisation: B737 NG
21:36
19:12
16:48
14:24 Mon
Tue
12:00
Up-Time
Wed
Thu
9:36 Fri
Sat
7:12 Sun
4:48
2:24
0:00
Nov Dec Jan Feb Mar
Period [Months] Up-Time: Includes Flight Time, Turn-around Time
and Time Between two flights (<3hrs)
From the figure above, it is not possible to determine whether the aircraft sampled
flies more frequently on specific days as compared to other weekdays. The only
observation that may follow from the graph is that Saturdays and Sundays are
relatively busier than weekdays.
A reason for the irregular variation lays in the fact that aircraft utilisation (in the low
season) is solely based on capacity demand. The type of aircraft utilised for a
particular destination will depend on the amount of seats required. A weekly
pattern can, therefore, be derived form the total fleet flight hours flown on a specified
day of the week. This would not be representative for the B737 NG.
0:00
21:36
19:12
16:48
Mon
14:24 Tue
Up-Time
W ed
12:00 Thu
Fri
9:36 Sat
Sun
7:12
4:48
2:24
0:00
Apr May Jun Jul Aug Sep Oct
Period [Months]
40
Chapter 3
As was the case with the low season, it is not possible to draw a conclusion from the
weekly pattern in the high season. However, by observing the months June to
August, it may be mentioned that Mondays seem to have a lower utilisation as
compared to other days of the week. The utilisation then seems to increase gradually
with the week.
A limiting factor in making use of the weekly utilisation pattern is the distribution of
hangar labour over the week (refer to section 2.3.1). Following the observation made
above on utilisation on Mondays, it may also be said that there is no labour available
for the hangar day shift on this day. The same case applies for Fridays and the whole
weekend.
Ad.3
Daily utilisation patterns give an indication of day periods when the aircraft is
least/not utilised, and the duration of this periods.
The daily utilisation is principally aircraft specific (Scheduled flights) and seasonal
dependent. The seasonal dependence is more evident from the departure times in the
two seasons (Figure 3.9). Aircraft-specific utilisation is entirely dependent on how
the Capacity-Planning Department plans in the aircraft.
UPTIME
APRIL 18:00
MAY 19:00
JUNE 18:15
JULY 19:15
AUGUST 17:30
SEPTEMBER 18:30
OCTOBER 18:00
NOVEMBER 13:45
DECEMBER 13:00
JANUARY 13:30
FEBRUARY 13:00
MARCH 15:00
0000. 0100 0200 0300 0400 0500 0600 0700 0800 0900 1000 1100 1200 1300 1400 1500 1600 1700 1800 1900 2000 2100 2200 2300 2400 0100
Figure 3.9 Average earliest departure and latest return times 737 NG (FLASH, 1999-2000)
Figure 3.9 above reveals that aircraft depart earlier and return late during the high
season, while the opposite is the case during the low season. Since the table contains
averages, the actual distribution of departure and return times may vary greatly,
especially during the low season.
41
Maintenance Planning Optimisation: B737 NG
Maintenance Slots
Fixed Slots Ad-hoc slots Fixed Slots Ad-hoc slots Fixed Slots Ad-hoc slots
It has already been mentioned that a specified number of slots (fixed See also Table
2.5) are made available for Line maintenance per year per (Long-term planning).
Specification of which aircraft should undergo maintenance is done on a short-term
basis (>48 Hours, < 3 Months before maintenance), based on the maintenance
demand of each aircraft. This possibility is also supplemented by the fact that the
737NG fleet is interchangeable. These slots are spread throughout the year, with no
variations between the high and low season.
Base maintenance slots are allocated on an annual basis (fixed slots). These slots are
allocated to specific aircraft tail numbers, depending on when the maintenance is due
to be performed.
Ad-hoc slots can be derived from periods that the aircraft remains unutilised at the
base station. Such slots can be utilised for the execution of opportunity maintenance
pending. These unutilised periods can be illustrated as follows:
0000. 0100 0200 0300 0400 0500 0600 0700 0800 0900 1000 1100 1200 1300 1400 1500 1600 1700 1800 1900 2000 2100 2200 2300 2400 0100
Figure 3.11 Sample times between outbound and inbound flights SPL (Source: FLASH)
42
Chapter 3
Assuming that the above pattern is sustained throughout the high season, it can be
concluded that no opportunities arise from the flight schedule (Opportunity = Time
between 2 flights 3hr 45 min. See also Table 2.5).
0000. 0100 0200 0300 0400 0500 0600 0700 0800 0900 1000 1100 1200 1300 1400 1500 1600 1700 1800 1900 2000 2100 2200 2300 2400 0100
Figure 3.12 Sample times between outbound and inbound flights SPL (Source: FLASH)
Notes:
1. The time between two flights is the difference between the return time of a
flight and the departure time of the next flight.
2. The number of flights indicated does not represent the total flight cycles
made by the aircraft on specified day.
From Figure 3.12, it can be seen that there are enough opportunities available for
planning in maintenance. Such maintenance can either be based on fixed slots or ad
hoc slots.
3.4 Conclusions
The maintenance clustering process may follow from two approach strategies,
namely the Top-down (When can maintenance be performed, and what maintenance
should be performed then?) and the Bottom-Up Approach (What maintenance should be
performed, and when should it be performed?). However, the utilisation of both
approaches is more practical for an airline, in terms of optimising maintenance
planning and execution.
For the optimisation of maintenance task intervals, it would be preferable to utilise
maintenance task packages for line maintenance, instead of the current Maintenance
Checks. Maintenance Task Packages are formed by considering the maintenance task
interval, task set-up activities and the Transavia (L/B) classification. The result of this
is a set of so-called maintenance task packages, which reflect a Task-Based
maintenance (MSG-3), and packages small enough to accommodate Fixed
Maintenance and Opportunity Maintenance.
It has also been illustrated that the high season offers little to no opportunities for
maintenance between flights, meaning that during the season, all maintenance has to
be on a fixed basis. On the other hand, the low season offers enough time between
flights for maintenance, either on a fixed or on an ad-hoc basis. Ad hoc slots may
present themselves on a weekly basis (low season), but they are limited during the
weekends.
43
Maintenance Planning Optimisation: B737 NG
44
Chapter 4
4.1 Approach
The process leading to the formation and evaluation of maintenance clusters is
illustrated in Figure 4.1 below, and is as follows:
Aircraft
Maintenance Maintenance
Utilisation
Items (a) Scenarios (c)
Scenarios (b)
Maintenance
Clusters (e)
Optimum
Cluster Evaluation
Maintenance
(Evaluation Process) (f)
Clusters
45
Maintenance Planning Optimisation: B737 NG
f) Task clusters resulting from (e.) above will then be evaluated (see also
sections 4.4.2 and 4.4.4), and from these, clusters deemed as optimum will be
identified
A choice for a cluster formation and evaluation method can be made between:
1. An analytical method:
The analytical method involves the design of a formal model of the process in
question. This model incorporates all relevant process parameters, and the
relationship between them is presented mathematically. Manipulating or
solving these equations achieves a solution to the problem [Bratley, 1987;
Hillston, 2001]
For this research, it is preferable to utilise the Hybrid Simulation for the creation and
evaluation of maintenance clusters.
The model will be utilised to generate events (maintenance items turning due) and
gather data (man-hours required and task interval de-escalations).
46
Chapter 4
6 [Smit, 1994]
47
Maintenance Planning Optimisation: B737 NG
MATLAB can only be used for its analytical power, but it has poor user-friendly
features. SimJava, on the other hand, is a perfect tool for carrying out simulations,
but the language is unknown to the researcher.
Hence, the software to be utilised in this research will be MS Excel 2000 and Visual
Basic 4 (VB4), in conjunction with the Visual Basic Editor (VBE) present in excel. VB4
will be utilised for programming purposed while MS Excel sheet properties will be
used for processing the inputs and presenting the outputs.
7. Event Routine This routine resets values to zero once the tasks are
performed
8. Library routine In order to make the evaluation process comprehensive
enough, different utilisation scenarios are simulated
(see sections 3.2.3, 3.3.1 and Ad.8. below)
9. Report Generator Data gathered from the simulations includes the
Maintenance demand, and the total de-escalation
Table 4.2 Simulation Components description
48
Chapter 4
Base maintenance is performed at large intervals, currently 1.5 years (see section
2.4.1). The simulations will therefore be aimed at evaluating the effect of increasing
this interval (the desired situation). Base maintenance frequencies will be set at:
Owing to the fact that the seasonal pattern has to be incorporated in the calculations,
it is preferred not to describe the utilisation with a normal (statistical) distribution.
Instead, The Monte-Carlo simulation method is utilised. A uniformly distributed
range is specified, and from this, a random number is drawn (see also Appendix J.3).
The VB4/Excel function utilised to generate values is the RANDBETWEEN [min,
max] function.
Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec
Hours Min 236 197 251 284 355 328 373 377 346 324 114 183
Max 305 288 337 405 398 385 429 407 388 396 280 307
Cycles Min 76 71 72 101 132 122 140 135 122 105 69 63
Max 114 109 126 133 141 142 153 149 149 143 131 101
FH/FC Min 2.1 1.8 1.9 2.1 2.5 2.3 2.4 2.5 2.3 2.3 0.9 1.8
Max 4.0 4.1 4.7 4.0 3.0 3.2 3.1 3.0 3.2 3.8 4.1 4.9
The simulation program samples the min/max FH or FC values for the each month
from the table above.
The FH/FC ratio tabulated serves to indicate the extremes of all possible
combinations
49
Maintenance Planning Optimisation: B737 NG
FH/FC ratio in the high season. The latter is of importance to Transavia, considering
the developments around their BASIQ AIR services.
For these reasons, various utilisation scenarios are developed to describe possible
utilisation patterns that may arise in the periods being considered.
Such utilisation scenarios may, therefore, be based on three assumptions, as
described by Bratley (1987). [Bratley, 1987]:
For each of the scenarios listed above, different Flight hour/cycle ratios are
considered in order to calculate flight cycles to correspond to the flight hours from
the three scenarios above.
The ratios mentioned above can be derived from the anticipated concentration of
Transavia activities to BASIQ AIR destinations. The current destinations of BASIQ
AIR are:
- Nice (2 hours flight time)
- Barcelona (2.2 hours flight time)
- Malaga (2.9 hours flight time)
The following ratio ranges are therefore chosen to represent various modes of
operation that may be witnessed within the airline:
a. Conservative: 1.9 - 2.1 FH/FC
b. Most Likely: 2.2 - 2.7 FH/FC (including the average utilisation)
c. Optimistic: 2.8 - 3.1 FH/FC
50
Chapter 4
From the flight hours and flight hour/cycle ratios above, it is possible to calculate
minimum and maximum utilisation ranges for each month. See also Appendix J.3
The simulation process described for the actual utilisation is also applicable for the
variable utilisation.
For the purpose of uniformity, the actual utilisation will be referred to as Scenario 10.
There are, therefore, 10 Scenarios to be considered in the calculations.
Figure 4.1 illustrates that maintenance clusters result from a clustering process,
which adds maintenance scenarios, utilisation scenarios, and boundary conditions to
the maintenance items.
This section will model the Hybrid Simulation method (selected in the previous
section) into a clustering model, referred to as the Maintenance Item Allocation
Model.
Process description:
The first section of Figure 4.1 describes the clustering process (while the second
section describes the optimisation process). This first section is elaborated as follows:
INPUTS PROCESS OUTPUTS
a). INPUTS/OUTPUTS
Considering that the inputs listed above comprise of System State Variables, the
Simulation Clock and the Library Routine (from Table 4.2), and considering that the
clustering process describes the Bottom-Up/ Top-Down approach, the following can
be said on about the inputs and outputs:
- The Maintenance Item interval induces a Time Since Last Performed variable
(in Days, Hours or Cycles, depending on the interval)
- The Maintenance Scenarios govern the frequency with which maintenance is
to be performed
- The aircraft utilisation not only follows a seasonal pattern, but it also
accumulates over the periods to be considered
- The Maintenance Clusters consist of a group of Maintenance Items (Line
Maintenance Task Packages or Base Maintenance Checks), which exhibit a
maintenance demand (in either Transavia Man-hours or KLM Man-hours),
and the total de-escalation resulting from the clustering (see section 4.3.5).
51
Maintenance Planning Optimisation: B737 NG
As described earlier in this chapter (see Table 4.2), the MIAM is modelled to serve the
following purposes:
1. To Simulate the aircraft utilisation
2. To calculate when a maintenance item turns due
3. To fit each maintenance items into a maintenance cluster
4. To generate maintenance clusters (see outputs, Figure 4.2)
The initial process within the MIAM (calculating when an item is due) can be
Reset TSLP to
0
Maintenance
Due item
Maintenance Item variable:
1. Time Since Last Performed OUTPUT
Maintenance Due Item
Maintenance Item Constants: 1. Date when THIS item is due
a. Maintenance Interval (Mx item) 2. Item Man-hours required to
b. Maintenance Item perform THIS item
Man-hours (Mx item)
illustrated as follows:
A Maintenance Item is introduced into the MIAM. Further, the library routine is
selected (utilisation, based on Scenarios 1-10, Table 4.5). A Maintenance Scenario
(Simulation Clock) is also selected from Table 4.3. The simulation increment is in
months.
The library routine draws a seed (monthly utilisation in flight hours and cycles) from
the scenario in question. The seed is added to the TSLP, where after an evaluation is
done on whether the items maintenance interval has been achieved. If this is the
case, the item is due for performance, it is marked as a maintenance due item. TSLP
is reset to 0, and the process is repeated. If the above is not the case, the process is
repeated until the maintenance interval is reached. If the item is a calendar-day
limited item, then the days of the month in question are utilised (i.e. 31 days for
January, 28 days for February etc), instead of the library routine.
The process it terminated at the end of the simulation period.
It is important to mention that for each month, the same seed is used to evaluate all
Maintenance Items. A new seed is drawn every time the calculation is repeated
The model described in Figure 4.3 does not result in a maintenance cluster, but it
does lay the basis on which the clustering process should be done. The outputs of the
model are:
52
Chapter 4
The next two sections will develop the model in Figure 4.3 into clustering models for
Line Maintenance (performed by Transavia) and Base maintenance (performed by an
MRO, in this case KLM).
Add package
To slot cluster
Package in
Package in
Window
OUTPUTS Slot cluster
Cluster
*= Due now or before the next maintenance interval
53
Maintenance Planning Optimisation: B737 NG
Maintenance
OUTPUT Cluster
The process illustrated in Figure 4.5 above is similar to that used in line maintenance
clustering. However, the determination of Maintenance Checks to execute per base
maintenance visit is much easier, mainly due to the large limits that Maintenance
Checks have.
Each of the above models results in a program that calculate the desired clusters. The
resulting programs (written in Visual Basic Code) are included in Appendix J.4.
TAVi lastp
de escalation = ; lastp TAVi
TAVi
54
Chapter 4
In terms of the exact cost of labour, the following values are applicable:
55
Maintenance Planning Optimisation: B737 NG
Verification: MIAM Design and MIAM Realisation
Verification is intended to ensure that the model does what is intended to do (often
referred to as debugging)
a). The model calculates all possible inputs; with the exclusion of line maintenance
frequencies lower than 4 weeks. If certain boundary conditions are violated (e.g.
escalation of maintenance interval limits), the model returns erroneous outputs
(N/A, #VALUE etc.). This ensures that no invalid results are evaluated further.
The results resulting from the MIAM can be grouped into two categories, each
resulting from the two models, namely the Line Maintenance and the Base
Maintenance MIAM. For both categories, the Current Situation (as described in
chapter 2) is analysed, after which the Proposed (New) Situation (as proposed in
chapter 3) is analysed. The results in section 4.4 are also included in Appendix K.
Unless otherwise stated, the following initial conditions will apply:
A. Current Situation:
The first results generated by the MIAM are based on the conditions that prevailed at
the start if this research, namely:
- Base Maintenance Items Include all Maintenance Checks with an interval equivalent
to, or larger than 540 Days, 3000 Cycles or 6000 Hours (See also and Table I. 2 in
Appendix I.1)
Items on Calendar Time: 1276 Base Man-hours
Items on Flight Cycles: 41 Base Man-hours
Items on Flight Hours: 100 Base Man-hours
- Base maintenance is performed every 18 months
56
Chapter 4
The purpose of this is to form a basis of evaluation and comparison with results
generated under different conditions.
The results will also be based on Scenario 10 (See Table 4.5). This is because they will
not be optimised further.
By utilising the MIAM, the following results are generated, under the conditions
stated above:
800
700
600
Base Mx Manhours
500
Cluster MHrs
400
Cluster De-Escalation
300
200
100
0
HMV1 [Sep 02] HMV2 [Mar 04] HMV3 [Sep 05] HMV4 [Mar 07] HMV5 [Sep 08] HMV6 [Mar 10]
Base Visit [Number, Date]
Table 4.8 and Figure 4.6 reveal that for every base maintenance visit, and average of
23% in terms of labour costs is wasted (see also Table 4.9 below).
57
Maintenance Planning Optimisation: B737 NG
It can also be observed that the first five visits (HMV1 HMV5) do not vary much in
terms of man-hour demand, as compared to the sixth visit. The sixth visit (HMV-6)
exhibits a large maintenance demand, mainly as a result of an increase in the number
of maintenance tasks that comes along with the ageing of the aircraft.
It was established in section 2.4.1 that the proportions of Routine work and non-
routine work resulting from routines are:
Routine Man-hours: 2.0 * Boeing Man-hours
Non-routines (from routines): 1.6 * Boeing Man hours
Total Routine Man-hours: 3.6 * Boeing Man-hours
By incorporating this information in Table 4.8, and by adding open and close man-
hours to the routine maintenance, the following may be seen as an illustration of the
proportions between routine and non-routine maintenance (resulting from routine
work):
1200
1000
800
Base Mx Man-hours
Routine MHrs
400
200
0
HMV1 [Sep 02] HMV2 [Mar 04] HMV3 [Sep 05] HMV4 [Mar 07] HMV5 [Sep 08] HMV6 [Mar 10]
Base Visit [Number, Date]
The maintenance demand illustrated above can be translated into maintenance costs
by incorporating the base maintenance rate stated in Table 4.6. The result of this is
given in Table 4.9 below. Observations from the table will be discussed in section
4.4.2 on optimisation.
58
Chapter 4
B. Proposed (New) Situation:
The proposed situation refers to the clustering that is done under the following
conditions:
- Base Maintenance items are not determined by their intervals (>3000 Cycles,
> 6000 Hours, > 540 Days), but by their Line/Base Classification. See also
section 3.2.1 and Figure 3.1.
Items on Calendar Time: 813 Base Man-hours
Items on Flight Cycles: 14 Base Man-hours
Items on Flight Hours: 65 Base Man-hours
- All maintenance scenarios on Table 4.5 may be considered, owing to the fact
that the model will be looking at the current situation and projecting into the
near future (0 10 years)
This situation utilises the simulation clock values, as listed in Table 4.3. Each interval
will be analysed separately.
Owing to the fact that different scenarios result in different cluster compositions
(illustrated on Figure 4.14), it can be said that an 18-Month Base Maintenance
Interval would demand a careful monitoring of the aircraft utilisation. Large
variations in the utilisation would require a constant revision of the maintenance
clusters that might have been pre-calculated and pre-planned. This is especially so
when the aircraft utilisation should not be limited by the anticipated maintenance.
When plotted out on Figure 4.8 below, it becomes clearer that the Optimistic
Scenarios slightly results in a higher maintenance demand than other scenarios. This
is as a result of the maintenance demand leaning slightly on utilisation, rather than
on calendar time, as is the case with other utilisation scenarios.
59
Maintenance Planning Optimisation: B737 NG
750
650
550
Base Mx Manhours
450
350
250
Cons
ML
150 Opt
Scen.10
Current
50
-50 HMV1 [Sep 02] HMV2 [Mar 04] HMV3 [Sep 05] HMV4 [Mar 07] HMV5 [Sep 08] HMV6 [Mar 10]
Figure 4.8 Cluster Maintenance demand by Scenario, as compared to the Current situation
It is also evident from Figure 4.8 that there is a large difference in the maintenance
demand between the current and the proposed situation. This is a logical
consequence of the Line/Base Maintenance redefinition (Appendix E), which
resulted in many tasks being classified as line maintenance tasks than was previously
the case.
As pertains to the maintenance costs associated with the clusters from the 18-Months
maintenance interval, the following is a tabulation of how these relate:
Mx Costs () Cons. ML Opt. Scenario 10
Labour De-esc Labour De-esc Labour De-esc Labour De-esc
Type Visit Cost Cost Cost Cost Cost Cost Cost Cost
HMV1 [Sep 02] 10900 997 10994 929 11931 1198 10994 896
HMV2 [Mar 04] 11931 1315 12306 1159 12306 1356 12306 1050
HMV3 [Sep 05] 15444 1356 15566 1275 17028 1595 15791 1294
HMV4 [Mar 07] 12428 1369 12925 1211 12644 1381 12700 1109
HMV5 [Sep 08] 15967 1112 15780 1002 16473 1271 15780 938
HMV6 [Mar 10] 28637 2765 28796 2558 29321 2836 29021 2540
Totals ()/ aircraft 95308 8915 96367 8133 99704 9637 96592 7828
Table 4.11 Base Maintenance Cluster Costs by Scenario - 18 Months
24 Month Interval for Base Maintenance Visits
As was with the other interval, the scenarios analysed will be limited to Scenario 1
(Scen-1), 5 (Scen-5), 9 (Scen-9), and the actual scenario.
Cons ML Opt Scenario 10
Type Visit Base MHrs % De-esc Base MHrs % De-esc Base MHrs % De-esc Base MHrs % De-esc
HMV1 [Mar 03] 82 9 100 12 107 10 107 13
HMV2 [Mar 05] 187 12 195 14 205 13 199 15
HMV3 [Mar 07] 95 8 111 9 113 8 119 10
HMV4 [Mar 09] 509 8 515 9 521 9 515 10
HMV5 [Mar 11] 327 2 341 3 348 2 348 3
Totals 1200 7 1261 8 1293 8 1287 9
Table 4.12 Base Maintenance Cluster man-hours by Scenario
60
Chapter 4
Table 4.12 shows that with a larger base maintenance interval, the maintenance
demand still varies between various scenarios. Further, the actual scenario tends to
move closer to the Optimistic utilisation. Further, there is a significant reduction in
the total maintenance demand as compared to the original situation (see Figure 4.1
below), and also as compared to the maintenance demand at an interval of 18
Months. The latter is the case; as a result of a reduction in work that recurs per
maintenance visit.
An additional advantage of the 24-Months interval is that there are only five base
maintenance visits in ten years, as compared to the previous six in the same period
750
650
550
Base Mx Manhours
450
350
250
Cons
ML
150 Opt
Scen.10
Current
50
-50 HMV1 [Mar 03] HMV2 [Mar 05] HMV3 [Mar 07] HMV4 [Mar 09] HMV5 [Mar 11]
The Cluster maintenance demand illustrated in Figure 4.9 shows relatively large
variations between various visits, and especially so during the fourth and the fifth
visit.
By comparing the total man-hour demand from an 18-month and a 24-month Base
maintenance interval (Scenario 10), it can be seen that the 24 month interval results in
a lower man-hour demand.
Base Mx at 18 Months Base Mx at 24 Months
Base Visit Base Mhrs Base Visit Base Mhrs
HMV1 [Sep 02] 211 HMV1 [Feb 03] 107
HMV2 [Mar 04] 236 HMV2 [Feb 05] 199
HMV3 [Sep 05] 303 HMV3 [Feb 07] 119
HMV4 [Mar 07] 244 HMV4 [Feb 09] 286
HMV5 [Sep 08] 303 HMV5 [Feb 11] 336
HMV6 [Mar 10] 557
Total Base Mhrs 1855 Total Base Mhrs 1046
Table 4.13 Base Man-hour demand Comparison: 18 Month and 24 Month Intervals
61
Maintenance Planning Optimisation: B737 NG
Following an analysis of the exact contents of the clusters illustrated the following
observations are made:
Bo-Mhrs
CHECKS
Cycles
Hours
Days
720 7742 2946 HMV1 HMV2 HMV3 HMV4 HMV5
D720* 720 4000 50 x x x x x
D1620* 1620 12000 14 x x
D1800 1800 3 x x
D1800* 1800 18000 63 x x
D2160* 2160 18000 2 x
D2880 2880 18 x
D3600 3600 51 x
D3600* 3600 36000 178 x
D4320 4320 15 x
D4320* 4320 36000 226 x
Figure 4.10 is a visualisation of how the MIAM identifies due items (x) and groups
them into clusters (HMV#).
The rows at the top stand for the aircraft age (in calendar days, hours and cycles),
while the first column of the left lists the maintenance checks. The second, third and
fourth columns are help columns, while the fifth column lists the base maintenance
man-hours.
From the extract, it can be seen that the D1800*, the D3600, D3600* and the D4320*
checks carry a significant amount of man-hours with them. These checks have,
however, been clustered for performance much earlier (~ 680 days) than they are
due. This is a large de-escalation of the check interval. Further, it is also evident that
in HMV4, there are many other checks that turn due for performance. The result of
this is a large maintenance cluster, hence the peak in Figure 4.9.
Figure 4.10 below illustrates the effect of the initial de-escalation on the total
Maintenance demand and on the total de-escalation (see also Table 4.14):
62
Chapter 4
400
1250
350
1050
300
Scen. 10 Mhrs
850 Cons Mhrs
De-escalation Manhours
250
Base Mx Mnhours
ML Mhrs
Opt Mhrs
De-esc. Scen 10 200
650
De-esc. Cons
De-esc. ML
De-esc. Opt 150
450
100
250
50
50 0
0 30 60 90 120
Initial De-escalation [Days]
On the other hand, the total de-escalation shows a sharper decline for an initial de-
escalation of 30 days (*). By values above this, the total de-escalation increases
gradually. The principle reduction (from 0-days), amounts to 26.9%
Scenario 1 shows the lowest maintenance demand and also the lowest de-escalation.
This follows from the fact that this scenario represents the lowest utilisation
considered in the MIAM.
Figure 4.12 below illustrates the effects of the initial de-escalation on the situation
illustrated in Figure 4.10. The maintenance checks that had previously been de-
escalated heavily are now clustered for performance at their exact due dates.
63
Maintenance Planning Optimisation: B737 NG
Base Mhrs
CHECKS
Cycles
Hours
Days
720 7604 2937 HMV1 HMV2 HMV3 HMV4 HMV5 HMV6
D720* 720 4000 50 x x x x x
D1620* 1620 12000 14 x x
D1800 1800 3 x x
D1800* 1800 18000 63 x x
D2160* 2160 18000 2 x
D2880 2880 18 x
D3600 3600 51 x
D3600* 3600 32000 178 x
D4320 4320 15
D4320* 4320 32000 226
At this point, it can be concluded that there are variations needed within the 24-
Month interval, i.e. the implementation of an initial de-escalation of between 1 - 4
Months. This means that the interval will not be a pure 2-year Base Maintenance
interval.
The maintenance demand resulting from the initial de-escalation on Figure 4.12 is
also illustrated on Figure 4.13 below.
Base Maintenance Demand: Before/After a 30-day Initial De-escalation (Scenario 10)
600
500
400
Base Mx Man-hours
300
0
HMV1 [Feb 03] HMV2 [Feb 05] HMV3 [Feb 07] HMV4 [Feb 09] HMV5 [Feb 11]
Base Visit [Visit#, Date]
Figure 4.13 Base Maintenance Demand: Before/ After 30 days Initial de-escalation (i.d)
Note: i.d.: Initial de-escalation
64
Chapter 4
It can be concluded at this point that an optimum clustering for base maintenance is
realised by a Base Maintenance frequency of 24 Months, and an initial de-escalation
of 30 days
It can also be said that it is possible to standardise Base maintenance clusters, owing
to the static nature of the cluster composition (mainly calendar-time driven). The
exact clusters belonging to the various utilisation scenarios can be derived from the
following cross-matrix scheme:
2003 2005 2007 2009 2011 2003 2005 2007 2009 2011 2003 2005 2007 2009 2011 2003 2005 2007 2009 2011
Feb Feb Feb Feb Feb Feb Feb Feb Feb Feb Feb Feb Feb Feb Feb Feb Feb Feb Feb Feb
Days 699 1429 2159 2889 3619 699 1429 2159 2889 3619 699 1429 2159 2889 3619 699 1429 2159 2889 3619
Hours 7111 14828 22125 29842 37301 5553 11315 17240 23002 28965 6839 14139 21375 28675 35968 8443 17377 26109 35043 43804
Cycles 2838 5771 8707 11640 14544 2742 5704 8577 11539 14441 2864 6023 9022 12181 15232 2862 5817 8777 11732 14701
BaseMhrs
CHECKS
Scen-1
Scen-5
Scen-9
Actual
HMV1 HMV2 HMV3 HMV4 HMV5 HMV1 HMV2 HMV3 HMV4 HMV5 HMV1 HMV2 HMV3 HMV4 HMV5 HMV1 HMV2 HMV3 HMV4 HMV5
D720* 49.68 x x x x x x x x x x x x x x x x x x x x
D1620* 13.824 x x x x x x x x
D1800 2.88 x x x x x x x x
D1800* 63.36 x x x x x x x x
D2160* 1.8 x x x x
D2880 18.36 x x x x
D3600 51.12 x x x x
D3600* 177.768 x x x x
C3500 0 x x x x
C9000* 14.4 x x x x x x x x x x x x x x x x x x x x
C24000* 49.068 x x x x
C36000* 19.8 x x x x
H8000 15.66 x x x x x x x x x x x x x x x x x x x x
H10000 1.8 x x x x x x x x x x x x x x x x x x x x
H12000 18 x x x x x x x x x x x x x x x x x
H15000 7.2 x x x x x x x x x x x
H20000 7.74 x x x x x x x
H22400 4.32 x x x x x
H25000 5.76 x x x x x
H30000 4.68 x x x x
0
Base maintenance clusters can also be classified as Phases, with each phase being a
composition of maintenance checks that are due for performance at the same time.
From Figure 4.14, the phasing of clusters should be done for three different
situations, namely Low, Normal, and High aircraft utilisation.
65
Maintenance Planning Optimisation: B737 NG
The optimum 2-year combinations (also denoted as HMV #) for all scenarios can be
made as follows (see Table 4.15, Table 4.16, and Table 4.17):
66
Chapter 4
As expected, Table 4.19 shows that a 30-day initial de-escalation leads to the highest
reduction in the total maintenance costs and de-escalation losses.
By considering the Actual utilisation scenario, the savings on maintenance cost
translates to an average of 2781 per Base Maintenance visit; with the whole fleet
(13 aircraft) taken into consideration, 36147 is saved per maintenance visit.
The results of the line maintenance analysis in the MIAM will be limited to a period
of five to six years. As stated in section 2.3, this is the period in which the aircraft is
considered as new.
A. Current Situation
As was the case with Base Maintenance, the current situation for line maintenance
encompasses all the maintenance checks bound by the following conditions
- All maintenance items with maintenance intervals lower than 540 Days, 3000
Cycles, and 6000 Flight Hours.
Items on calendar Time: 37 Transavia Man-hours
Items on Cycles: 14 Transavia Man-hours
Items on Flight Cycles: 124 Transavia Man-hours
- Hangar maintenance is performed every 5 weeks, in the high and in the low season,
and it follows the sequence on Table 2.5
The conditions above results in the following maintenance demand per year (based
on Days, Hours and Cycles) in terms of Transavia man-hours (1.7 x Boeing Man-
hours):
The variation in the Scenario maintenance man-hours required per year is not large,
and this is also the case with the total maintenance demand for the period
considered. However, there is a significant difference between maintenance resulting
from the conservative utilisation and from the other scenarios. For illustration
purpose only, scenarios 1, 4, 7, and 10 are plotted out on Figure 4.15 below to
illustrate these facts.
67
Maintenance Planning Optimisation: B737 NG
550
500
450
Transavia Ma- hours
400
250
200
2001 2002 2003 2004 2005 2006
Period [Years]
The maintenance demand from year 2001 is lower than that from others years, but
this is a direct result of the fact that the airplane calculated is introduced in the fleet
in the course of that year. However, this year has to be taken into consideration
because maintenance performed in that year has an effect on maintenance
performance in the future (time since last performed).
From Table 4.21 above, it can be concluded that the maintenance-planning pattern on
Table 2.5 in chapter 2 is sufficient for the execution of routine line (hangar)
maintenance. This planning provided a maximum of 885 Transavia man-hours per
year for routine hangar maintenance (for each aircraft). This routine maintenance
includes non-routines generated from the routine maintenance.
However, the demand for maintenance slots portrays a different sequence than that
on Table 2.5. Table 4.22 below gives an example of this slot demand.
68
Chapter 4
Table 4.22 Slot planning vs. slots required for routine maintenance
The Shaded area in Table 4.22 represents the high season. The maintenance during
this season is more utilisation-dependent than calendar-time-dependent, hence the
increase in the man-hour demand. This is in contrast with the planning, which
assumes a specific rhythm all year round. It can also bee seen that the slots in the
high season leave little room for non-routine maintenance arising from reasons other
than routine maintenance (e.g. from postponed non-routine maintenance on
Dispatch Deviation Sheets - DDS).
If the extra slots required in Mar-02, Jun-02, Sep-02 are not granted, it will be
necessary to have the aircraft back in the hangar between two consecutive hangar
visits. This will be for the purpose of completing the unfinished work from the slot
previous slot.
Table 4.22 and Table 2.5 may explain the differences exhibited by the realisation of
the planning on Table 2.6, i.e. the large differences between the maintenance
planning and realisation.
By comparing Table 4.20 and Table 4.21, it can be seen that the maintenance demand
resulting from performing line maintenance at a 5-week frequency is higher than the
theoretical maintenance demand performing individual maintenance tasks as they
turn due without clustering or de-escalating their intervals (See Figure 4.16 below).
The reason for this can be attributed to the fact that a lot of maintenance is performed
more frequently that it ought to; this is an inevitable consequence of task clustering.
69
Maintenance Planning Optimisation: B737 NG
4000.0
3500.0
Total Mx Man-hour demand (5 yrs)
3000.0
2500.0
Theoretical
2000.0
Maintenance
Demand
1500.0
5-week Hangar
Visit
Maintenance
1000.0 Demand
500.0
0.0
Scen-1 Scen-2 Scen-3 Scen-4 Scen-5 Scen-6 Scen-7 Scen-8 Scen-9 Actual
Utilisation Scenarios
It may be said that the total de-escalation is large, as compared to the de-escalation
witnessed from Base Maintenance. It is worth noting that the de-escalation decreases
with utilisation, indicating that items mainly de-escalated are those based on the
utilisation, and especially the Flight-hour items.
70
Chapter 4
Maintenance Demand:
New Situation Cons. ML Opt.
Scen-1 Scen-2 Scen-3 Scen-4 Scen-5 Scen-6 Scen-7 Scen-8 Scen-9 Scen-10
2001 218 218 218 218 218 218 218 231 218 218
2002 467 467 467 512 513 513 533 520 533 533
2003 389 401 376 439 446 531 545 550 534 527
2004 483 470 508 540 519 435 568 542 579 438
2005 451 443 438 555 582 594 601 623 601 594
2006 474 494 494 521 534 521 427 427 427 520
Total Transavia MHrs 2480.9 2494.3 2501.6 2785.2 2811.6 2811.6 2891.6 2892.4 2892.0 2829.7
As was the case in base maintenance, and in the current situation, the variations
within the utilisation patterns (conservative, most likely, optimistic) are minimal.
However, the conservative utilisation is considerably less (11%) than the other
utilisation patterns.
By considering samples from the utilisation on Table 4.24, the following comparison
can be made between the demand from the current and the new situation:
3500.0
3000.0 2892.0
2829.7
2811.6
2480.9 2524.4
2429.6 2455.5
2500.0
Transavia Mx Man-hours
2111.8
2000.0
Maintenance Demand:
Current Situation
1500.0
Maintenance Demand:
Proposed Situation
1000.0
500.0
0.0
Cons ML Opt Scen-10
Utilisation Scenario
Figure 4.17 Line Maintenance Demand per aircraft, first 6-yrs: Current vs. proposed
Situation
It follows from the comparison above that the new situation leads to a maintenance
demand increase of 372 Transavia man-hours per aircraft, over six years (on
average).
Table 4.25 below show how the above translates on an annual basis for scenario 10:
71
Maintenance Planning Optimisation: B737 NG
Table 4.25 Line (routine) maintenance demand change: Current vs. proposed Situation
It follows that the increase on Table 4.25 is time-dependent. This results from the fact
that maintenance items previously included in base maintenance are gradually
introduced in line maintenance. This happens gradually because current base
maintenance checks (before Line/Base Redefinition Appendix E) normally posses
large maintenance intervals, and are mainly Calendar-Time-dependent (see also
Table 3.6).
5000.0
4413.8
4235.2
4181.2
4500.0
3927.7
3845.9
3826.4
3788.6
3601.7
3558.3
Maintenance Demand [Transavia Man-hours]
4000.0
3428.5
3407.4
3164.4
3500.0
2892.0
2829.7
2811.6
Pre-calculated
2480.9
3000.0 Maintenance
Demand
2500.0
4 Weeks Mx.
Frequency
2000.0
6 Weeks Mx.
500.0 Frequency
0.0
Cons ML Opt Scen-10
Scenarios
Figure 4.18 5-yr Line Maintenance Demand per aircraft, by a varying line maintenance
interval
The pre-calculated maintenance demand represents the maintenance demand
illustrated on Figure 4.17.
A high maintenance frequency leads to a higher maintenance demand, a fact that can
be drawn from Figure 4.18 above. It can also be seen that by increasing the
maintenance interval, the de-escalation of maintenance intervals is reduced
considerably. In the case above, the total de-escalation may be viewed as the
difference between the Pre-calculated maintenance-demand (1st column) and the
maintenance resulting from the maintenance frequency.
However, the increase in the maintenance interval is not unlimited. Scenarios leading
to utilisation higher than 11 FH/day limit the maintenance frequency to 6-weeks, for
above this, the lowest maintenance interval will be exceeded, and this will be in
violation of one boundary condition. A 7-week frequency is, therefore, only
applicable for the conservative utilisation
72
Chapter 4
900
800
700
Transavia Man-hours
4 weeks
600 5 weeks
6 weeks
500
400
300
2001 2002 2003 2004 2005 2006
Period [Years]
180
166
160
3
133
140
128
Tra nsa via M a n-hours
120
2002
100
2003 2 2002
82
80
80
72
2003
60
59
57
60
51
43
43
42
41
40
39
39
38
35
40 1
31
20
0
0
Jan Feb M ar A pr M ay Jun Jul A ug S ep Oc t Nov Dec 0
73
Maintenance Planning Optimisation: B737 NG
By considering two years, namely 2002 and 2003, the following man-hour/slot
demand is established:
HV MHrs Per a/c Slots Per a/c HV MHrs Per a/c Slots Per a/c
2002 2003 2002 2003 2002 2003 2002 2003
Jan 39 35 1 1 Jan 35 0 1
Feb 38 0 1 Feb 0 77 1 2
Mar 80 133 2 3 Mar 104 106 2 2
Apr 42 41 1 1 Apr 26 30 1 1
May 40 43 1 1 May 149 0 3
Jun 166 51 3 1 Jun 0 67 2
Jul 0 57 1 Jul 73 40 2 1
Aug 114 128 3 3 Aug 98 145 2 3
Sep 59 82 1 2 Sep 54 0 1
Oct 31 0 1 Oct 0 75 3
Nov 39 43 1 1 Nov 52 47 1 1
Dec 60 72 1 2 Dec 47 0 1
Totals 709 685 16 16 639 586 15 15
Table 4.26 Man-hour/slot demand: 5-weeks Table 4.27 Man-hour/slot demand: 6-weeks
M an-hour de mand: 6-we e ks Hangar fre que ncy Slot De mand: 6-we e k Hangar fre que ncy
160
149
145
3.5
140
3
120
Tra nsa via M a n-hours
106
104
2.5
98
100
2
77
2002 2002
75
80
73
67
2003 2003
1.5
60
54
52
47
47
40
1
35
40
30
26
20 0.5
0
0 0
Jan Feb M ar A pr M ay Jun Jul A ug S ep Oc t Nov Dec Jan Feb M ar A pr M ay Jun Jul A ug S ep Oct Nov Dec
P e riod [M onths] P e riod [M onths]
It can be seen from Figure 4.21 that even though the slot distribution is not similar for
the two years considered, both years demand 15 slots in total. The total man-hours
demanded in both situations (5 weeks and 6 weeks) still remain under the maximum
available man-hours for routine maintenance (885 HV Man-hours Table 2.5).
It should be said that the calculation of the slot demand on Table 4.26 and Table 4.29
is done in such a way that the upper limit of the man/hour slot demand is
considered. For example, 39 man-hours = 1 slot. It has already been given that 1 slot
74
Chapter 4
< 55 man-hours (Table 2.5). The surplus man-hours my be made available for non-
routine maintenance, such as that from DDS, MIs and AMs.
It was mentioned at the beginning of this chapter that the optimisation process
mainly involves the reduction of maintenance man-hours, and a maximum
utilisation of the maintenance item interval. From reduced man-hours follows a
reduction in maintenance downtime, and from increased item interval utilisation
follows a decrease in the interval de-escalation.
Line maintenance optimisation will also be aimed at reducing the amount of
maintenance performed in the high season.
It was illustrated in the previous section that an increase in the maintenance interval
also leads to a decrease in the man-hour demand. The 6-weeks maintenance
frequency leads to the following maintenance demand:
180
160
140
120
Transavia Manhou
100
80
60
40
20
0
May-01
May-02
May-03
May-04
May-05
May-06
Nov-01
Nov-02
Nov-03
Nov-04
Nov-05
Nov-06
Sep-01
Sep-02
Sep-03
Sep-04
Sep-05
Sep-06
Feb-02
Mar-02
Feb-03
Mar-03
Feb-04
Mar-04
Feb-05
Mar-05
Feb-06
Mar-06
Aug-01
Dec-01
Jan-02
Aug-02
Dec-02
Jan-03
Aug-03
Dec-03
Jan-04
Aug-04
Dec-04
Jan-05
Aug-05
Dec-05
Jan-06
Aug-06
Dec-06
Apr-01
Jun-01
Apr-02
Jun-02
Apr-03
Jun-03
Apr-04
Jun-04
Apr-05
Jun-05
Apr-06
Jun-06
Oct-01
Oct-02
Oct-03
Oct-04
Oct-05
Oct-06
Jul-01
Jul-02
Jul-03
Jul-04
Jul-05
Jul-06
Period [mm/yy]
Section 3.2.4 mentioned the fact that line maintenance could either be utilisation or
calendar-time driven. The high season is characterised by an increase in aircraft
utilisation, hence an increase in maintenance demand (based on utilisation).
75
Maintenance Planning Optimisation: B737 NG
High Season: April-October (7 Months/214 days, approx 2620 FH/960 FC Scen. 10)
Period Task Package Maintenance interval Occurrence HV man-hours
214 Days 60 Days (0.06 HV MHrs) 3x 0.18
60 Days/400C (5.44 HV MHrs) 3x 16.32
100 Days (22.4 HV MHrs) 2x 44.8
100Days/1000C (2.44 HV MHrs) 2x 4.88
2626 FH 450 Hours (26 HV MHrs) 5x 130
1000 Hours (13.1 HV MHrs) 2x 26.2
1600 Hours (7.7 HV MHrs) 1x 7.7
2000 Hours (6.42 HV MHrs) 1x 6.42
960 FC 300 Cycles (0.68 HV MHrs) 3x 1.02
Minimum Man-hours required 238.54
The occurrence of the task packages listed on Table 4.29 above creates a large
demand for man-hours, and especially so in the high season. Task packages based on
calendar days can be de-escalated (strategy utilised for base maintenance) once, such
that they will always recur in the low season or at specific times of the later year.
As for the FH and FC task packages, their recurrence is based on the utilisation. By
referring to Scenarios 1, 5, 9, and, the following recurrence cycles can be calculated:
76
Chapter 4
outside the high season. Reason: the next due date is bound to fall in the high season,
and the de-escalation process is bound to be repeated annually.
Table 4.32 Line Maintenance Man-hour Demand per aircraft: Optimised situation
77
Maintenance Planning Optimisation: B737 NG
200
180
160
140
Transavia Man-hours
120
100
80
60
40
20
0
Jun-01
Jan-02
Jun-02
Jan-03
Jun-03
Jan-04
Jun-04
Jan-05
Jun-05
Jan-06
Jun-06
Dec-03
May-01
Aug-01
Sep-01
Dec-01
Feb-02
Mar-02
May-02
Aug-02
Sep-02
Dec-02
Feb-03
Mar-03
May-03
Aug-03
Sep-03
Feb-04
Mar-04
May-04
Aug-04
Sep-04
Dec-04
Feb-05
Mar-05
May-05
Aug-05
Sep-05
Dec-05
Feb-06
Mar-06
May-06
Aug-06
Sep-06
Dec-06
Apr-01
Apr-02
Apr-03
Apr-04
Apr-05
Apr-06
Jul-01
Jul-02
Jul-03
Jul-04
Jul-05
Jul-06
Oct-01
Oct-02
Oct-03
Oct-04
Oct-05
Oct-06
Nov-01
Nov-02
Nov-03
Nov-04
Nov-05
Nov-06
P eriod [Months]
The maintenance demand above leads to a maintenance slot demand that increases
gradually over the years, to reflect the increase in the amount of maintenance arising
from maintenance task packages with large maintenance intervals. With the
exception of the first two years (2001, 2002), all high seasons show a demand of 8
slots per year. This trend is tabulated here below, and plotted out on Figure 4.24
below.
78
Chapter 4
3.5
2.5
Maintenance Slots
1.5
0.5
0
May-01
May-02
May-03
May-04
May-05
May-06
Nov-01
Nov-02
Nov-03
Nov-04
Nov-05
Nov-06
Sep-01
Feb-02
Mar-02
Sep-02
Feb-03
Mar-03
Sep-03
Feb-04
Mar-04
Sep-04
Feb-05
Mar-05
Sep-05
Feb-06
Mar-06
Sep-06
Jan-02
Jan-03
Jan-04
Jan-05
Jan-06
Aug-01
Dec-01
Aug-02
Dec-02
Aug-03
Dec-03
Aug-04
Dec-04
Aug-05
Dec-05
Aug-06
Dec-06
Apr-01
Jun-01
Jul-01
Apr-02
Jun-02
Jul-02
Apr-03
Jun-03
Jul-03
Apr-04
Jun-04
Jul-04
Apr-05
Jun-05
Jul-05
Apr-06
Jun-06
Jul-06
Oct-01
Oct-02
Oct-03
Oct-04
Oct-05
Oct-06
Period [M onths]
In terms of de-escalated man-hours, it can be seen from Table 4.34 that depending
on the scenario, the de-escalation lies between 10-19%.
Utilisation Labour Costs- De-esc. Costs % De-esc Labour Cost - De-esc. Costs % De-esc
Current () - Current () Proposed () Proposed ()
Cons. 94394 27323 29 100501 12498 12
ML 101489 24325 24 108218 18916 17
Opt. 106596 26421 25 114391 22540 20
Scen-10 101140 23153 23 108890 19018 17
Total Cost 403619 101222 25 432000 72972 17
Table 4.35 Total Maintenance cost per aircraft, over a 6-year period: Current vs. Proposed
situation
The proposed situation shows a higher man-hour demand than the current
situation. This is a direct result of the increase in the amount of maintenance work
79
Maintenance Planning Optimisation: B737 NG
4.5 Summary
The application of the two approaches also leads to the conclusion that Base
maintenance can be performed optimally at a frequency of 24 Months. However, this
optimum is achieved through the application of an initial de-escalation, which
schedules the performance of the first base visit at a date, not later than 23 Months
after the introduction of the aircraft into the fleet. This leads to the clustering of
maintenance checks to clusters closest tot their maintenance intervals. A 30-day
initial de-escalation leads to the least total de-escalation of check-intervals, and
consequently to the most optimum clusters for Base maintenance.
The clustering of base maintenance checks leads to static maintenance clusters,
which may also be viewed as phases in maintenance. Such phases are of importance,
especially in the planning of long-term maintenance.
80
Conclusions and Recommendations
5.1 Conclusions
The optimisation of the maintenance of the B737 NG can be sub-divided into Line
Maintenance Optimisation and Base Maintenance Optimisation. This follows from
the fact that Transavia performs Line Maintenance in house, and contracts out Base
Maintenance.
Line Maintenance
The shift from the current situation, (where line maintenance is limited by 3000 Flight
Cycles, 6000 Flight Hours and 540 Days), to a Line-Base classification results in an
increase in maintenance demand of between 66-113 man-hours per year.
The demand for maintenance slots does not follow a specific pattern, but is also
dependent on the utilisation. The departure situation was a rhythmic distribution of
maintenance slots (1-1-2-1-1-4-1-1-2-...), irrespective of the season or utilisation. In
reality, the actual demand and distribution of maintenance slots varies from year to
year. This leads to distributions such as 3-2-1-2-1-1-1-3- or 2-2-1-2-1-2-1-1-2-.
However, the total demand for maintenance slots remains at 15 in all cases.
81
Maintenance Planning Optimisation: B737 NG
escalation will have to be done every new season, and the cost of this will be
dependent on the utilisation of the aircraft.
Base Maintenance
Base maintenance items are characterised by large maintenance intervals and a high
demand for man-hours and ground time. The scheduling of this is, therefore, done
only in the low season.
The optimisation of base maintenance follows from the reduction in the frequency (of
the maintenance), a reduction in the number of man-hours contracted out, and a
minimisation in the interval (man-hour) de-escalation.
Maintenance labour forms a component of the direct maintenance costs. It also has a
direct proportionality with the downtime, and the material costs, which are also
elements of direct maintenance costs. In turn, direct maintenance costs are a part of
the cost of ownership. Hence, the reduction of maintenance labour (man-hours)
directly leads to the reduction of the cost of ownership
82
Conclusions and Recommendations
5.2 Recommendations
In order to maximise the maintenance advantages that came along with the B737 NG,
the following is recommended:
4. The use of initial de-escalation at the first base maintenance visit. This should
serve as a tool to reduce the total maintenance man-hours contracted out, and
the total de-escalation resulting from maintenance clustering.
83
References
References
[2] Patton, J.D., Preventive Maintenance, 2nd edition, ISA, North Carolina
(USA), 1995
[3] Kelly, A., Maintenance Planning and Control, Butterworths, Manchester
(UK), 1984
[15] Stam, T., Lems, W., Line and Base Maintenance Redefinition, Maintenance
program concept redefinition, Schiphol East, 2001
[16] Dijkhuizen, G. van, Maintenance Meets Production: On the Ups and
Downs of a Repairable System, Dissertation TU Twente, pg. 23-29, Print
85
Maintenance Planning Optimisation: B737 NG
86
Appendix A
A. Assignment
Background With the introduction of the 737 NG, the Boeing Company has
ceased prescribing the C-check interval. This has been
replaced by a large number of ungrouped tasks that have got
to be planned into the maintenance program by the aircraft
operator.
87
Appendix B
B. Transavia Airlines
B.1. History
Transavia Limburg N.V. was founded in 1965. It received an airline licence in 1966,
with which it could operate charter flights from Beek (Maastricht) and Zestienhoven
(Rotterdam) airports. Shortly thereafter, an American company, the American Boreas
Corporation, bought all company shares. The companys name was later changed to
Transavia Holland so that it could also operate from Schiphol airport. Despite
various problems with the acquisition of licences, the airline transported 21.000
passengers in its first year of operation.
In 1968, the airline began operating charter flight to the United States, and it also
stated leasing out its aircraft and cabin/cockpit crew. By 1971, the Transavia
passenger count had reached 447.000, and it had 420 employees. In 1972, the Royal
Dutch Steamship Company (KNSM) took over 40% of the companys shares. In the
same year, the Caravell fleet was replaced with the Boeing 737-200, at the cost of fl.
20 million per aircraft. KNSM acquired all the company shares in 1977.
Transavia Holland changed its name to Transavia airlines in 1986, and in the same
year, it acquired its first B737-300. During the same year, the company transported its
millionth passenger. An introduction of a line service to London Gatwick led to the
introduction of Business Class seats for the fleet.
A merger of the KNSM and Ned Lloyd in 1988 resulted in the sale of 40% of the
shares to KLM. Transavia airlines also introduced a line service to holiday
destinations along the Mediterranean Coast.
1991 saw the acquisition of 80% of the shares by KLM. A year later, the airline
received a delivery of its first Boeing 757-200.
The airline has its head office at Amsterdam Airport Schiphol. Its main operation
base is also Schiphol Airport. Its other base is at Rotterdam Airport.
Staff
Currently, Transavia has an average of 1500 employees, serving on board, as ground
staff and as administrative staff.
89
Maintenance Planning Optimisation: B737 NG
B.4. Organisation
The following diagram illustrates the organisation of Transavia airlines as a whole.
90
Appendix B
President
Corporate Personnel
Communications (PR) & Organisation
Controllers
Marketing
Department
Quality & Fleet
Safety Assurance Management Financial Charter
Administration Sales
Commercial
Insurance
Realisation
Facility
Management Lease
The technical department is subdivided into three departments, namely Purchases and
Logistics, Aircraft Maintenance and Engineering.
Figure B.1 illustrates the organisational structure of the E&M
B - Aircraft Maintenance
The Aircraft Maintenance department has the following objectives:
1. To ensure that all operating aircraft are in serviceable condition as regards to
airworthiness, punctuality, sitting configuration and passenger/crew comfort
2. Support the Lease department (Lease in/out, sale) and the Fleet Management
(aircraft purchases) technically.
a. Line maintenance
- Arrival services (aircraft reception from cockpit crew)
- Aircraft inspection as per instructions
- Defect analysis
- Defect correction or deferring
- Aircraft transfer to cockpit crew
- Departure services
91
Maintenance Planning Optimisation: B737 NG
Head of
Maintenance and Engineering (A)
Maintenance
Training Engineering Technical
Purchases Supply shop Planning and
co-ordination team library
Support
Project Technical
purchaser
Support
Technical
purchaser
Maintenance Seat
team workshop
Analysts
b. Hangar maintenance
- Towing aircraft from the Gate to the Hangar
- Parking aircraft in the Hangar
- Opening of inspection access panels
- Performing line maintenance inspections and modifications
- Defect analysis
- Defect correction or deferring
- Closing of inspection access panels
- Moving aircraft from Hangar
- Towing aircraft to the Gate
c. Seat maintenance
- Seat inspection
- Seat defect correction
C Engineering
92
Appendix B
Objectives:
1. Setting technical standards and specifications for new or modified aircraft and
components so that aircraft maintenance may be performed in accordance with
Transavia and NLA maintenance execution requirements, while keeping cost as
low as possible
2. Supporting the aircraft maintenance department in faultfinding and rectification
of complex faults. Also issue repair instructions for specific tasks, and help in
investigating the course of the most common defects
3. Participate technically in the purchase of new aircraft, and in making
specifications on aircraft modification in order to fulfil the wishes of the Flight
and cabin services, marketing and lease department.
a. Maintenance programs
b. Power Plants
c. Structures/Systems
d. Interiors
e. Avionics
Objectives
Realisation of all commercial activities (purchases, outsourcing and sales) of goods
and services (including personnel) and all logistical services in order to support the
E&M in maintaining its JAROPS1 and JAR145 standard and low cost of ownership
strategies.
Activities
a. Project purchases: purchasing specified goods and services for maintenance
activities and modifications from specialised vendors. This also involves
preparing contracts for out-source work and pooling activities.
b. Purchases: Purchasing goods and materials for (ad hoc) maintenance activities
and toping up component stocks for normal activities. Purchases also deals with
the disposal (sale) of unnecessary materials.
c. Analysis: Determining optimum stock and spare-parts levels and re-order levels
and monitoring contracts and co-operation agreements.
d. Warranty management: Monitoring aircraft and component warranty as a tool of
minimising spending on maintenance.
e. Supplies shop: Mainly deals with the reception, storage and distribution of
materials/articles and technical consignments.
93
Maintenance Planning Optimisation: B737 NG
The Transavia operational year is typically divided into two periods: the summer
period (high season) and the winter period (low season).
The high season begins in May and runs through October. During this season, the
airline operates at its maximum capacity (utilises all aircraft) and, where necessary,
leases in aircraft. Consequently, minimum ground time (turn around time and
maintenance time) is desired in order to suit operational demands.
The low season begins in November and runs through April, the following year. This
season demands the availability of at least half of the fleet, hence offering more
ground time for maintenance activities. It is during this season that heavy
maintenance is carried out.
94
Appendix B
95
Appendix C
The B747 Maintenance Steering Group (MSG) created a new analytic approach to
maintenance using three control processes:
1. Hard Time limit: A maximum interval for performing maintenance tasks on a
part or unit. Such intervals apply to overhaul, but also to the total life of the
part or unit.
2. On-Condition: A repetitive inspections or tests to determine the condition of
units or systems, comprising servicing, inspecting, testing, calibrating and
replacement.
3. Conditioning Monitoring: Applies to items that have neither Hard Time limits
nor On Condition maintenance as their primary maintenance policies
Is there a
failure NO
Mode that has direct
Adverse effects on
Flight safety?
Is there an
Indication to the flight crew NO
YES When a function ceases
To perform upon
Demand?
YES
Is there an
adverse relationship between
Age and reliability?
Is there a
Is there a maintenance or shop
reduction in failure resistance NO Check that will assure a high
NO NO
Detectable by Probability of continued Function or
Maintenance? function on demand?
HT HT
OC CM OC
97
Maintenance Planning Optimisation: B737 NG
MSG-3 Revision 1
This was a refining of the original MSG-3 in order to make it user-friendlier. It led to
the elimination of the Operating Crew Monitoring strategy in the original MSG-3
MSG-3 Revision 2
This is a revision of the MSG-3 policy. It provides additional clarification and
guidelines on development of Corrosion Prevention Control Programs (CPCP).
98
Appendix C
As mentioned earlier, the logical flow diagram classifies maintenance tasks into five
categories, namely:
- Category 5: Evident Safety Effect
- Category 6: Evident Operational Effect
- Category 7: Evident Economic Effect
- Category 8: Hidden Safety Effect
- Category 9: Hidden Non-Safety Effect
YES NO
Evident Functional Failure Hidden Functional Failure
Does the functional failure or secondary Doe s the combination of a hidden functional failure
Damage resulting from the functional and one additional failure of a system related
Failure have a direct adverse effect on the or back-up function have and adverse effect on
Operating safety? Operational safety?
YES NO YES NO
YES NO
99
Maintenance Planning Optimisation: B737 NG
The MSG-3 principle is also utilised in identifying Structural Significant Items (SSI). It
begins with considering the whole aircraft structure.
The aircraft is sub-divided into zones and/or areas using industry standards and
definitions resulting in manageable sections for analysis. Thereafter, all structural
items are identified and classified as Structural Significant Items, or as Other Structure.
AIRCRAFT
STRUCTURE
DEFINE AIRCRAFT
ZONES OR AREAS
DEFINE AIRCRAFT
STRUCTURAL ITEMS
IS ITEM
NO
SIGNIFICANT STRUCTURAL
ITEM?
YES
CATEGORISE AND
LIST AS SSI
For all SSI, Accidental Damage (AD) and Environmental Deterioration (ED)
evaluations are performed, resulting in the MRB tasks. Fatigue evaluations usually
form part of the damage tolerance evaluation for certification, and will result in
supplementary inspections (Airworthiness Limitations), where the baseline MRB are
not adequate.
The Zonal maintenance program covers inspection requirements for Other Structure.
100
Appendix D
Objectives
Manufacturer
To provide the operator with packaging options for fleet maintenance
program
Increase the production efficiency of an organisation
Operator
Increase aircraft availability
Reduce production costs
Advantages:
Reduces spare requirements at out stations
101
Maintenance Planning Optimisation: B737 NG
Disadvantages
Manpower requirements can be adversely affected if airplane fleet planning cannot
support back-to-back scheduling of airplanes.
Component parts may be replaced several hundred hours before replacement is
necessary. However, the benefit quite often outweighs the disadvantages. Operator
scheduling constraints are typically the reason for early replacements.
Block check programs increase the risk of the airplane becoming a parts department
in support of in-service airplanes.
Advantages
Equalised packaging simplifies manpower-planning requirements. This is because,
moving tasks from one check package to another minimises peaks and valleys in
manpower requirements. However, when considering equalisation, it is important to
weigh all operational considerations.
102
Appendix D
Disadvantages
As with the block check, if taken to an extreme, disadvantages can offset the
efficiency gained in another area if all aspects of the program are not carefully
considered. Equalised work packages increase open/close access requirements. This
in turn increases the stress on fasteners, causing a consequent replacement due to
induced damage. Additionally, over time, total man-hours as compared to a block
check will be higher.
Other disadvantages include:
Increase in production planning and scheduling workload
Careful planning is required due to the limited ground time
There is limited time for the accomplishment of modifications
Short jobs are inherently inefficient
Limited time to identify and rectify non-routines
103
Appendix E
E.1. Cause
The MPD of the 737-800 does not make a distinction between A and C checks.
Instead, it just lists all maintenance tasks that need to be performed on the aircraft.
This allows the operator to set up his own maintenance program, depending on the
aircraft utilisation.
Transavia opted to define line maintenance task as all tasks with maintenance limits
of less than 540 Days, 3000 Cycles or 6000 Hours. All tasks with limits above these
were classified as base maintenance tasks. Such a definition was not reasonable,
owing to the fact that there were many tasks with higher intervals, which Transavia
was fully capable of performing.
The JAA Leaflet No. 6 on the JAR-145 Definition of Line Maintenance also supported
this view. Point 4 of this leaflet states: Taking into account the wide range of aircraft
used in commercial aircraft transport, e.g. light single engine, helicopter and large transport
airplanes, and furthermore considering their different maintenance programmes, it is not
appropriate to use hours, letter checks, or calendar time as a divider between line and base
maintenance.
Transavia would therefore prefer to perform some base maintenance tasks at its own
hangar, in order to reduce this de-escalation.
E.2. Aim
To carry out an evaluation on which tasks, classified under the current definition (see
E.1 above) as Base Maintenance tasks, can be performed by Transavia, a JAR-145
Line Maintenance organisation (JAR-145 Section 2, paragraphs 3.2.1-3.2.4).
Transavia Airlines is currently certified as a Line Maintenance Organisation, and
hence licensed to perform all Line Maintenance tasks. Line Maintenance referred to is
as defined in this document, and as approved by the National Aviation Authority
(Inspectie Verkeer and Waterstaat - Divisie Luchtvaart, IVW-DL, formally RLD), in
accordance to JAR-OPS 1.910(b) and 3.910(b).
E.3. Approach
All maintenance tasks Transavia is capable of performing will be evaluated by
considering the following aspects:
Staff
Whether the Transavia Certified staff (JAR-66, Category B1) has the knowledge and
expertise to perform such tasks, and to rectify defects where necessary.
Means
Whether Transavia has the necessary components, materials, tools, equipment and
documentation needed to perform such tasks, and to rectify defects where necessary.
Materials
Whether Transavia has the necessary facilities required to perform such tasks, and to
rectify defects where necessary.
105
Maintenance Planning Optimisation: B737 NG
Methodology
Whether Transavia, as an organisation, is structured in such a way that it can
perform such tasks and rectify defects where necessary.
E.3.1. Engineering
The engineering department evaluates whether maintenance for various systems can
be executed within the means of Transavia, or whether a JAR-145 base maintenance
organisation should be contracted
The following are restrictions on the evaluation on the execution of tasks, based on
MSG-3 rev.2 task categories
System engineers are responsible for determining the necessary means and materials,
and if necessary, can request intervention of the tooling commission.
106
Appendix E
The engineering department will use the following flow diagram (Figure E. 1) to
determine whether Transavia can perform a specific maintenance task. The results
will then be recorded in a compliance checklist.
MRI
NO
2. Does Transavia have the tools
and the equipment required
to perform this task? Are the YES/ NOT APPLICABLE
Certified staff from (1) above
capable to operate these
tools/equipment?
NO
NO
YES
Transavia Cannot Perform this task Transavia can perform this task
107
Appendix F
The reliability program is a close loop cycle, accomplished by applying the following
steps:
1. Identification of performance parameters that reflect airplane reliability.
2. Collection, analysing and reporting of data gathered from service experience and
reflecting airplane reliability.
3. Problems are investigated and identified.
4. Corrective actions are proposed and applied.
5. Corrective actions are monitored to ensure that maintenance cycle problems are
solved.
A Reliability Program Flow Chart (Figure F. 1 below) will include the following
steps:
1. Identification of performance parameters.
2. Collection of. Service data
3. Reporting and analysing of service data
4. Decisions are made if performance standards are met
5. Engineering investigates alerts and determines corrective actions
6. Reliability Control Board (RCB) approves corrective action.
7. Engineering issue engineering order (EO) to correct problem.
8. Maintenance accomplishes EO on airplane.
9. Cycle repeat it self.
The reliability program supports the long and short-term surveillance of fleet
reliability through a progressive schedule of activities, which include a daily
109
Maintenance Planning Optimisation: B737 NG
Data Collected
No
Incorporate
EO Alert
Engineering
Investigation
Maintenance
Corrective
RCB Approval Program
Action
Changes
Performance Parameters
A performance parameter is a quantitative measure of airplane reliability for which the
following can be established:
1. A trend value
2. An upper control limit
3. A lower control limit
110
Appendix F
Pilot Reports*1000
Fleet Technical Performance =
Airtime per evaluation interval
b. System level: performance parameters used are Pilots reports (PIREPS) per 100
landings and/or Mechanical delays/Cancellations per 100 Revenue
Departures, the ATA chapters considered when calculating the rate of these
two parameters are ATA 21 through 36, 38 and 49.
c. Power plant level: parameters used are In-flight shutdowns events (IFSD) per
1000 engine hours and/or unscheduled removal per 1000 engine hours.
Transavia airlines also works with an engine removal rate parameter, which
is:
No. of engines removed*1000
Engine Removal Rate =
Engine Hours
The ATA chapters considered when calculating the rate of these parameters
are ATA 71 through 80.
111
Maintenance Planning Optimisation: B737 NG
The ATA chapters considered when calculating the rate of these two
parameters are ATA chapters 21 through 36, 38, 49 for systems and 73, 75, 77,
79 and 80 for power plant.
The MTBUR is used more often because gathering failure information would
require shop reports, which usually take a long time to gather.
where X =
X
N
( X ) 2
X2 N
= SD =
N 1
K = multiplier
The mean value ( X ) is calculated by dividing the sum of the entire failure rate for
the period under consideration by the number of months in the period.
Alert values will exist whenever a three-month average rate of occurrence exceeds
the upper control limit.
112
Appendix G
The Standards for design and operation are issued as regulations. These can apply to
both Manufacturers and Operators. Approvals by the Regulatory authority showing
compliance with these standards are shown by means of issuing certificates.
The FAA issues its standards through the Federal Aviation Regulations (FARs). The
Code of Federal Regulation is divided into 50 titles that represent broad areas subject
to federal regulation. The Aviation Regulations are found in the Code of Federal
Regulations Title 14, Chapter I, Parts 1 through 199.
The major Parts from Chapter I of the Code that have an impact on Maintenance are:
Parts 25, 33, 34, and 36 contain the design standards that are used as the
basis of Certification.
Parts 39 and 121 are applicable to operating airplanes
Standards for Safety in Design are contained in Part 25 and mandatory changes to
the certified Design are covered by part 39. Part 25.1529 (Appendix H) contains the
requirement for Instructions for Continued Airworthiness. This means that
Maintenance Manuals must support a new design. This must include an approved
Maintenance Program.
The standards in the FARs that relate to how an operator must perform maintenance
are found in part 121 and Part 43. Part 121.365 requires a certificate holder to have an
adequate maintenance and inspection organization. Part 121.379 requires certificate
holders to use approved data for major repairs and alterations.
Part 43 requires that methods, techniques, and practices acceptable to the
administrator be used for inspection and repair of civil aircraft. FAA approved
services documents are the major source that fill this requirement. This FAR also
identifies what is required (by paper work sign offs) to return an aircraft back to
service also defines a Major and Minor Repair or Alteration.
Advisory Circulars are the FAA primary means of distributing information of a non-
regulatory nature to the public. They provide a description of non-regulatory
guidance, policy, and information. There is a Master Index that lists all effective ACs.
There are several Advisory Circulars that have an impact on Maintenance Programs.
These include:
1. AC 120-16C: Continuous Airworthiness Maintenance Programs
This AC gives airline the privilege and responsibility of aircraft
maintenance. It provides information and guidance on Continuous
Airworthiness Maintenance Programs. Five elements of an airworthiness
maintenance program include:
- Performance and approval of maintenance & alterations (routine
and non-routine maintenance)
113
Maintenance Planning Optimisation: B737 NG
114
Appendix G
G.1.3. Airworthiness
Two conditions must be met for an airplane to be considered airworthy:
The Joint Aviation Authorities (JAA) is an associated body of the European Civil
Aviation Conference (ECAC) representing the civil aviation regulatory authorities of
a number of European States who have agreed to co-operate in developing and
implementing common safety regulatory standards and procedures. This co-
operation is intended to provide high and consistent standards of safety and a "level
playing-field" for competition in Europe. Much emphasis is also placed on
harmonising the JAA regulations with those of the USA (FAA).
The JAA develops and implements Joint Aviation Requirements (JARs) in the fields
of:
- Aircraft design and manufacture,
- Aircraft operations and maintenance,
- Licensing of aviation personnel
115
Maintenance Planning Optimisation: B737 NG
116
Appendix H
Assumptions:
No maintenance is performed on days other than those agreed upon with CPV
Slot Time Hangar Time = 3hrs 45 min consists of:
Arrival last flight SPL: Crew/passengers/Catering out, cleaning 50 minutes
Towing 30 minutes
Hangar activities (miscellaneous) 40 minutes
117
Maintenance Planning Optimisation: B737 NG
118
Appendix I
a. The letters D, H and C stand for Days, Hours and Cycles, respectively
b. Inspections/checks marked with a * have a dual limit: The checks are
performed when either of the limits is reached, i.e. whichever limit comes
first
c. The 38 preceding the letters D, H, and H indicates that the check is
applicable for the B737-800, abbreviated as B738. The checks apply to all
next generation aircraft (- 700 and 800) in the Transavia fleet
119
Maintenance Planning Optimisation: B737 NG
38C9000B* 900D 8 4
8C24000B* 2880D 14 13.63
38C25000B 2 4
8C36000B* 2880D 3 5.5
38H8000B 7 4.35
38H10000B 1 0.5
38H12000B 2 5
38H15000B 2 2
38H20000B 4 2.15
38H22400B 1 1.2
38H25000B 4 1.6
38H30000B 2 1.3
The Maintenance task packaging results from an analysis of all effective OMP tasks,
based on the procedure illustrated above.
ALL Tasks
Line?
Base?
Line Maintenance Base Maintenance
Tasks Tasks
Transavia L/B
Classification Criteria
Set-up properties
Transavia Interval
(CT/FH/FC)
Maintenance Maintenance
Task Packages Checks
Figure I. 1 Maintenance Task Packaging Process
120
Appendix I
The following is a summary of all Maintenance Task packages. A complete list of the
contents of each task package follows in Table I. 4.
It is worth mentioning that most tasks in a task package are from the same ATA 100
chapter (level 1), even though this does not always have to be the case. The reason for
this may be attributed to the relationship between ATA 100 and aircraft Zones
Tasks not classified include all engine-specific tasks that require to be performed
separately through the application of the staggering principle. This is done in order
to ensure power unit reliability, as recommended by Boeing and the engine
manufacture, CFMI.
Task Package TAV Check No. of MRIs MPD MHrs MSG-3 Class. Skill MSG-3 Cat ATA Chapter
C300 38C300 2 2 GV M 8 25-
C1000 38C1000 4 1.2 LU M 5 27-
C2000A 38C2000 1 0.3 RS M 6 21-
C2000B 38C2000 6 0.34 GV M 0 70-
C3000 38C3000 14 5.5 M 32-
C3500 38C3500 2 0.2 GV M 9 25-
C4000A 38C4000 2 0.6 LU M 6 27-
C4000B 38C4000 4 0.3 2GV, 2DT M 0,6 70,74
C4000A* 38C4000* 7 1.85 5DT, 2GV M 20,32
C4000B* 38C4000* 10 1.08 GV M 0 52
C4000C* 38C4000* 12 2.49 GV M 0 53
C4000D* 38C4000* 9 4.16 GV M 0 53
C4000E* 38C4000* 3 0.59 GV M 0 53
C4000F* 38C4000* 4 0.54 GV M 0 53,54
C4000G* 38C4000* 23 3.96 GV M 0 55
C4000H* 38C4000* 54 11.21 GV M 0 57
C5000A 38C5000 3 1 SV M 6 27
C5000B 38C5000 2 0.2 VC M 9 72
C9000* 38C9000* 6 1.18 GV M 0 54
C10000 38C10000 1 0.1 OP A 0 24
C12000 38C12000 2 0.8 DT M 6 27
C12500 38C12500 1 0.25 OP M 9 32
C15000 38C15000 1 0 OP M 8 32
C25000 38C25000 5 3.9 3SV, 2OP M 8 27
121
Maintenance Planning Optimisation: B737 NG
Task Package TAV Check No. of MRIs MPD MHrs MSG-3 Class. Skill MSG-3 Cat ATA Chapter
D720F* 38D720* 5 20 DT M 0 53
D1080A 38D1080 1 0.5 FC M 9 26
D1080B 38D1080 5 3 2OP, 2DT M 8 28
D1080* 38D1080* 2 0.6 DT M 8 27
D1440A 38D1440 1 1.5 FC A 9 23
D1440B 38D1440 1 2 VC M 9 35
D1800A 38D1800 1 4 RS M 9 38
D1800B 38D1801 2 0.2 OP M 9 71
D1800A* 38D1800* 8 4 DT M 6 20
D1800B* 38D1800* 10 15.4 GV M 0 54
D2880A 38D2880 2 1.1 GV M 0 53
D2880B 38D2880 2 2 GV M 0 55
D1440 38D2880 2 1 GV M 0 55
D2880C 38D2880 1 0.5 GV M 0 57
D2880* 38D2880* 23 17.9 M 1 52
122
Appendix I
Task Package TAV Check No. of MRIs MPD MHrs MSG-3 Class. Skill MSG-3 Cat ATA Chapter
H10000H 38H10000 1 0.4 DS M 8 29
H10000I 38H10000 1 0.2 DS M 8 29
H10000J 38H10000 1 0.2 DT A 6 34
H10000K 38H10000 4 0.85 3GV, 1OP M 49
H12000 38H12000 1 2 FC M 8 28
H12500 38H12500 2 1.9 FC M 8 27
H15000A 38H15000 4 3 2DT, 2GV A 8 20
H15000B 38H15000 1 1 GV A 8 20
H15000C 38H15000 2 0.4 DT A 8 20
H20000 38H20000 1 1.5 DT A 8 20
H25000A 38H25000 5 2.45 M 27
H25000B 38H25000 7 1.6 6FC, 1DS M 29
H25000C 38H25000 6 2.4 FC M 8 29
H25000D 38H25000 2 0.4 FC, OP M 32
H30000 38H30000 1 0.5 GV A 8 20
H4000A 38H4000 2 0.2 OP A 9 21
123
Maintenance Planning Optimisation: B737 NG
1000 HRS MI 1.5 IN M 0 MI-28-20-8006
C4000A* 4000A* CYC 20-140-00 38C4000* 0.15 D480* 480* DYS 27-222-01 38D480* 0.3
C4000A* 4000A* CYC 20-150-01 38C4000* 0.25 D480* 480* DYS 27-222-02 38D480* 0.3
C4000A* 4000A* CYC 20-150-02 38C4000* 0.25
C4000A* 4000A* CYC 20-160-01 38C4000* 0.6 D540A 540A DYS 25-380-00 38D540 0.05
C4000A* 4000A* CYC 20-160-02 38C4000* 0.6 D540B 540B DYS 25-400-00 38D540 0.05
C4000A* 4000A* CYC 32-806-01 38C4000* 0 D540C 540C DYS 32-200-00 38D540 0.6
C4000A* 4000A* CYC 32-810-02 38C4000* 0 D540C 540C DYS 32-215-01 38D540 0.9
C4000B* 4000B* CYC 52-804-02 38C4000* 0.13 D540C 540C DYS 32-215-02 38D540 0.9
C4000B* 4000B* CYC 52-808-02 38C4000* 0.13
C4000B* 4000B* CYC 52-812-01 38C4000* 0.13 D720A 720A DYS 23-050-00 38D720 0.2
C4000B* 4000B* CYC 52-814-01 38C4000* 0.03 D720B 720B DYS 25-150-00 38D720 0.1
C4000B* 4000B* CYC 52-818-01 38C4000* 0.03 D720C 720C DYS 25-230-00-01 38D720 0.2
C4000B* 4000B* CYC 52-824-01 38C4000* 0.13 D720C 720C DYS 25-230-00-02 38D720 0.2
C4000B* 4000B* CYC 52-828-02 38C4000* 0.17 D720C 720C DYS 25-230-00-03 38D720 0.2
C4000B* 4000B* CYC 52-830-02 38C4000* 0.08 D720C 720C DYS 25-230-00-04 38D720 0.2
C4000B* 4000B* CYC 52-834-02 38C4000* 0.08 D720D 720D DYS 25-330-00 38D720 0.05
C4000B* 4000B* CYC 52-840-02 38C4000* 0.17 D720E 720E DYS 25-370-00 38D720 0.05
C4000C* 4000C* CYC 53-802-00 38C4000* 0.07 D720F 720F DYS 34-010-00 38D720 2.4
C4000C* 4000C* CYC 53-804-00 38C4000* 0.07 D720G 720G DYS 49-010-00 38D720 0.1
C4000C* 4000C* CYC 53-806-00 38C4000* 0.1 D720H 720H DYS 52-011-01 38D720 0.6
C4000C* 4000C* CYC 53-808-00 38C4000* 0.13 D720H 720H DYS 52-011-02 38D720 0.6
C4000C* 4000C* CYC 53-810-00 38C4000* 0.08 D720H 720H DYS 52-012-01 38D720 0.6
124
Appendix I
C4000C* 4000C* CYC 53-822-00 38C4000* 0.13 D720H 720H DYS 52-012-02 38D720 0.6
C4000C* 4000C* CYC 53-828-00 38C4000* 0.33 D720H 720H DYS 52-120-00 38D720 0.2
C4000C* 4000C* CYC 53-838-00 38C4000* 0.25 D720H 720H DYS 52-220-00 38D720 0.4
C4000C* 4000C* CYC 53-842-00 38C4000* 0.5 D720H 720H DYS 52-230-00 38D720 0.15
C4000C* 4000C* CYC 53-846-00 38C4000* 0.33 D720H 720H DYS 52-240-00 38D720 0.4
C4000C* 4000C* CYC 53-848-00 38C4000* 0.33 D720H 720H DYS 52-250-00 38D720 0.4
C4000C* 4000C* CYC 53-852-00 38C4000* 0.17 D720H 720H DYS 52-260-00 38D720 0.2
C4000D* 4000D* CYC 53-858-00 38C4000* 0.5
C4000D* 4000D* CYC 53-860-00 38C4000* 0.33 D720A* 720A* DYS 27-138-01 38D720* 0.2
C4000D* 4000D* CYC 53-864-00 38C4000* 0.5 D720A* 720A* DYS 27-138-02 38D720* 0.2
C4000D* 4000D* CYC 53-866-00 38C4000* 0.33 D720B* 720B* DYS 52-510-00 38D720* 1
C4000D* 4000D* CYC 53-874-00 38C4000* 0.5 D720C* 720C* DYS 52-540-00 38D720* 0.3
C4000D* 4000D* CYC 53-876-00 38C4000* 0.67 D720C* 720C* DYS 52-610-00 38D720* 1
C4000D* 4000D* CYC 53-882-00 38C4000* 0.5 D720C* 720C* DYS 52-610-00 38D720* 1
C4000D* 4000D* CYC 53-884-00 38C4000* 0.5 D720C* 720C* DYS 52-610-00 38D720* 1
C4000D* 4000D* CYC 53-896-00 38C4000* 0.33 D720C* 720C* DYS 52-610-00 38D720* 1
C4000E* 4000E* CYC 53-898-00 38C4000* 0.17 D720C* 720C* DYS 52-670-00 38D720* 1
C4000E* 4000E* CYC 53-900-00 38C4000* 0.25 D720C* 720C* DYS 52-670-00 38D720* 1
C4000E* 4000E* CYC 53-902-00 38C4000* 0.17 D720C* 720C* DYS 52-730-00 38D720* 0.8
C4000F* 4000F* CYC 53-904-01 38C4000* 0.17 D720C* 720C* DYS 52-730-00 38D720* 0.8
C4000F* 4000F* CYC 53-906-02 38C4000* 0.17 D720D* 720D* DYS 53-010-00 38D720* 1.2
C4000F* 4000F* CYC 54-802-01 38C4000* 0.1 D720D* 720D* DYS 53-010-00 38D720* 1.2
C4000F* 4000F* CYC 54-810-02 38C4000* 0.1 D720D* 720D* DYS 53-010-00 38D720* 1.2
C4000G* 4000G* CYC 55-802-00 38C4000* 0.07 D720D* 720D* DYS 53-010-00 38D720* 1.2
C4000G* 4000G* CYC 55-804-00 38C4000* 0.1 D720E* 720E* DYS 53-030-00 38D720* 0.4
C4000G* 4000G* CYC 55-806-00 38C4000* 0.07 D720E* 720E* DYS 53-030-00 38D720* 0.4
C4000G* 4000G* CYC 55-810-00 38C4000* 1 D720E* 720E* DYS 53-030-00 38D720* 0.4
C4000G* 4000G* CYC 55-812-00 38C4000* 0.07 D720F 720F* DYS 53-310-00 38D720* 4
C4000G* 4000G* CYC 55-814-00 38C4000* 0.5 D720F 720F* DYS 53-310-00 38D720* 4
C4000G* 4000G* CYC 55-816-00 38C4000* 0.07 D720F 720F* DYS 53-310-00 38D720* 4
C4000G* 4000G* CYC 55-818-00 38C4000* 0.07 D720F 720F* DYS 53-310-00 38D720* 4
C4000G* 4000G* CYC 55-820-00 38C4000* 0.03 D720F 720F* DYS 53-310-00 38D720* 4
C4000G* 4000G* CYC 55-822-01 38C4000* 0.07
C4000G* 4000G* CYC 55-824-01 38C4000* 0.07 D1080A 1080A DYS 26-290-00 38D1080 0.5
C4000G* 4000G* CYC 55-826-01 38C4000* 0.13 D1080B 1080B DYS 28-020-01 38D1080 1
C4000G* 4000G* CYC 55-828-01 38C4000* 0.5 D1080B 1080B DYS 28-020-02 38D1080 1
C4000G* 4000G* CYC 55-830-01 38C4000* 0.07 D1080B 1080B DYS 28-030-01 38D1080 0.5
C4000G* 4000G* CYC 55-832-01 38C4000* 0.08 D1080B 1080B DYS 28-030-02 38D1080 0.5
C4000G* 4000G* CYC 55-834-01 38C4000* 0.07
C4000G* 4000G* CYC 55-836-02 38C4000* 0.07 D1080* 1080* DYS 27-034-01 38D1080* 0.3
C4000G* 4000G* CYC 55-838-02 38C4000* 0.07 D1080* 1080* DYS 27-034-02 38D1080* 0.3
C4000G* 4000G* CYC 55-840-02 38C4000* 0.13
C4000G* 4000G* CYC 55-842-02 38C4000* 0.5 D1440A 1440A DYS 23-030-00 38D1440 1.5
C4000G* 4000G* CYC 55-844-02 38C4000* 0.07 D1440B 1440B DYS 35-060-00 38D1440 2
C4000G* 4000G* CYC 55-846-02 38C4000* 0.08 D1440C 1440 DYS 55-120-01 38D2880 0.5
C4000G* 4000G* CYC 55-848-02 38C4000* 0.07 D1440C 1440 DYS 55-120-02 38D2880 0.5
C4000H* 4000H* CYC 57-804-01 38C4000* 0.08
C4000H* 4000H* CYC 57-806-01 38C4000* 0.17 D1800A 1800A DYS 38-040-00 38D1800 4
C4000H* 4000H* CYC 57-808-01 38C4000* 0.08 D1800B 1800B DYS 71-040-01 38D1800 0.1
C4000H* 4000H* CYC 57-810-01 38C4000* 0.08 D1800B 1800B DYS 71-040-02 38D1800 0.1
C4000H* 4000H* CYC 57-812-01 38C4000* 0.07
C4000H* 4000H* CYC 57-814-01 38C4000* 1 D1800A* 1800A* 20-180-00 38D1800* 0.5
C4000H* 4000H* CYC 57-816-01 38C4000* 0.17 D1800A* 1800A* 20-190-00 38D1800* 0.9
C4000H* 4000H* CYC 57-818-01 38C4000* 0.03 D1800A* 1800A* 20-200-00 38D1800* 0.45
C4000H* 4000H* CYC 57-824-01 38C4000* 1.25 D1800A* 1800A* 20-210-00 38D1800* 0.3
C4000H* 4000H* CYC 57-828-01 38C4000* 0.5 D1800A* 1800A* 20-220-00 38D1800* 0.65
C4000H* 4000H* CYC 57-832-01 38C4000* 0.05 D1800A* 1800A* 20-230-00 38D1800* 0.6
C4000H* 4000H* CYC 57-834-01 38C4000* 0.08 D1800A* 1800A* 20-240-00 38D1800* 0.45
C4000H* 4000H* CYC 57-836-01 38C4000* 0.05 D1800A* 1800A* 20-250-00 38D1800* 0.15
C4000H* 4000H* CYC 57-838-01 38C4000* 0.1 D1800B* 1800B* 54-010-01 38D1800* 0.2
C4000H* 4000H* CYC 57-840-01 38C4000* 0.05 D1800B* 1800B* 54-010-02 38D1800* 0.2
C4000H* 4000H* CYC 57-842-01 38C4000* 0.1 D1800B* 1800B* 54-015-01 38D1800* 0.2
C4000H* 4000H* CYC 57-844-01 38C4000* 0.05 D1800B* 1800B* 54-015-02 38D1800* 0.2
C4000H* 4000H* CYC 57-846-01 38C4000* 0.1 D1800B* 1800B* 54-030-01 38D1800* 0.3
C4000H* 4000H* CYC 57-848-01 38C4000* 0.17 D1800B* 1800B* 54-030-02 38D1800* 0.3
C4000H* 4000H* CYC 57-852-01 38C4000* 0.08 D1800B* 1800B* 54-070-01 38D1800* 3
C4000H* 4000H* CYC 57-854-01 38C4000* 0.08 D1800B* 1800B* 54-070-02 38D1800* 3
C4000H* 4000H* CYC 57-858-01 38C4000* 0.17 D1800B* 1800B* 54-080-01 38D1800* 4
C4000H* 4000H* CYC 57-860-01 38C4000* 0.25 D1800B* 1800B* 54-080-02 38D1800* 4
C4000H* 4000H* CYC 57-862-01 38C4000* 0.25
125
Maintenance Planning Optimisation: B737 NG
C4000H* 4000H* CYC 57-864-01 38C4000* 0.05 D2880A 2880A DYS 53-020-00 38D2880 0.1
C4000H* 4000H* CYC 57-868-01 38C4000* 0.33 D2880A 2880A DYS 53-050-00 38D2880 1
C4000H* 4000H* CYC 57-870-01 38C4000* 0.08 D2880B 2880B DYS 55-080-01 38D2880 1
C4000H* 4000H* CYC 57-874-02 38C4000* 0.08 D2880B 2880B DYS 55-080-02 38D2880 1
C4000H* 4000H* CYC 57-876-02 38C4000* 0.17 D2880C 2880C DYS 57-030-00 38D2880 0.5
C4000H* 4000H* CYC 57-878-02 38C4000* 0.08
C4000H* 4000H* CYC 57-880-02 38C4000* 0.08 D2880* 2880* DYS 52-530-00 38D2880* 1
C4000H* 4000H* CYC 57-882-02 38C4000* 0.07 D2880* 2880* DYS 52-550-00 38D2880* 0.5
C4000H* 4000H* CYC 57-884-02 38C4000* 1 D2880* 2880* DYS 52-570-00 38D2880* 0.3
C4000H* 4000H* CYC 57-886-02 38C4000* 0.17 D2880* 2880* DYS 52-620-00 38D2880* 1
C4000H* 4000H* CYC 57-888-02 38C4000* 0.3 D2880* 2880* DYS 52-620-00 38D2880* 1
C4000H* 4000H* CYC 57-894-02 38C4000* 1.25 D2880* 2880* DYS 52-620-00 38D2880* 1
C4000H* 4000H* CYC 57-898-02 38C4000* 0.5 D2880* 2880* DYS 52-620-00 38D2880* 1
C4000H* 4000H* CYC 57-902-02 38C4000* 0.05 D2880* 2880* DYS 52-650-00 38D2880* 1
C4000H* 4000H* CYC 57-904-02 38C4000* 0.08 D2880* 2880* DYS 52-650-00 38D2880* 1
C4000H* 4000H* CYC 57-906-02 38C4000* 0.05 D2880* 2880* DYS 52-650-00 38D2880* 1
C4000H* 4000H* CYC 57-908-02 38C4000* 0.1 D2880* 2880* DYS 52-650-00 38D2880* 1
C4000H* 4000H* CYC 57-910-02 38C4000* 0.05 D2880* 2880* DYS 52-680-00 38D2880* 1
C4000H* 4000H* CYC 57-912-02 38C4000* 0.1 D2880* 2880* DYS 52-680-00 38D2880* 1
C4000H* 4000H* CYC 57-914-02 38C4000* 0.05 D2880* 2880* DYS 52-710-00 38D2880* 1
C4000H* 4000H* CYC 57-916-02 38C4000* 0.1 D2880* 2880* DYS 52-710-00 38D2880* 1
C4000H* 4000H* CYC 57-918-02 38C4000* 0.17 D2880* 2880* DYS 52-740-00 38D2880* 1.2
C4000H* 4000H* CYC 57-922-02 38C4000* 0.08 D2880* 2880* DYS 52-740-00 38D2880* 1.2
C4000H* 4000H* CYC 57-924-02 38C4000* 0.08 D2880* 2880* DYS 52-760-00 38D2880* 0.6
C4000H* 4000H* CYC 57-928-02 38C4000* 0.17 D2880* 2880* DYS 52-760-00 38D2880* 0.6
C4000H* 4000H* CYC 57-930-02 38C4000* 0.25 D2880* 2880* DYS 52-816-01 38D2880* 0.08
C4000H* 4000H* CYC 57-932-02 38C4000* 0.25 D2880* 2880* DYS 52-820-01 38D2880* 0.08
C4000H* 4000H* CYC 57-934-02 38C4000* 0.05 D2880* 2880* DYS 52-832-02 38D2880* 0.17
C4000H* 4000H* CYC 57-938-02 38C4000* 0.33 D2880* 2880* DYS 52-836-02 38D2880* 0.17
C4000H* 4000H* CYC 57-940-02 38C4000* 0.08
C5000A 5000A CYC 27-132-00 38C5000 0.3 C10000 10000 CYC 24-110-00 38C10000 0.1
C5000A 5000A CYC 27-140-01 38C5000 0.35
C5000A 5000A CYC 27-140-02 38C5000 0.35 C12000 12000 CYC 27-172-01 38C12000 0.4
C5000B 5000B CYC 72-100-01 38C5000 0.1 C12000 12000 CYC 27-172-02 38C12000 0.4
C5000B 5000B CYC 72-100-02 38C5000 0.1 C12500 12500 CYC 32-310-00 38C12500 0.25
C15000 15000 CYC 32-285-00 38C15000 0
C9000* 9000* CYC 54-800-01 38C9000* 0.25
C9000* 9000* CYC 54-804-01 38C9000* 0.17 C25000 25000 CYC 27-134-00 38C25000 0.5
C9000* 9000* CYC 54-806-01 38C9000* 0.17 C25000 25000 CYC 27-142-01 38C25000 1.5
C9000* 9000* CYC 54-808-02 38C9000* 0.25 C25000 25000 CYC 27-142-02 38C25000 1.5
C9000* 9000* CYC 54-812-02 38C9000* 0.17 C25000 25000 CYC 27-168-01 38C25000 0.2
C9000* 9000* CYC 54-814-02 38C9000* 0.17 C25000 25000 CYC 27-168-02 38C25000 0.2
H1000A 1000A HRS 25-125-00 38H1000 0.2 H4000A 4000A HRS 21-020-00 38H4000 0.1
H1000B 1000B HRS 26-320-00 38H1000 0.1 H4000A 4000A HRS 21-030-00 38H4000 0.1
H1000C 1000C HRS 28-060-01 38H1000 0.25 H4000B 4000B HRS 25-010-01 38H4000 0.5
H1000C 1000C HRS 28-060-02 38H1000 0.25 H4000B 4000B HRS 25-010-02 38H4000 0.5
H1000C 1000C HRS 28-060-03 38H1000 0.5 H4000B 4000B HRS 25-050-00 38H4000 0.3
H1000D 1000D HRS 32-300-00 38H1000 0.05 H4000B 4000B HRS 25-080-00 38H4000 0.3
H1000E 1000E HRS 72-020-01 38H1000 0.5 H4000B 4000B HRS 25-090-00 38H4000 0.3
H1000E 1000E HRS 72-020-02 38H1000 0.5 H4000B 4000B HRS 25-105-00 38H4000 0.2
H1000F 1000F HRS 78-120-01 38H1000 0.01 H4000C 4000C HRS 27-011-00 38H4000 0.5
H1000F 1000F HRS 78-120-02 38H1000 0.01 H4000D 4000D HRS 27-013-01 38H4000 0.6
H1000G 1000G HRS 79-020-01 38H1000 0.05 H4000D 4000D HRS 27-013-02 38H4000 0.6
H1000G 1000G HRS 79-020-02 38H1000 0.05 H4000D 4000D HRS 27-026-01 38H4000 0.5
H1250 1250 HRS 27-216-00 38H1000 0.7 H4000D 4000D HRS 27-026-02 38H4000 0.4
H1250 1250 HRS 27-224-00 38H1000 0.05 H4000D 4000D HRS 27-136-01 38H4000 0.4
H4000D 4000D HRS 27-136-02 38H4000 0.4
H1600A 1600A HRS 24-010-01 38H1600 1 H4000D 4000D HRS 27-182-00 38H4000 0.1
H1600A 1600A HRS 24-010-02 38H1600 1 H4000D 4000D HRS 27-182-01 38H4000 0.4
H1600A 1600A HRS 24-040-01 38H1600 0.5 H4000D 4000D HRS 27-182-02 38H4000 0.4
H1600A 1600A HRS 24-040-02 38H1600 0.5 H4000D 4000D HRS 27-215-01 38H4000 0.4
H1600A 1600A HRS 24-050-01 38H1600 0.2 H4000D 4000D HRS 27-215-02 38H4000 0.4
H1600A 1600A HRS 24-050-02 38H1600 0.2 H4000D 4000D HRS 28-010-00 38H4000 0.5
H1600B 1600B HRS 26-050-00 38H1600 0.05 H4000E 4000E HRS 29-060-00 38H4000 0.05
H1600C 1600C HRS 27-102-00 38H1600 0.4 H4000E 4000E HRS 29-070-00 38H4000 0.2
H4000E 4000E HRS 29-080-00 38H4000 0.7
H2000A 2000A HRS 24-090-00 38H2000 0.2 H4000E 4000E HRS 35-010-00 38H4000 0.2
126
Appendix I
H2000A 2000A HRS 24-120-00 38H2000 0.2 H4000F 4000F HRS 49-132-00 38H4000 0.2
H2000A 2500A HRS 27-154-00 38H2000 0.2 H4000G 4000G HRS 52-051-01 38H4000 0.1
H2000A 3500 HRS 25-030-00 38H2000 0.2 H4000G 4000G HRS 52-051-02 38H4000 0.1
H2000A 3000A HRS 26-300-00 38H2000 0.2 H4000G 4000G HRS 52-052-01 38H4000 0.1
H2000B 3000B HRS 27-218-00 38H2000 0.3 H4000G 4000G HRS 52-052-02 38H4000 0.1
H2000B 2000B HRS 29-050-01 38H2000 0.3 H4000G 4000G HRS 52-101-00 38H4000 0.1
H2000B 2000B HRS 29-050-02 38H2000 0.3 H4000G 4000G HRS 52-102-00 38H4000 0.1
H2000C 2000C HRS 35-050-00 38H2000 0.06 H4000G 4000G HRS 52-130-00 38H4000 0.1
H2000C 2500B HRS 38-030-00 38H2000 0.3 H4000G 4000G HRS 52-200-00 38H4000 0.4
H2000D 2000D HRS 49-190-00 38H2000 0.1 H4000G 4000G HRS 52-210-00 38H4000 0.2
H2000D 2000D HRS 49-200-00 38H2000 0.05 H4000H 4000H HRS 71-020-01 38H4000 0.1
H2000D 3200 HRS 49-090-00 38H2000 0.1 H4000H 4000H HRS 71-020-02 38H4000 0.1
H2000D 3200 HRS 49-110-00 38H2000 0.1
H2000D 3200 HRS 49-140-00 38H2000 0.2
H2000E 2000E HRS 79-030-02 38H2000 0.1
H2000E 3000C HRS 78-090-01 38H2000 0.1
H2000E 3000C HRS 78-090-02 38H2000 0.1
H2000F 2000F HRS 79-030-01 38H2000 0.1
H6000A 6000A HRS 26-250-00 38H6000 0.05 H8000A 8000A HRS 29-025-01 38H8000 0.4
127
Maintenance Planning Optimisation: B737 NG
H6000B 6000B HRS 26-090-00 38H6000 0.6 H8000A 8000A HRS 29-025-02 38H8000 0.4
H6000B 6000B HRS 26-150-00 38H6000 0.6 H8000A 8000A HRS 32-240-00 38H8000 0.4
H6000B 6000B HRS 26-170-00 38H6000 0.6 H8000B 8000B HRS 35-090-00 38H8000 0.5
H6000B 6000B HRS 28-080-00 38H6000 0.5 H8000B 8000B HRS 35-100-00 38H8000 0.1
H6000B 6000B HRS 78-100-01 38H6000 0.2 H8000C 8000C HRS 49-102-00 38H8000 0.2
H6000B 6000B HRS 78-100-02 38H6000 0.2 H8000C 8000C HRS 49-120-00 38H8000 0.3
H6000C 6000C HRS 26-190-00 38H6000 0.05 H8000D 8000D HRS 52-111-00 38H8000 0.3
H8000D 8000D HRS 52-112-00 38H8000 0.3
H6500A 6500A HRS 38-010-00 38H6500 0.2 H8000D 8000D HRS 52-140-00 38H8000 0.1
H6500A 6500A HRS 38-050-00 38H6500 0.1
H6500B 6500B HRS 38-060-00 38H6500 0.3
H10000A 10000A HRS 20-030-01 38H10000 1 H20000 20000 HRS 20-050-00 38H20000 1.5
H10000A 10000A HRS 20-030-02 38H10000 1 H25000A 25000A 27-084-00 38H25000 0.2
H10000B 10000B HRS 21-070-00 38H10000 0.1 H25000A 25000A 27-104-00 38H25000 0.05
H10000C 10000C HRS 23-090-00 38H10000 0.3 H25000A 25000A 27-166-00 38H25000 0.2
H10000D 10000D HRS 27-156-00 38H10000 0.1 H25000A 25000A 27-192-01 38H25000 1
H10000E 10000E HRS 27-194-00 38H10000 0.5 H25000A 25000A 27-192-02 38H25000 1
H10000F 10000F HRS 28-040-00 38H10000 1 H25000B 25000B 29-110-00 38H25000 0.4
H10000G 10000G HRS 29-210-00 38H10000 0.4 H25000B 25000B 29-120-00 38H25000 0.2
H10000H 10000H HRS 29-220-00 38H10000 0.4 H25000B 25000B 29-130-00 38H25000 0.2
H10000I 10000I HRS 29-270-00 38H10000 0.2 H25000B 25000B 29-140-00 38H25000 0.2
H10000J 10000J HRS 34-100-00 38H10000 0.2 H25000B 25000B 29-150-00 38H25000 0.2
H10000K 10000K HRS 49-162-00 38H10000 0.1 H25000B 25000B 29-160-00 38H25000 0.2
H10000K 10000K HRS 49-180-00 38H10000 0.5 H25000B 25000B 29-180-00 38H25000 0.2
H10000K 10000K HRS 49-212-00 38H10000 0.05 H25000C 25000C 29-170-00 38H25000 0.2
H10000K 10000K HRS 49-220-00 38H10000 0.2 H25000C 25000C 29-260-00 38H25000 0.4
H25000C 25000C 29-280-00 38H25000 0.5
H12000 12000 HRS 28-070-00 38H12000 2 H25000C 25000C 29-290-00 38H25000 0.4
H25000C 25000C 29-300-00 38H25000 0.2
H12500 12500 HRS 27-082-00 38H12500 0.2 H25000C 25000C 29-310-00 38H25000 0.7
H12500 12500 HRS 27-092-00 38H12500 1.7 H25000D 25000D 32-230-00 38H25000 0.3
H25000D 25000D 32-420-00 38H25000 0.1
H15000A 15000A HRS 20-010-01 38H15000 0.5
H15000A 15000A HRS 20-010-02 38H15000 0.5 H30000 30000 HRS 20-020-00 38H30000 0.5
H15000A 15000A HRS 20-080-01 38H15000 1
H15000A 15000A HRS 20-080-02 38H15000 1
H15000B 15000B HRS 20-100-00 38H15000 1
H15000C 15000C HRS 20-120-01 38H15000 0.2
H15000C 15000C HRS 20-120-02 38H15000 0.2
128
Appendix I
Unclassified: Inconsistent Task Interva
H450 500 HRS 21-140-01 38H450 0.05
H450 500 HRS 21-140-02 38H450 0.05
H450 1000 HRS 21-150-00 38H450 0.1
H450 400 HRS 24-020-01 38H450 0.1
H450 400 HRS 24-020-02 38H450 0.1
H450 400 HRS 24-030-01 38H450 0.1
H450 400 HRS 24-030-02 38H450 0.1
H450 550 HRS 25-020-00 38H450 0.6
H450 750 HRS 27-161-00 38H450 0.1
H450 500 HRS 29-030-01 38H450 0.4
H450 500 HRS 29-030-02 38H450 0.4
H450 500 HRS 29-090-00 38H450 0.1
H450 800 HRS 71-010-01 38H450 0.1
H450 800 HRS 71-010-02 38H450 0.1
H450 500 HRS 72-060-01 38H450 0.05
H450 500 HRS 72-060-02 38H450 0.05
H450 500 HRS 72-260-01 38H450 0.05
H450 500 HRS 72-260-02 38H450 0.05
H450 500 HRS 72-320-01 38H450 0.15
H450 500 HRS 72-320-02 38H450 0.15
129
Appendix J
J.1. Necessity
System state: State variables that describe the system state at a particular time
Event List: A list containing the next time when each type of event will occur
Statistical Counters: Variables used for storing statistical information about system
performance
Event Routine: A subprogram that updates the system state when a particular type of
event occurs (there is one event routine for each event type).
Library routines: a set of subprograms used to generate random observations from the
probability distributions determined as part of the simulation model
Main program: This is the subprogram that invokes the routines to determine the next
event and update the system state appropriately.
[Hilston, 2001]
Table J. 1 below tabulates the specific annual utilisation of the B737 NG Fleet, over
the period 1999-2000. It only includes that part of the fleet that was present for the
whole calendar year, i.e. from January till December.
131
Maintenance Planning Optimisation: B737 NG
The years 1998 (introduction year) and 2001 (data collection year) are not
considered.
C y cle s ZA 114 109 109 124 136 133 149 143 122 134 131 67 1471
2000 ZB 88 90 86 133 138 142 143 135 139 105 69 63 1331
ZC 89 78 126 101 135 140 144 149 140 132 131 107 1472
ZD 77 79 72 106 139 131 153 146 142 129 69 70 1313
ZE 76 82 81 103 140 135 146 148 145 128 71 82 1337
ZF 102 99 113 113 132 134 142 143 132 136 82 89 1417
1999 ZA 79 76 104 111 136 122 145 144 147 143 100 75 1382
ZB 87 71 94 103 137 128 150 141 149 134 97 106 1397
ZC 81 79 99 119 141 130 140 143 128 142 96 86 1384
The following diagram is an illustration of how the utilisation (in flight hours) varied
over the year 2000, per aircraft tail number.
500.00
450.00
400.00
350.00
Average-2000
Monthy Flight Hours
300.00 Reg. ZA
Reg. ZB
250.00 Reg. ZC
Reg. ZD
200.00 Reg. ZE
Reg. ZF
150.00
100.00
50.00
0.00
Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec
Period [Months]
132
Appendix J
500
450
400
350
Flight Hour-Flight Cycles
300
Min. Flight Hours
Max. Flight Hours
250
Min. Flight Cycles
Max. Flight Cycles
200
150
100
50
0
Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec
Calendar Months
Variable utilisation is based on the fact that the utilisation of the fleet may change
differ from that witnessed in the actual utilisation.
In order to incorporate such changes, the following utilisation scenarios are
developed:
1. Conservative Scenario The utilisation to be anticipated will be below
average
2. Most Likely Scenario The utilisation to be anticipated will be about
average
3. Optimistic Scenario The utilisation to be anticipated will be above
average
These scenarios apply for both flight hours and flight cycles.
133
Maintenance Planning Optimisation: B737 NG
Owing to the fact that the increment used in the simulation is in months, the monthly
utilisation should be determined. Further, the seasonal increment and decrement
should also be incorporated in the resulting monthly utilisation.
By considering the average aircraft utilisation, the proportion that each month
contributes is:
Suppose ai is the average aircraft utilisation (in hours or cycles) per month
(calculated for the period between 1999-2000), and that Av is the annual average for
the same period. The proportion (P) each month contributes on the utilisation curve
is:
ai
Pi = ; where i is the month being calculated
Av
12
From this follow the fact that: P =1
1
i
The values from table x. above are then added to the simulation program for the
generation of the required utilisation.
134
Appendix J
Option Explicit
'Attribute VB_Name = "Module1"
'UMXSMP/PICCO/Anthony\Visual Basic code
'
'
'
'UTDAYS = Days limit on check
'UTHOURS = Hours limit on check
'UTCYCLES = Cycles limit on check
'VISITS = Hangar visit date
'FREQUENCY = Hangar visit frequency
'RDAYS = Days in a Calendar month
'RHOURS = Production hours (following utilisation) corresponding to the indicated
month
'RCYCLES = Production cycles (following utilisation) corresponding to the indicated
month
'For the leading column applies: If Value>UTLIMIT then MOD(Value,UTLIMIT)
k = RVALUE + NextRw
j = PrevVal + PrevRw
l = j - UTLIMIT
If Value = UTLIMIT Then
'DueDate = l + RVALUE
DueDate = RVALUE
'Else
'If j < UTLIMIT Then
'DueDate = RVALUE
Else
'If (NextVis < Frequency) And (Value + k < UTLIMIT) Then
'DueDate = Value + RVALUE
If Value + k < UTLIMIT Then
DueDate = Value + RVALUE
Else
'If Visit = Frequency And (Value + RVALUE > UTLIMIT) Then
If Visit = Frequency Then
135
Maintenance Planning Optimisation: B737 NG
DueDate = UTLIMIT
Else
If NextVis = Frequency Then
DueDate = Value + RVALUE
Else
DueDate = Value + RVALUE
End If
'End If
End If
End If
End If
End Function
j = RDAYS + NextRd
k = RHOURS + NextRh
Else
If Visit = Frequency Then
TDDY = UTDAYS
Else
If NextVis = Frequency Then
TDDY = Value + RDAYS
End If
End If
End If
End If
136
Appendix J
End If
Else
If (Value2 + k < UTHOURS) Then
TDHR = Value2 + RHOURS
Else
If Visit = Frequency Then
TDHR = UTHOURS
Else
If NextVis = Frequency Then
TDHR = Value2 + RHOURS
End If
End If
End If
End If
If Switch = 1 Then
DUALDH = TDHR
Else
DUALDH = TDDY
End If
End Function
j = Value + RDAYS
137
Maintenance Planning Optimisation: B737 NG
Else
If Value < Freq And (Value + RDAYS >= Freq) Then
i = Freq
Else
If Value = Freq And (PrevVal = Freq) Then
i = RDAYS
Else
If Value = Freq And (PrevRd - (Freq - PrevVal) + RDAYS >= Freq) Then
i = Freq
Else
If Value = Freq And (PrevRd - (Freq - PrevVal) + RDAYS < Freq) Then
i = PrevRd - (Freq - PrevVal) + RDAYS
End If
End If
End If
End If
End If
End If
ToHangar5 = i
End Function
i = PrevVal + RDAYS
Else
LoSSeS = 0
End If
End Function
End Function
138
Appendix J
End Function
139
Maintenance Planning Optimisation: B737 NG
End Function
i = Value + UTVALUE
j = Freq
k = RVALUE
Else
If (ADDIT = 1) And (i > UTLIMIT) Then
CheckDue = UTLIMIT
Else
CheckDue = Value + RVALUE
End If
End If
End Function
140
Appendix J
i = RVALUE
j = RVALUE2
k = Value + UTVALUE
l = Value2 + UTVALUE2
If Switch = 1 Then
DUALDC = CYCSD
Else
DUALDC = DAYSD
End If
End Function
i = PrevVal + RVALUE
141
Maintenance Planning Optimisation: B737 NG
End Function
REAL LIFE
SYSTEM DATA
SYSTEM
ANALYSIS
INPUT CONCEPTUAL
MODEL MODEL
DATA MODEL
VALIDATION VALIDATION
(Replication) (Predictive)
QUANTITATIVE
MODEL
PROGRAM PROGRAMMING
VERIFICATION
COMPUTER
COMPUTER
VISUAL BASIC/
VISUAL BASIC/ EXCEL DATA
EXCEL DATA MODEL
MODEL
SIMULATION
Source: Syllabus-Digital Simulations A
J.5.1. Validation
Model validation is done in order to ascertain that the model is a reasonable
representation of the real life process. This follows from the fact that various aspects
of the real life situation are either simplified or left out, and in other cases, various
assumptions are made.
The following aspects of model validation can be considered (and tabulated here
below):
142
Appendix J
a. Assumptions
b. Input parameter values and distributions
c. Output values
[Hillston, 2001]
Table J. 5 Assumptions
Assumptions Real Life Model
1. Aircraft utilisation - Aircraft do not perform flight every day of - The aircraft performs flights daily
the year, especially not during the low season,
when the flight demand is low
- During the low season, a part of the fleet is - The aircraft considered is present in
leased out. The specific tail number of the the Transavia fleet for the period
aircraft to be leased out depends on the choice considered (not leased out)
of the lessee, and on the approval of the
technical department.
2. Aircraft utilisation - This is basically seasonal, as was illustrated - Follows the same seasonal pattern
pattern in Figure J. 1 above. witnessed in reality
- The utilisation strongly varies per tail number - Varies over time. Each simulation
and is not repeated over the years. No tail run results in a different utilisation
number flies to specific destinations. quantity (flight hours cycles)
- If an aircraft is leased out, the utilisation by - The aircraft utilisation considered
the lessee may differ strongly from that of always follows the Transavia pattern
Transavia,
3. Execution of - This varies from much in advance, to past - All clusters are performed on time.
maintenance clusters the due dates. In cases where a maintenance No extensions are incorporated in the
cluster cannot be performed on time, model
extensions are requested.
- The METALS system considers maintenance - The model considers tasks as done,
tasks as performed at the moment the aircraft once they have been clustered.
enters the maintenance facility.
3. Aircraft utilisation - Based on the exact utilisation of - Based on assumptions derived from
the aircraft. Varies with the reality, and projected situation;
seasons (see also Figure J. 1) assumptions modelled into utilisation
and per aircraft scenarios. Varies with seasons.
- Uniform distribution used in the
simulation process meant to represent
patterns on Figure J. 1.
143
Maintenance Planning Optimisation: B737 NG
From the data above, it may be assumed that the model created idealises the real
situation, without misrepresenting it. Hence, it may be sat that the model is valid.
J.5.2. Verification
Verification of a model is meant to certify that the program runs as intended, and
that it produces the desired outputs.
The following kinds of verification will be performed:
1. Verification of Logic/ Check against known solutions
2. Modular testing: Ensuring that each sub-routine produces sensible output for
all possible input
3. Sensitivity testing: Checking the effect of a parameter on the performance of
the model
4. Stress testing: Ensuring that the model works for extreme values of all input
parameters
Scenario 1
Random Values 1 Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec Total
Hours 178 205 227 243 285 240 286 304 271 290 198 197 2924
Cycles 93 90 115 127 127 141 152 155 146 132 91 84 1453
Random Values 2 Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec Total
Hours 179 190 244 275 276 246 322 305 310 283 212 163 3005
Cycles 97 98 100 120 125 139 148 158 154 135 89 90 1453
Random Values 3 Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec Total
Hours 191 205 243 265 239 284 333 324 290 323 199 180 3076
Cycles 101 87 106 128 143 130 156 145 146 144 88 89 1463
Scenario 2
Random Values 1 Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec Total
144
Appendix J
Hours 219 167 239 227 277 269 278 272 299 287 195 203 2932
Cycles 85 94 112 109 143 151 151 128 152 160 95 99 1479
Random Values 2 Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec Total
Hours 176 177 206 226 290 276 283 284 289 312 194 172 2885
Cycles 111 78 93 106 168 131 146 150 169 144 90 95 1481
Random Values 3 Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec Total
Hours 193 206 247 231 242 269 263 264 276 323 194 162 2870
Cycles 95 100 92 110 129 151 145 151 165 139 90 83 1450
Scenario 3
Random Values 1 Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec Total
Hours 191 189 244 250 286 288 286 312 291 310 193 179 3019
Cycles 97 93 106 125 142 141 156 155 149 145 95 90 1494
Random Values 2 Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec Total
Hours 186 203 213 229 288 297 304 323 267 322 211 182 3025
Cycles 100 92 110 119 135 145 153 149 147 146 93 88 1477
Random Values 3 Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec Total
Hours 182 174 203 230 283 232 262 330 281 317 181 174 2849
Cycles 95 96 106 119 134 144 152 157 145 147 94 90 1479
Scenario 4
Random Values 1 Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec Total
Hours 245 239 277 298 333 328 399 391 337 376 237 243 3703
Cycles 99 93 110 112 140 132 154 160 150 147 101 88 1486
Random Values 2 Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec Total
Hours 239 241 248 334 334 341 371 367 385 330 253 228 3671
Cycles 96 95 110 126 127 135 141 158 141 131 89 87 1436
Random Values 3 Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec Total
Hours 236 229 257 277 340 360 363 346 335 364 229 211 3547
Cycles 102 94 100 115 125 139 139 142 156 144 87 92 1435
Scenario 5
Random Values 1 Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec Total
Hours 255 245 283 303 307 353 369 380 385 365 217 246 3708
Cycles 117 111 127 118 147 147 144 178 148 132 90 108 1567
Random Values 2 Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec Total
Hours 230 251 272 298 304 306 396 344 345 383 219 217 3565
Cycles 116 97 132 129 146 153 177 145 122 149 103 98 1567
Random Values 3 Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec Total
Hours 251 226 263 332 354 321 387 406 370 377 252 241 3780
Cycles 112 109 92 137 129 144 127 151 148 163 84 97 1493
145
Maintenance Planning Optimisation: B737 NG
Scenario 6
Random Values 1 Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec Total
Hours 258 217 252 313 326 347 366 379 348 387 217 226 3636
Cycles 99 95 108 124 140 143 155 150 151 139 94 87 1485
Random Values 2 Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec Total
Hours 247 256 264 276 341 355 369 366 323 336 219 246 3598
Cycles 100 94 106 121 139 143 151 156 150 147 97 88 1492
Random Values 3 Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec Total
Hours 245 232 258 327 353 299 365 361 327 375 242 218 3602
Cycles 98 94 107 125 134 143 152 153 144 144 96 91 1481
Scenario 7
Random Values 1 Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec Total
Hours 305 277 321 350 425 408 471 477 456 404 302 259 4455
Cycles 103 87 99 127 143 139 159 159 140 143 97 83 1479
Random Values 2 Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec Total
Hours 288 259 350 375 421 410 448 472 394 437 266 292 4412
Cycles 94 99 110 114 131 139 155 142 140 137 97 84 1442
Random Values 3 Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec Total
Hours 304 292 317 363 376 401 464 410 433 443 267 271 4341
Cycles 102 95 111 113 127 142 159 159 142 145 97 82 1474
Scenario 8
Random Values 1 Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec Total
Hours 301 297 312 380 376 424 424 439 439 430 307 271 4400
Cycles 87 86 117 116 136 173 187 172 142 162 79 97 1554
Random Values 2 Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec Total
Hours 299 264 344 345 427 394 475 456 430 439 277 270 4420
Cycles 120 79 115 105 156 141 175 138 142 168 99 91 1529
Random Values 3 Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec Total
Hours 322 302 333 344 418 369 416 480 409 434 273 258 4358
Cycles 89 104 133 122 155 121 176 131 171 129 81 110 1522
Scenario 9
Random Values 1 Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec Total
Hours 282 270 350 380 399 425 475 412 396 401 285 275 4350
Cycles 97 93 106 125 142 145 154 149 152 144 93 87 1487
Random Values 2 Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec Total
Hours 306 262 335 380 387 414 414 483 392 450 304 278 4405
Cycles 99 96 111 122 142 143 156 157 148 141 96 92 1503
146
Appendix J
Random Values 3 Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec Total
Hours 307 267 331 368 373 381 472 457 429 452 296 281 4414
Cycles 98 97 112 121 135 144 149 152 149 144 97 87 1485
Scenario 10
Random Values 1 Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec Total
Hours 245 199 282 320 390 370 381 389 383 332 178 253 3722
Cycles 105 80 123 120 138 122 148 148 130 143 120 95 1472
Random Values 2 Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec Total
Hours 302 208 271 299 357 369 395 390 376 366 235 194 3762
Cycles 99 83 113 129 140 126 143 147 131 142 101 103 1457
Random Values 3 Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec Total
Hours 301 201 334 312 373 356 394 401 360 356 191 290 3869
Cycles 106 78 110 128 139 137 149 149 145 143 126 100 1510
It may be said that the random values drawn remain within the specified ranges
(Table J. 3). The extreme minimums and maximums are not attained, owing to the
large variation in the random numbers drawn. This does compare to the reality (see
also Figure J. 1).
It can therefore be said that the utilisation simulations produce the desired outputs.
In order to establish whether the maintenance scenarios have an effect on the base
maintenance man-hour demand, a number of runs are made for each scenario.
After three consecutive runs, the following base maintenance man-hours are
established for an 18-month interval:
Table J. 9 Base Maintenance Demand variation by scenario
Type Visit Scen-1 Scen-2 Scen-3 Scen-4 Scen-5 Scen-6 Scen-7 Scen-8 Scen-9 Scen-10
HMV 1 [Sep 02] 209 209 209 211 211 211 229 229 229 211
HMV 2 [Mar 04] 229 229 229 236 236 236 236 236 236 236
HMV 3 [Sep 05] 297 297 297 299 299 299 327 327 327 303
HMV 4 [Mar 07] 239 239 239 248 248 248 243 243 243 244
HMV 5 [Sep 08] 307 307 307 303 303 303 316 316 316 303
HMV 6 [Mar 10] 550 550 550 553 553 553 563 563 563 557
Total MHrs 1830 1830 1830 1851 1851 1851 1915 1915 1915 1855
From Table J. 9 above, it is clear that there is no variation in the maintenance demand
within a given utilisation (Conservative, Most Likely, Optimistic). The reason behind
this can be attributed to the fact that the differences in flight hours and cycles within
scenarios are relatively small compared to the large interval between two consecutive
base maintenance visits.
Further, it is evident that there is minimal difference between the Most Likely
utilisation and scenario 10. This follows from the fact that the Most Likely utilisation
range forms a large part of the actual utilisation See also Table J. 2 and Table J. 3.
147
Maintenance Planning Optimisation: B737 NG
Scen-1 Scen-2 Scen-3 Scen-4 Scen-5 Scen-6 Scen-7 Scen-8 Scen-9 Scen-10
2001 218 218 218 218 218 218 218 218 231 218
2002 410 422 410 467 455 448 475 475 462 467
2003 333 326 317 372 385 378 502 471 481 476
2004 397 405 426 460 452 460 463 494 463 356
2005 342 333 330 468 468 481 481 481 481 481
2006 412 421 424 458 452 426 385 373 407 458
Total MHrs 2111.8 2125.6 2124.8 2443.3 2429.6 2410.9 2524.0 2512.2 2524.4 2455.5
Despite the fact that the differences in the maintenance demand from various
maintenance scenarios are minimal, it can be said that the maintenance demand does
vary with the utilisation.
148
Appendix J
Table J. 13 Maintenance Demand variation from 3 runs: Scenario 2, 6-weeks Mx frequency
149
Maintenance Planning Optimisation: B737 NG
Table J. 13 till Table J. 15 clearly show that no three runs will result in identical
clusters all year round. However, the total maintenance demand remains more or
less stable.
The purpose of the simulation model was to ensure that the maintenance demand
follows the aircraft utilisation, and that this varies over the years and over the
selected periods (Months). The tables above indicate that this purpose is achieved.
Modular Testing:
Sensitivity testing:
The model works well when a single parameter is altered.
Stress testing:
Model inputs include: the maintenance frequency, the maintenance interval limits,
and the aircraft utilisation.
- The aircraft utilisation varies as listed in Table J. 3 above. The model behaviour
under this conditions it included in Table J. 8 above
- The maintenance frequency is defined to lie between the following ranges:
Line Maintenance 4 7 Weeks:
At a three-week frequency, the model functions inappropriately because it is
designed to cluster at intervals of 4 Weeks and above. Three weeks translates to
21 days, while the model works with values above 28 days (See also: Model
Validation).
At an eight-week frequency, the model functions properly as pertains to the
maintenance frequency sub-routine. However, this frequency leads to the
violation of a number maintenance limits, hence rendering all outputs invalid
- As for the maintenance interval limits, the model will calculate all limits not
lower than 28 days, 250 flight hours and 50 flight cycles. This is because it is
modelled to work with the lowest maintenance interval indicated in the
OMP. It will also not calculate Fight Cycle intervals exceeding 32780. The
reason behind this is not clear, but it may be attributed to the program (MS
Excel) itself. However, such high limits may be ignored because Transavia
150
Appendix J
does not operate airplanes with over 50,000 flight hours or 25,000 flight
cycles.
151
Appendix K
153
Maintenance Planning Optimisation: B737 NG
154
Appendix K
155
Maintenance Planning Optimisation: B737 NG
156
Appendix K
The quantities shown on Table K. 5 above are estimations of what the situation
would look after the application of an initial de-escalation. In reality, the exact
maintenance demand will differ from that tabulated above, owing to the fact that an
evaluation of tasks already performed needs to be done. The last base maintenance
visit also has to be reviewed.
157
Maintenance Planning Optimisation: B737 NG
158
Appendix K
Total 3164.4 3198.9 3175.5 3410.4 3407.4 3397.8 3615.8 3599.3 3601.7 3428.5
159
Appendix L
L. Appendix L
161