Content uploaded by Miguel Mujica Mota
Author content
All content in this area was uploaded by Miguel Mujica Mota on Jul 08, 2019
Content may be subject to copyright.
Simulation-Based Analysis of Operational Capacity for Airport Configurations of the System Santa Lucia-
Mexico City
M. Mujica Mota, A. Di Bernardi, P. Scala, D. Delahaye
SIMULATION-BASED ANALYSIS OF CAPACITY FOR A MULTI-AIRPORT SYSTEM:
MEXICO CITY CASE STUDY
M
iguel Mujica Mota
Alejandro Di Bernardi
Aviation Academy
Department of Aeronautics, Faculty of Engineering
Amsterdam University of Applied Sciences
National University of La Plata
Amsterdam, The Netherlands
La Plata, Argentina
Paolo Scala
Daniel Delahaye
Amsterdam School of International Business
-
IT &
Logistics Department
ENAC Lab
Amsterdam University of Applied Sciences
Ecole National
e
de l
’Aviat
ion Civile
Amsterdam, The Netherlands
Tolouse, France
ABSTRACT
The following paper presents the analysis of the potential operational capacity of the airport system in
Mexico composed by Santa Lucia Airport and Mexico City considering the most important elements that
compose the system: two airports and the common airspace that is shared by both infrastructure. The study
is based on stochastic simulation and it reveals that the capacity of growth is unleashed in the metropolitan
region of Mexico City once the new airport of Santa Lucia is constructed by the government; however the
study also reveals some limitations that the system will have once the traffic increases in the metropolitan
region of Mexico City.
Keywords: multiairport, NAICM, AICM, airport capacity, optimization, airports, passengers, congestion
1 INTRODUCTION
Mexico City airport is the main gateway to the country since years ago. However, its growth has been
hampered by the saturation of the airport which in most slots of the day it is impossible to accommodate
more traffic. The latter can be only achieved by performing three activities: infrastructure expansion,
optimization of the current resources or by managing the system under a different paradigm such as the
multi-airport system approach.
The previous government in Mexico decided for the first choice by constructing a new infrastructure
in an old lake which made the project a risky business with uncertain outcomes. For the previous reasons
and other consequences mainly environmental (Reuters, 2019) the new government (by late 2018) decided
to cancel the project and betted for a less risky approach by expanding an old military facility and changing
the approach to a multi-airport system which will be composed by Santa Lucia Airport, Mexico City Airport
and eventually Toluca which is also in the vicinity of Mexico City.
The opposition to this solution claim that the multi-airport system approach will not be able to solve
the original problem of saturation in Mexico city airport while at the same time maintaining a steady traffic
growth and also that the multi-airport system will not be equivalent in capacity to the previous option of a
completely new airport.
Mujica Mota, Di Bernardi, Scala, and Delahaye
The present study aims at answering some of the questions raised by the critics to the project, in
particular, it will answer the questions of whether this proposal is able to cope with the expected demand
in the coming years and if it is also able to solve the congestion problems in the current airport of Mexico
city. In addition, this study will provide some light in the expected performance indicators of the new
facilities and the limitations that can be expected once the expected traffic becomes a reality.
The present analysis involves the following aircraft fields
• Benito Juárez International Airport of Mexico City (MEX) which is the main airport for Mexico City
• Military Base "General Alfredo Lezama Álvarez" of Santa Lucia (NLU) which is currently a military
base, but it will be upgraded to attract commercial traffic.
The following figure illustrates the geographical location of NLU and MEX and the characteristics of
both airports.
Figure 1
Figure 1. Geographical location of NLU and MEX
Figure 2. Aerial view of system NLU-MEX
The study considered the two airports in their current phase and the expected developments for the
short, medium and long term horizons.
2 STATE OF THE ART IN MULTI-AIRPORT SYSTEMS
The topic of Multi-Airport Systems (MAS) has been gaining some attention the last few years as many
issues regarding complex airport systems have been studied. The concept of a MAS is defined as one main
Mujica Mota, Di Bernardi, Scala, and Delahaye
airport with another or more secondary airports that together serve a metropolitan region and it has diverse
issues that require attention, such as capacity, coordination, selection, sustainability and feasibility among
others.
Regarding the definition and feasibility, the seminal paper of de Neufville (1995) introduced the
analysis of the viability of MAS by defining that the air traffic of a metropolitan area should exceed 10
million originating passengers per year so that a MAS could be economically and operationally viable,
however this number has increased up to 15 million in some cases. On the other hand, the paper of Martin
and Voltes-Dorta (2011) provides some caution for the development and use of MAS. They suggest,
considering a financial approach, that some MAS worldwide are operating inefficiently and that the
consolidation of air traffic of the whole MAS into one airport could provide a better performance regarding
operating costs. However, their conclusions did not consider that, as the utilization of any capacity-
constrained resource increases in a stochastic environment, the service levels of the system rapidly
deteriorate with a non-linear function (Hopp and Spearman, 2000) as fig 3 illustrates.
Figure 3. Example of an AirTraffic/DelayTimes curve in stochastic environments
Furthermore, Fasone et al. (2012) and Yang et al. (2016) suggested that the viability of a MAS is
intertwined with the development of other transport infrastructure, such as, railways, roads and bus services,
that connects customers and cargo with the various airports in the system so that customers of the MAS
could have accessible options to use any of the airports in the system and change their initial preference
regarding the principal airport.
As mentioned, de Neufville and Odoni (2013) state the viability threshold, which in 2013 they calculated
was 15 million passengers per year for originating passengers, discarding the transfer ones.
Regarding the issue of airport selection, the subject of the main factors involved influencing selection
among customers has been extensively studied using statistical methods (Hess and Polak, 2005; Loo, 2008;
Ishii et al., 2009; Marcucci and Gatta, 2011; de Luca, 2012; Fuellhart et al., 2013; Nesset and Helgesen,
2014). These papers found that air fare, access time, flight frequency, the number of airlines and the
availability of particular airport–airline combinations were statistically significant factors in customer
choice of airport. Interestingly, airport access time was found to be more important for business travellers
than for leisure travellers. In contrast, leisure travellers were found to be more sensitive to price changes
than business travellers.
The specific issue of multi-airport capacity has only been studied by Ramanujam and Balakrishnan (2009).
The study by Ramanujam and Balakrishnan focuses on the definition of capacity envelopes for the MAS
of NYC, based on Gilbo (1993) proposal. Using quantile regression and historical data, they modelled the
relation between arrival and departure rates at singular airports considering the arrival rate as the
independent variable, as arrivals are given priority over departures at singular airports. In addition, they
also modelled the relation between arrival and departure rates of different airports, because the (airspace)
Mujica Mota, Di Bernardi, Scala, and Delahaye
approach and departure paths of different airports in MAS could interfere with each other. They found that
the visibility factor is significant for arrivals but not for departures and that the capacity envelope area is
increased when using one runway for arrivals and a different runway for departures, instead of a mixed use
of runway for arrivals and departures. They also found that airside capacity is more significant for defining
airport capacity than airspace as approach path overlap factor was not found to be statistically significant
for capacity envelope definition.
Regarding Operational capacity of an airport, there are diverse studies attempting to estimate it for
a singular airport resource, such as, runway capacity (Gilbo, 1993; Hockaday and Kanafani, 1974; Bäuerle
et al., 2007; Janic, 2008; Janic, 2014), terminal capacity (Janic and Tosic, 1982; Solak et al., 2009) and
terminal capacity (Mujica Mota and Zuniga Alcaraz, 2015; Mujica Mota, 2015). In addition, there are few
attempts to model the actual capacity of the whole airside operations of a singular airport, i.e., runways,
taxiways and apron operations. Modelling the complete set of capacity-constrained resources of an airport
could provide practitioners and researchers with better decision tools for design and management of the
complete airside facilities of an airport as the interactions among different serialized queues could create
different behaviour patterns than singular resources. In literature, only the work of Mujica et al. (2014)
present this approach. The paper by Mujica et al. (2014) analyses, using Discrete Simulation, the capacity
and performance of Lelystad Airport assuming that some traffic will be diverted from the highly utilized
Schiphol Airport in Amsterdam. They modelled the capacity of Lelystad Airport considering historic data
of wind visibility and airport traffic and considering various operative restrictions, such as, the separation
criteria between aircraft operations, weather conditions, mix of aircrafts and type of taxiways. They found
that the use of rapid exit taxiways could increase the throughput of a singular airport.
Thus, this literature review shows that regarding airport systems some studies have covered different
aspects of single airports and by using some mathematical techniques some aspects of the MAS, but no
authors have modelled a Multi-Airport System considering an integral approach (two or three airport
operations together with airspace and/or different elements of the airport). This type of study could provide
great insight in understanding the consequences and potential problems that might appear once the system
of airports is operational. Consequently, the objective of the paper is to address this gap as it will be focused
on studying a twin system that will be operational in the near future in Mexico considering the current
Mexico City International Airport and the future Santa Lucia Airport.
Mujica Mota, Di Bernardi, Scala, and Delahaye
3 METHODOLOGY
The methodology followed in this work, is the one presented by Mujica et al. (2018) for developing a multi
model system in which a combination of models are developed in order to create one that minimizes the
uncertainty associated with the modelling process (Fig.4).
Figure 4. Methodology of the n-model virtual cycle approach for airport capacity
The developed model considers the following elements:
• Benito Juárez International Airport (IATA Code: MEX). For the modelling of this airport, we used a high-
detailed model developed by the authors for diverse studies (Mujica Mota et al., 2018).
• Santa Lucia Airport (IATA Code: NLU). For this airport the airport model considers the characteristics
reported in public documents of the Mexican government and the development project of the airport. A
macro model of this airport is developed.
• Current aeronautical routes. For the traffic approaching to the two airports, the current aeronautical routes
have been considered and once NLU is operational, a modification based on the expertise and experience
of the authors, has been proposed.
• Current capacities of MEX runway system regarding runway, taxiways and terminal stands
Mujica Mota, Di Bernardi, Scala, and Delahaye
• Capacities of the NLU airport regarding runway and stands for the macro model.
The analysis is performed by developing different scenarios considering different assumptions and taking
the expected demand of traffic as an input for the model together with current restrictions and traffic mix.
All the technical restrictions that correspond to airspace and ground operations of the airports under study
have been considered.
In the first instance, Scenario 0 is analyzed, This scenario represents the baseline for comparison of the
following 6 scenarios that are used for characterizing the capacity of the aeronautical operation. This
scenario is the current situation of the airport system of Mexico City.
3.1 Boundary Conditions and Analysis Criteria
The analysis is carried out considering the following operational assumptions:
• Current MEX airport layout.
• Sequential configuration of runways, taxiways and platforms for NLU.
• Current traffic mix for MEX airport.
• Current airspace based on Mexico AIP will be considered for the base case scenario.
• A feasible redesign of airspace to allow operation of NLU and MEX together is proposed.
The analysis is also made based on the following general considerations:
• The mix indexes% (C + 3D) of the AICM are maintained.
• 44,320,000 Pax / year is adopted as the current passenger level in Mexico City.
• 414,000 Movements / year is adopted. Starting with 590 arrivals/day.
• 18 hours of operation and 120 Pax / Average aircraft for both airports is assumed
• An annual operation is assumed for both airports
• The slot management model is maintained.
• The parking spaces available at MEX are maintained at 103
• 33 Parking places available in NLU are assumed from scenario 0 to 5a
• The considerations for the aeronautical capacity and associated airspace are conceptual.
All the simulations carried out consisted of simulations of 30 hours of operation and for each scenario 30
replications were made for obtaining the statistical indicators.
4 SCENARIO ANALYSIS AND RESULTS
The following section presents the different scenarios evaluated for the current study, starting with a base-
case scenario and progressively modifying it for evaluating different situations.
4.1 Scenario 0
This scenario serves as the base case and represents the current situation of the airport in Mexico City. It
makes use of a low-level simulation model, based on MEX’s AIP, as well as using current air traffic values
to compare the operational results with the subsequent scenarios.
The model includes the following elements:
MEX Air Space
Runway system 05R-23L and 05L-23R
Terminal 1 and Terminal 2 with 103 aircraft parking positions in total
Taxiway system with speed restrictions
Flight plan for an average day of operation
Mujica Mota, Di Bernardi, Scala, and Delahaye
Traffic mix that includes low cost airlines (LCCs) and full service airlines (FSCs) and different
aircraft equipment
For modelling the complete system, a simulation model of the current airspace together with the model of
MEX were used. Figure 5 illustrates the airspace of the MEX model. The current routes reported in the AIP
of Mexico (SENEAM, 2016) are used.
Figure 5. Elements of the airspace model
Figure 6 illustrates the low-level operating model implemented in MEX. With this model, all the
emergent dynamics can be identified, as well as conflict situations that limit the capacity of the system
like runway occupancy times, runway crossings, delays, congestion among others.
Mujica Mota, Di Bernardi, Scala, and Delahaye
Figure 6. Airside model of Mexico City Airport
By running experiments with this model, we obtained performance indicators for the system, and we
validated them statistically comparing them with historical data. Figure 7 presents the evolution of traffic
during the day for the airport where clearly it can be seen that the levels of congestion are reached from 10
am until late at night (such as the real situation).
Figure 7. ATM evolution during the day
Regarding the remaining performance indicators for the elements that compose the system, the
following table present different values obtained with the experiments.
Mujica Mota, Di Bernardi, Scala, and Delahaye
Table 1. Performance indicators of MEX
MEX
Avg Value
Max Value
ATM/Hr
38.6
62
Gate Occupancy
55
%
78
%
Aircraft Waiting Runway
6
15
Aircraft Waiting Gate
0
6
Total Annual Passengers
30.4
MILL
48
MILL
It can be seen that the maximum value of ATM/HR corresponds to the declared limitation of the airport
by the government. In addition, it can also be perceived that the remaining elements of the system are not
fully congested during the day, only during peak times, revealing the effect of the business models of the
companies that operate in the airport. From the analysis, it can also be perceived that the runway is the
bottleneck of the system as it has been known for years now in Mexico.
4.2 Scenario 1
For this scenario, a redesigning of the air routes that respect the operational restrictions was made, so that
aircraft can fly from north and south to both airports without infringing safety restrictions.
It is assumed that low cost airlines LCCs transfer their operation to NLU and the Hub operation of legacy
carriers is kept in MEX. The same airspace design was maintained for later scenarios.
Figure 8 illustrates the simulated airspace as well as the location of the two airports under study.
Figure 8. Air routes re-design for the system NLU-MEX
Mujica Mota, Di Bernardi, Scala, and Delahaye
Table 2 shows the values of the air routes used for the study. These were designed in such a way that they
respect the separations imposed by ICAO for avoiding safety violations.
Table 2. Description of the landing routes for the MEX and NLU airports
Destination
airport
Route
Waypoint
Altitude
[ft]
Speed
[kts]
MEX
DATUL/AVSAR/TUMAL/KOBEK/
DOS ALFA
LUCIA
12.000
220
MATEO (IAF)
12.000
220
PLAZA (FAF) 8.800 130
MEX
MEX TRES ALFA
MEX
FL 240*
250
D23
-
MEX
18.000
250
MATEO (IAF)
12.000
220
PLAZA (FAF)
8.800
130
MEX
VIVER CUATRO
VIVER
12.000
250
MEX
12.000
220
MATEO (IAF)
12.000
220
PLAZA (FAF)
8.800
130
NLU
Route LUCIA 1
LUCIA2 12.000 220
LUCIA3
12.000
220
LUCIA4 (FAF)
8.800
130
NLU
Route LUCIA 2
MEX
FL 240*
250
D23
-
MEX
18.000
250
LUCIA3
12.000
220
LUCIA4 (FAF)
8.800
130
This scenario would correspond to a stage prior to the construction of the two runways referred in the Ante
Obra de Santa Lucía (Grupo Rioboo, 2018). This scenario would also free up MEX capacity without
affecting traffic growth. In addition, the function of the hub of MEX (in which full-service airlines operate)
is not affected, and the growing demand of the low-cost airlines that service domestic demand is not
hampered by this strategy. This scenario assumes that some LCC companies and charter ones will operate
in NLU like the following: Magnicharter, Viva Aerobus, Interjet and Volaris; even some LCCs from the
US could operate in this airport like Southwest or JetBlue. The same premise is maintained for subsequent
scenarios in which we increase the volume of traffic.
The following figures illustrate how the evolution of demand would behave in this scenario during a day
for both airports.
Mujica Mota, Di Bernardi, Scala, and Delahaye
Figure 9. Evolutions of movements in NLU and MEX
As it can be seen, once the airports are operating independently, the saturation of MEX is solved as in
the worst situations the number of ATMs are 45 ATM/hr. The following table presents the remaining
performance indicators for this scenario.
Table 2. Performance indicators for Scenario 1
MEX
NLU
AVG Value
MAX Value
.
AVG Value
MAX Value
Traffic Share
57%
43%
ATM/Hr
23
44
17
41
Gate Occupancy
47%
59%
48%
100%
Aircraft Waiting
Runway
1
6
0
0
Aircraft Waiting
Gate
0 0 0 8
Total Annual
Passengers
18 MILL 34.7 MILL 13.4 MILL 30.7 MILL
Mujica Mota, Di Bernardi, Scala, and Delahaye
Under the mentioned assumptions, the demand is shared by the two systems, and the maximum
expected traffic with the type of equipment assumed would imply a maximum of 34 million of passengers
for MEX and 30.7 million for NLU, making a total of almost 65 million passengers for the combined
system.
4.3 Scenario 2
This scenario would correspond to the time-horizon when two runways have been completed in NLU, and
the HUB operation of FSC would move to NLU while the LCCs would move to MEX. In this scenario, a
simultaneous operation of landings and takeoffs in NLU will be possible (due to the two runways), and
MEX would have enough room to absorb the growth of low-cost airlines as it is seen in the results of the
simulations.
Figure 10 presents the evolution of traffic of a simulated day for NLU and MEX respectively.
Figure 10. Evolution of Traffic in NLU and MEX
In this scenario, MEX remains as an airport for low cost airlines (LCCs); it can be seen that MEX is totally
decongested, since the maximum number of movements would be 37 ATM / Hr or 60% of its capacity. On
Mujica Mota, Di Bernardi, Scala, and Delahaye
the other hand, NLU would have two tracks and operate at the time of maximum demand at 47 ATM / hr
which suggests that it would not be even close to the current situation of MEX.
With this type of demand it could be assumed that some 67 million passengers could be moved annually
between the two airports with very reasonable operational indicators as the following table suggests.
Table 3. Performance indicators for Scenario 2
MEX
NLU
AVG Value
MAX Value.
AVG Value
MAX Value
Traffic Share
43%
57%
ATM/Hr
14
37
25
47
Gate Occupancy
19%
44%
70%
100%
Aircraft Waiting
Runway
0 4 0 1
Aircraft Waiting
Gate
0
0
5
23
Total Annual
Passengers
11 MILL 29 MILL 19.7 MILL 37 MILL
It is important to note that it is perceived that the first bottleneck for NLU will be the aircraft parking spaces,
since it is observed that there would be an average value of 5 cases of aircraft that do not have a gate when
landing. Furthermore, in times of high demand this number can go up to 23 aircraft.
Scenario 3 and the following ones are designed to determine the growth limits and the elements that
would restrict growth due to operational and capacity constraints.
4.4 Scenario 3
In this scenario, the same proportion of traffic mix between FSCs and LCCs is maintained, the variation
of traffic consists of an increase of 10% in the demand for LCCs and 10% in FSC for both MEX and
NLU. This scenario would correspond to a mid-term scenario assuming the current traffic growth trend.
Mujica Mota, Di Bernardi, Scala, and Delahaye
Figure 11. Evolution of Traffic in NLU and MEX with a 10% increase of demand
It is observed that when increasing the demand, MEX does not present major setbacks. In the case of NLU,
it is clear that the number of operations increases to a max of 50 ATM / hr at times of maximum demand,
although on the other hand it is already clear that to avoid problems of reactive or induced delays it will be
necessary to implement remote posts or add more gates to the infrastructure since the problems of aircraft
without gate is evident (Table 4).
Under this scenario, some 70 million passengers could be expected (60% more than what MEX alone is
currently receiving).
Mujica Mota, Di Bernardi, Scala, and Delahaye
Table 4. Performance indicators for Scenario 3
MEX
NLU
AVG Value
MAX Value.
AVG Value
MAX Value
Traffic Share
43%
57%
ATM/Hr
15.5
40
27.5
49.6
Gate Occupancy
19%
44%
75.8%
100%
Aircraft Waiting
Runway
0
3
0
1
Aircraft Waiting
Gate
1 1 14.4 45
Total Annual
Passengers
12.2 MILL 31.5 MILL 21.7 MILL 39 MILL
In this situation the problem of lacking gates for NLU is evident as the results illustrate. During peak
hour, the problems could be severe (45 Aircraft waiting) which might be translated in delays for the flights
at certain moments of the day.
4.5 Scenario 4
The increase in air traffic corresponds to 30% of LCCs and 30% of FSCs in MEX and NLU respectively.
As previously mentioned, LCCs traffic can grow in MEX and FSC in NLU with a Hub-Spoke business
model. This scenario would correspond to a medium-term horizon as well.
Figure 12. Evolution of Traffic in NLU and MEX with 30% increase
Mujica Mota, Di Bernardi, Scala, and Delahaye
It can be seen that NLU has grown to a similar level of the current situation of MEX, however, in a more
efficient way since it can operate in a simultaneous fashion the two runways. It can also be realized that
MEX does not present major problems, since it would be operating at a daily average of 18 ATM / Hr with
peaks of 45 ATM/hr. However, in the case of NLU, it is already clear that in terms of runways there would
be no problems, but it would be necessary to find a solution to the lack of Gates as Table 5 illustrates.
Table 5. Performance indicators for Scenario 4
MEX
NLU
AVG Value
MAX Value.
AVG Value
MAX Value
Traffic Share
43%
57%
ATM/Hr
17.7
45
31
65
Gate Occupancy
25%
49%
83%
100%
Aircraft Waiting
Runway
0
3.6
0
2
Aircraft Waiting
Gate
0 1 44 119
Total Annual
Passengers
14 MILL 35.5 MILL 24.4 MILL 51 MILL
Mujica Mota, Di Bernardi, Scala, and Delahaye
4.6 Scenarios 5a and 5b
In scenario 5a and 5b air traffic increases by 70% in MEX and NLU respectively. This scenario would
correspond to the assumption that traffic would grow as predicted in the next 50 years. This would be a
long-term scenario, and would allow to evaluate the operation and limitations to absorb the expected
traffic.
Scenario 5a
Figure 13. Evolution of Traffic in NLU and MEX with 70% increase in demand
As it has been mentioned NLU presents numbers of operations according to the type of airport (2
independent runways), and MEX reveals that under this configuration, it could grow without major
problems for the next 50 years. However, NLU would have severe problems in case no gates expansion is
performed as Table 6 illustrates.
Mujica Mota, Di Bernardi, Scala, and Delahaye
Table 6. Performance indicators for Scenario 5a
MEX
NLU
AVG Value MAX Value. AVG Value MAX Value
Traffic Share
43%
57%
ATM/Hr
24
49.5
36
89
Gate Occupancy
28%
70%
85%
100%
Aircraft Waiting
Runway
0.7 8 0 2.8
Aircraft Waiting
Gate
0 1 112 256
Total Annual
Passengers
19 MILL 39 MILL 28.4 MILL 70 MILL
Under the expected traffic, NLU gate infrastructure would be severely limited (as results reveal), for this
reason, an alternative scenario (5b) is proposed which contemplates an expansion of the parking positions
for aircraft and terminal building. In 5b, the number of parking spaces is doubled (66 Gates).
Mujica Mota, Di Bernardi, Scala, and Delahaye
Scenario 5b
In 5b, the amount of traffic is similar, but with the double capacity of gates for NLU. Under this
circumstance, the problem of lack of gates is partially solved but not completely. From Table 7 it can be
noticed that there are still some problems during some days as still some aircraft do not find a gate. The
latter suggests that the double of gates probably is not enough for the operation, instead it is necessary to
invest in more than 33 gates.
Figure 14. Evolution of Traffic in NLU and MEX with 70% traffic increase + facility expansion
Table 7 presents the complementary indicators for the long-term scenario. As it can be seen, MEX
reveals the limitation of the Runway as some aircraft will be limited by the runway, however with a proper
management of the sequence of the expected traffic mix, this can be minimized. In the case of NLU, the
gates are the limiting factor for growth.
Together in the most optimistic scenario, it could be expected to absorb a maximum of approximately
120 mill of passengers in the metropolitan region.
Mujica Mota, Di Bernardi, Scala, and Delahaye
Table 7. Performance indicators for Scenario 5b
MEX
NLU
AVG Value
MAX Value.
AVG Value
MAX Value
Traffic
Share
43%
57%
ATM/Hr
24
50
42
105
Gate Occupancy
28%
70%
57%
100%
Aircraft Waiting
Runway
0.7
8
0
4
Aircraft Waiting
Gate
0 1 8 50
Total Annual
Passengers
19 MILL 39.5 MILL 33 MILL 82.7 MILL
5 DISCUSSION AND CONCLUSION
The study presented for the first time a methodology for performing a simulation-based analysis of a Multi-
airport system. We presented the case of Santa Lucia and Mexico City Airport which has become a key
development for the country. The study consisted in different scenarios based on public information and
governmental plans using three models: one for Mexico City airport, another for Santa Lucia and another
one for the airspace that connects both airports. The experiments with the different scenarios gave light to
some important issues regarding the development of the facilities such as the capacities of the system and
the limitations that will appear when the growth in traffic takes place in the airport system. Some important
results about the scenarios are discussed.
Regarding Scenario 0, we could identify that the main bottleneck is the runway, which coincides with what
has been discussed publicly in the media. Depending on the time of the day, the effect of the runway is
more or less severe. In addition, we could also identify that the limit of the capacity of 61ATM/hr can be
reached sometimes. Assuming these operating levels, it can be estimated that this airport could absorb a
capacity of 48 million passengers assuming the average aircraft type with an occupancy of 120 passengers,
maintaining continuously 61 ATM / Hr, which is currently unfeasible.
Scenario 1 gives light on the operational levels of the system NLU-MEX system with one runway in NLU.
Under the assumptions presented, NLU can operate with values of 41 ATM / Hr without major problem
(using only one runway). In the case of MEX, it can be seen that the congestion problem is solved as it
operates with an average of 40 ATM / hr, or what is the same at 65% of its current capacity. With the release
of capacity, it would be expected that the problems of flight delays would be drastically reduced, and in
addition to that, it would also be expected that the Mexican national airport network would operate without
major setbacks with the consequence of deactivating the Ground Delay Program (GDP) which is currently
active due to congestion (Mujica and Romero, 2018).
With the following scenario where the traffic is increased (Scenario 3 to 5) they reveal that NLU will suffer
from a lack of gate capacity, with the consequence of flights waiting for a parking position, inducing delays
to the airlines and the national network. Scenario 5b reveals that the investment in the medium term in more
gate capacity would alleviate partially the limitation, but it would suggest that the double of gates would
not be sufficient for solving the problem. The different scenarios reveal that by implementing the system
NLU-MEX and with the proper timely investments, the growth in the metropolitan region of Mexico City
is unleashed and it has potential to grow up to a three digit level in terms of passengers.
The study presented, revealed the different capacities the system will have at different time horizons; the
short term that consists of the current situation and some 5 years more, medium term for approximately 30
yrs and a long term scenario of 50+ years in which we can identify the amount of passengers, number of
Mujica Mota, Di Bernardi, Scala, and Delahaye
movements and potential problems that will arise during the operational life of the system. The analysis
provides enough information for giving light about the potential areas of improvement and requirements of
expansion in the coming years. This could not be achieved without the use of simulation technology; for
this reason, the authors strongly encourage the use of this methods and technology during the planning
phase of any critical infrastructure.
ACKNOWLEDGMENTS
The authors would like to thank the Amsterdam University of Applied Sciences for the support in this
research, as well as the Dutch Benelux Simulation Society (www.DutchBSS.org) and EUROSIM for
disseminating the results of this work.
The publication reflects the views only of the authors and it does not necessarily reflect the views of the
institutions that provided support for the realization of the study.
REFERENCES
de Luca, S. (2012). Modelling airport choice behaviour for direct flights, connecting flights and different
travel plans. Journal of Transport Geography, vol. 22, pp. 148-163. doi:
http://dx.doi.org/10.1016/j.jtrangeo.2011.12.006.
de Neufville, R. (1995). Management of multi-airport systems. doi:http://dx.doi.org/10.1016/0969-
6997(95)00035-6.
de Neufville, R., Odoni, A.R. (2013). Airport Systems: Planning, Design, and Management, 2nd eds,
McGraw-Hill Education, New York.
Fuellhart, K., O’Connor, K., Woltemade, C. (2013). Route-level passenger variation within three multi-
airport regions in the USA. Journal of Transport Geography, vol. 31, pp. 171-180. doi:
http://dx.doi.org/10.1016/j.jtrangeo.2013.06.012.
Gilbo, E. P. (1993). Airport capacity: representation, estimation, optimization. IEEE Transactions on
Control Systems Technology, vol. 1, no. 3, pp. 144-154. doi: 10.1109/87.251882.
Grupo Rioboo (2018). Ante Programa de Obra, Ante Presupuesto, Ante Proyecto Arquitectónico AISL
Hess, S., Polak, J.W. (2005). Mixed logit modelling of airport choice in multi-airport regions. Journal of
Air Transport Management, vol. 11, no. 2, pp. 59-68. doi:
http://dx.doi.org/10.1016/j.jairtraman.2004.09.001.
Hockaday, S.L.M., Kanafani, A.K. (1974). Developments in airport capacity analysis.
doi:http://dx.doi.org/10.1016/0041-1647(74)90004-5.
Hopp, W., Spearman, M. (2000). Factory Physics, Second edn, McGraw-Hill.
Ishii, J., Jun, S., Van Dender, K. (2009). Air travel choices in multi-airport markets.
doi:http://dx.doi.org/10.1016/j.jue.2008.12.001.
Janić, M., Tošić, V. (1982). Terminal airspace capacity model. doi:http://dx.doi.org/10.1016/0191-
2607(82)90052-8.
Janic, M. (2008). Modelling the capacity of closely-spaced parallel runways using innovative approach
procedures. Transportation Research Part C: Emerging Technologies, vol. 16, no. 6, pp. 704-730. doi:
http://dx.doi.org/10.1016/j.trc.2008.01.003.
Janic, M. (2014). Modeling effects of different air traffic control operational procedures, separation rules,
and service disciplines on runway landing capacity. Journal of Advanced Transportation, vol. 48, no.
6, pp. 556-574. doi: 10.1002/atr.1208.
Loo, B.P.Y. (2008). Passengers’ airport choice within multi-airport regions (MARs): some insights from a
stated preference survey at Hong Kong International Airport. Journal of Transport Geography, vol. 16,
no. 2, pp. 117-125. doi: http://dx.doi.org/10.1016/j.jtrangeo.2007.05.003.
Mujica Mota, Di Bernardi, Scala, and Delahaye
Marcucci, E., Gatta, V. (2011). Regional airport choice: Consumer behaviour and policy implications.
Journal of Transport Geography, vol. 19, no. 1, pp. 70-84. doi:
http://dx.doi.org/10.1016/j.jtrangeo.2009.10.001.
Martin, J.C., Voltes-Dorta, A. (2011). The dilemma between capacity expansions and multi-airport
systems: Empirical evidence from the industry’s cost function. Transportation Research Part E:
Logistics and Transportation Review, vol. 47, no. 3, pp. 382-389. doi:
http://dx.doi.org/10.1016/j.tre.2010.11.009.
Mujica Mota, M., P. Scala, G. Boosten (2014). Simulation-based capacity analysis for a future airport. 2014
Asia-Pacific Conference on Computer Aided System Engineering
(APCASE),10.1109/APCASE.2014.6924479.
Mujica Mota, M. (2015). Check-in allocation improvements through the use of a simulation–optimization
approach. Transportation Research Part A: Policy and Practice, vol. 77, pp. 320-335. doi:
http://dx.doi.org/10.1016/j.tra.2015.04.016.
Mujica Mota, M., Zuniga Alcaraz, C. (2015). Allocation of Airport Check-in Counters Using a Simulation-
Optimization Approach. Applied Simulation and Optimization: In Logistics, Industrial and
Aeronautical Practice, eds. M. Mujica Mota, I.F. De La Mota & D. Guimarans Serrano, Springer
International Publishing, Cham, pp. 203-22910.1007/978-3-319-15033-8_7.
Mujica Mota, Di Bernardi A., Scala P., Ramirez-Diaz G. (2018). Simulation-based Virtual Cycle for Multi-
level Airport Analysis. Aerospace, April 2018
Mujica Mota, M., Romero R. (2018) Analysing the Decision-Rules for a Ground Delay Program: Mexican
Airport Network. In Proc. of Int.Conf of Research in Air Transport 2018, Castelldefels, Spain
Nesset, E., Helgesen, Ø. (2014). Effects of switching costs on customer attitude loyalty to an airport in a
multi-airport region. Transportation Research Part A: Policy and Practice, vol. 67, pp. 240-253. doi:
http://dx.doi.org/10.1016/j.tra.2014.07.003.
Ramanujam, H. Balakrishnan (2009). Estimation of arrival-departure capacity tradeoffs in multi-airport
systems. Proceedings of the 48h IEEE Conference on Decision and Control (CDC) held jointly with
2009 28th Chinese Control Conference,10.1109/CDC.2009.5400462.
Reuters (2019). Building of New Mexico City Airport suspended, but some works continue. January 3,
2019
SENEAM, MEXICO CITY AIP, 2016
Solak, S., Clarke, J.B., Johnson, E.L. (2009). Airport terminal capacity planning. Transportation Research
Part B: Methodological, vol. 43, no. 6, pp. 659-676. doi: http://dx.doi.org/10.1016/j.trb.2009.01.002.
Yang, Z., Yu, S., Notteboom, T. (2016). Airport location in multiple airport regions (MARs): The role of
land and airside accessibility. Journal of Transport Geography, vol. 52, pp. 98-110. doi:
http://dx.doi.org/10.1016/j.jtrangeo.2016.03.007.
AUTHOR BIOGRAPHIES
MIGUEL MUJICA MOTA holds a Ph.D and M.Sc. in industrial informatics and Operation research from
the AUB in Barcelona and Mexico respectively. He is an Associate Professor at the Aviation Academy of
the Amsterdam University of Applied Sciences in the Netherlands. He was subdirector of aviation at the
Autonomous University of Barcelona. He is chair of the Dutch Benelux Simulation Society and executive
board member of the EUROSIM federation of simulation societies of EUROPE. Dr. Mujica has given
several courses in modelling, simulation and optimization in different countries for industrial and academic
audiences. He collaborated with industry for practical projects. He is the co-author of four books and
numerous scientific papers on simulation, operations research and aviation. His research interests lie in the
use of simulation and heuristics for the optimization and performance analysis of aeronautical and logistic
operations. Currently he studies diverse problems of the Aviation Network of Mexico City and the Multi-
airport System of North-Holland among other activities. Recently he has been appointed as Researcher I in
the National Council of Research of Mexico. His email address is: m.mujica.mota@hva.nl
Mujica Mota, Di Bernardi, Scala, and Delahaye
ALEJANDRO DI BERNARDI has obtained his Master in Airport Systems in Madrid Polytechnic
University (Spain), as well as three specialist titles in airport subjects, in this same University. He graduated
as an Aeronautical Engineer in National University of La Plata (Argentina), where he teaches the subjects
“Reaction Engines”, “Alternative Engines”, “Airports and Flight Operations”, “Planning and Design of
Airports”. Director of the postgraduate specialization “Airport Projects”. He is also Director of the "Unit of
Research, Development, Extension and Transfer" UIDET GTA.-GIAI. "GIAI" “Group of Engineering
Applied to Industry" and G.T.A. "Air Group Transport” of the Department of Aeronautics.
Instructor in seminaries ICAO/Aena in CAR-SAM region. ICAO expert in the fields of “Aerodrome
Engineer” , “Airport Master Plan Expert”, “Airport and Engine Emissions Expert”, and ”Aviation Trainer
Expert”. Advisor and instructor in the Central American Institute of Aeronautical Capacitation (ICCAE) of
COCESNA. Responsible of the design and management of two airport diplomats. He has taught courses,
classes, conferences or talks at more than 35 institutions and universities in 17 countries. Co-author of
several publications in the area of airports, air transport and aeronautical engines.
Director of different investigation and engineering projects, having worked on more than 100 airports and
heliports in Argentina, Brazil, Chile, Colombia, Ecuador, El Salvador, España, Peru and Uruguay and
having participated in the elaboration of 37 airport master plans in Argentina. He participate in the
development of various studies on safety assessment, and on sustainable air transport, and the elaboration
of long-term air transport strategic plans in Argentina and in Peru, His email address is
cadibern@ing.unlp.edu.ar
PAOLO SCALA received the B.Sc. and M.Sc. in industrial engineering at the University of Calabria,
Rende, Italy in 2011 and 2014 respectively; He is currently a PhD candidate with the Ecole Nationale de
l’Aviation Civile(ENAC)/Universite Paul Sabatier, Touluose, France in the subject of applied mathematics.
He is a Researcher and Lecturer at the Amsterdam School of International Business – Department of IT &
Logistics at the Amsterdam University of Applied Sciences, The Netherlands. He has several publications
in International Journals, and a chapter in a book. He published and presented in many international
conferences. He is the Secretary of the Dutch Benelux Simulation Society. His research interests lie in the
use of simulation and optimization in air traffic management. His email address is: p.m.scala@hva.nl
DANIEL DELAHAYE received the B.Eng. degree from Ecole Nationale de l’Aviation Civile (ENAC),
Toulouse, France; the M.Sc. degree in signal processing from National Polytechnic Institute of Toulouse,
in 1991; and the Ph.D. degree in automatic control from Aeronautics and Space National School, Toulouse
in 1995. In 1996 he was Postdoctoral Researcher with the Department of Aeronautics and Astronautics,
Massachusetts Institute of Technology, Cambridge, MA, USA. In 2012 he received his tenure in applied
mathematics from the Universite Paul Sabatier, Toulouse, France. He is currently the Head of the
Optimization Group with the Mathematiques Appliquees, Informatique et Automatique pour l’Arien
(MAIAA) Laboratory, ENAC, where he is conducting research on stochastic optimization for airspace
design and large-scale traffic assignment.