Skip to main content

Full text of "An Analysis of Delay and Travel Times at Sao Paulo International Airport (AISP/GRU): Planning Based on Simulation Model"

See other formats


Santana, E, S. M 
Mullen C 



AN ANALYSIS OF DELAY AND TRAVEL TIMES AT SAO PAULO EVTX AIRPORT 
(AISP/GRU): PLANNING BASED ON SIMULATION MODEL 

Erico Soriano Maitins Santana <ehco@ipv,ct£Lbr> 

SiiBiilatioii Lftbontory 

Insdtnto de Prote^ ao Vdo 

Sao Jose dos Campos * SP - Brazfl 

Carlos Muller <muller@infrajta.br> 

AeroBantical Inftmstimctiirc Eiighicariiig XKvislon 

Institiito Tecnologico de Aemnaiitica 

Sio Jos^ dos Campos - SP ~ Bran! 

ABSTRACT 

The occurrence of flight delays in Brazil, mostly verified at the ground (airfield), is re^onsibfe for serious 
disruptions at the airport level but also for the unchaining of problems in all the airport system, affecting also the 
air^ace. 

The present study develops an analysis of delay and travel times at Sao Paulo International Airport / Guarulhos 
(AISP/GRU) airfield based on simulation model. 

the conqjotadonal tool developed to represent aircraft operation in the airspace and airside of airports, was used to 
perfonn diese analysis. 

The study was mainly focused on aircraft operations on ground, at the airport runway, taxi -lanes and aprons. The 
visualization of the operations with increasing demand facilitated the analyses. 

The results generated in this work certify the viabiiiiy of the methodology, ihey also indicated the solutions capable 
to solve the delay prt^lem by travel time analysis, thus diminishing the costs for users mainly airport authority. It 
also indicated alternatives for airport operations, assisting the decision-making process and in the appropriate tuning 
of the proposed changes in the existing infrastructure. 

Key Words: Airports, Capacity, Simulation, Demand, and Delays 



L INTRODUCTION 

An increasing number of users of air transportation in the whole world can be observed along the 
years (ICAO, 1999). Al&ough the attacks in New York (United States) on September 1 1th, 2001 
have generated conservation (in some cases fell) in these numbers, the wlK)le aviation sector 
e:q>ects the retaking of the growth of indexes in a short tenn. 

In spite of the apparent pause in the growth, traffic jams in the airspace and airports continues to 
h^pen and causes frequent delays, harming all partic^)ants of tiie airport system. The search for 
solutions mobilizes from airlines to air traffic control or^nizations. 

To have an idea of the losses, in November 2000, the airport operations in Brazil generated losses 
in a Brazilian airline that reached the value of US$ 1 miUion (Targa, 2001). These losses. 



Santana, E. S, M. 
Muller, C. 



generated in concentrated way due to delays in the flights, most of the time they were verified in 
the ground (airports), unchaining problems in every system, reaching even the airspace. 

Due to the observation of the problem, analyses were made in possible operational scenarios in 
the International Airport of Sao Paulo / Guarulhos (AISP/GRU), main airport of South America 
in passengers' movement, based on a simulation model attesting its viability as a methodology in 
airport planning systems, varying the growth of the demand expressed for the aircraft's 
movement. 

For this purpose it was used, the computational tool SIMMOD. A tool developed by the Federal 
Aviation Administration - FAA, civil aviation authority in the United States, capable to show 
through simulation the operation in the airspace and airside of airports, producing resuhs that 
allow the analysis of such systems. 

This methodology is consecrated in the field "airports operation analysis", with several works 
al-eady implemented in the worid (TRANSSOLUTIONS, 2000; Delcaire, B., Feron, E., 1997; 
Irani, A. A., Wing-Ho, F., 1997), including Brazil (Pereira et al., 2000; Barros, 1994). 

The models in SIMMOD are discrete, dynamic and stochastic. SIMMOD represents airports and 
airspace system as a series of nodes connected by links. A node is defined as the position in a 
system of coordinates where the simulation evaluates the location of an aircraft with regard to 
other aircrafts from the system. All events happen according to the existent characteristics in the 
construction of such links. 

The operation of the models developed in the tool is described in the Figure 1, being part of their 
events, attributes and entities considered as input data. In the tool there is also a great amount of 
information related to the operational performance of the aircrafts that can be classified as entities 
in the system besides an algorithm well buih as for the airport operations. 

2. JUSTIFICATION 

The Sao Paulo Terminal Area, where AISP/GRU is inserted, is being reason for concerns from 
planners, administrators and operators of the airports, due to the "bottlenecks" existent in air 
operational procedures generating a traffic jam situation and problems as for its viability on the 
ground or in the airspace. 

Due to the complexity of the problem, the users of the system are recurring to computer science 
more frequently, which subsidizes the development of new capacity analysis methodologies, 
where the simulation has its outstanding status. As main justification for the use of such a 
methodology, it is verified that the modeling using simulation techniques became quite efficient 
in the resolution of complex problems, possessing a very low cost if compared to the 
accomplishment of an experience with the real system. In the context of the air transportation, 
SIMMOD has been showing quite satisfactory results. 



Santana. E. S. M. 
Muller, a 



However the modeling of the object in study, its identification and the search for the best 
guidehnes in Hit present researches were not tasks easy to be acconq)lished, once tiie naain 
analyses were resulting of several technical visits to the control tower of the airport, exhausting 
dialogues with the responsible for the traffic in the ground (airside of the airport). Ml learning of 
the tool SIMMOD and, above all, the understanding and construction of operational scenarios for 
the new tffrminal of passengers (TPS - 3) and of the new track of landings and takeofifs. Their 
strategy of operation is still not known. 

For the CGSStruction of &e scenarios it w*as considered the increases in the demand, farming an 
affiance of this criterion with the changes of the physical layout (improvement of the 
infiBstructiire), a new runway for landings and takeoffe, and irxq)lantation of a new terminal of 
passengers (TPS - 3). The sensibility analysis were made confronting the demand increase in the 
flights against the value of delays and displacement times accomphshed in the operations of 
landings and takeoffe called as travel times, measuring all the influence of tiie improvement of 
the airport infrastructure in the simulated alternatives. 

The results will be able to assist fte decisionmaking process and in the appropriate timing of 
the pyroposed changes in the existing infrastructure. 

3. AIRPORT SYSTEM CAPACITY 

The airport capacity is defined as "the maYimimi number of operations of aircrafrs, established 
for a certain aerodrome, for specific periods, support^ by the airport infi:astructure" (Siewerdt, 
2001). Starting from diis concept ll^ decisions taken in relation to the development of the 
infrastructure are decisive for the visualization of the coii^)lete increase of system capacity. 

However, the capacity is also limited fay the weakest factor of the system where it is inserted, 
since tiie airspace edacity, or a new runway of landings and takeofifs, or apron area, or terminal 
of passengers, or simply the accessibility of the passengers to the terminal. 

According this information it can be noticed how big is the problem of the increase in the 
capacity of airport systems. As consequence, it is observed a necessity of searching solutions, for 
the con5)lete system - airspace until the access to tiie terminal - or sinq>ly for each fector that 
involves it. 

4. SAO PAULO INTERNATIONAL AIRPORT (AISP/GRU) 

The biggest airport conq)lex in Brazil, AISP/GRU was conceived originally to assist the 
metropolitan area of Sao Paulo to d^ demand of domestic flights that used the Congonhas 
airport, except for the shuttle service between Rio de Janeiro - Sao Paulo, as well as the 
international flights related wiA the integral countries of the South Cone, also serving as an 
alternative for the Campinas airport 

Although it has been planned to serve a certain scenario and objective, the adninistration 
aufliority of the Airport couldn*t maintain the initial idea, making necessary the inq>lantation of 
new strategies equable to absorb the unpredicted demand. 



Santana, E, 5. M. 
Mullen C. 



The AISP/GRU, the Brazil's biggest "gate of entrance", is responsible for more than 70% of the 
international passengers (origin/destiny) processed. It concentrates, in Brazil, the largest traffic of 
regular passengers, the first relative placement to the total volume of freight and postal transport, 
besides the great movement of aircrafts of total regular traffic (INFRAERO, 2000). 

4,1. New facilities 

Looking forward to solve these problems that are happening, mainly in the part of infirastructure 
of AISP/GRU, Brazilian Airports Infrastructure Company - INFRAERO - is proposing as 
alternatives, the construction of a new terminal of passengers (TPS - 3), and another runway of 
landings and takeoffs, absorbing simultaneous operations of the aircrafts, identified as the 
bounding factor for the increase of system capacity. 

5. METHODOLOGY - THE MODELS CONCEPTION 

Starting from the definition of the problem to be studied, the first step of the methodology 
consists of listing the data and necessary information for the conception of the models and tihe 
means to obtain them. 

For that, basically, the model demands that are detailed the mherent aspects to the demand, 
operation of the aircrafts (Brazilian Ministry of Aeronautics, 1987) and geometric configuration 
of the airport infrastructure and of the terminal airspace (Brazilian Ministry of Aeronautics, 1993) 
existent today. The construction of the model is a computational work in which all this 
information is inserted in SIMMOD (Pereira et al., 2001). 

The last phase for conclusion of the basic model consists of the validation and verification of it. 
Some metrics are generated and selected by the simulation for comparison with real data of the 
airport operation, verifying if the obtained deviations are within the maximxmi limits of tolerance 
stipulated. 

The conception of the models followed the plans of enlargement of the infrastructure of 
AISP/GRU, which have been proposed by INFRAERO. 

In the study now presented were conceived 6 models, relative to two new configurations, besides 
the simulation of the current situation. Each one of the configurations was simulated with the 
present demand (registered on March of 2000) and with increments that vary from 2% to 30% on 
the current demand, totaling 15 increment values. The results for each one of them were 
converted in delays (cost), and confronted against the growth of the demand. The comparative 
analysis was generated from these results. 

The demand on March 2000 had in its peak day 555 operations; while in February 2002 the 
number of operations in the peak day was 534 operations (INFRAERO, 2002). In other words, on 
these 2 years, the change was not significant; there was conservation in the numbers. These times 
were originated by facts as the attacks in New York on September 11, 2001, besides the exit of 



Santana, E. S. M, 
Muller, C. 



the airline Transbrasil tiiat paralyzed a great range of the market with domestic and international 
flights. 

The 6 models biiih and simulated with the tool SIMMOD were: 

• Ml: Current situation; 

- M2: Hypothesis with tte addition of the third runway of landings and takeoffs; 

' M3: Hypodiests wiA tte a<kUtion of die third runway and dmd terminal - first operational 

strategy; 

M4: Hypothesis with ti^ addition of the third runway and third terminal - second operational 
strategy; 

• M5: Hypothesis with the addition of the third terminal - first operational strategy; 

• M6: Hypotiiesis with the addition of the third terminal - second operational strategy 

Due to die inexistence of information about the behavior and the operation of the new terminal of 
passengers, in development, in tiiis work two cations of operational strategies were adopted, 

5.1. Current Situation Model 

The model built to show the current situation of AISP/GRU, followed the data from AIP 
(Brazilian Aeronautical Command, 2000). It has two parallel runways of landings and takeoff, 
being ov£^ with 3000m of length (09R/27L) and tlK other with 3700m (09L/27R). The separation 
between the axes of the runways is 375m, In the direction 09, the thresholds are 500m staggered, 
runway 09R is used primarily fta: the operations of landings (85%) (lAC, 1999), and runway (^L 
is used most of the tin^ for takeoff operations. 

The sta^ered runway configuration results an effective separation among runways of 485m 
(addition of 3Qm for each 15Qm of staggering). Howevar, the miTiirriUni separatioa suggested by 
ICAO is of 760m to segregated dependents operations of landings and takeoffs in parallel 
runways, in visual flight rules (Barros apud Nagid, 1994). Though, runways with separation 
larger than 300m can operate in the system dual lane, in other words, authorized takeoffs i^en 
the aircraft touch the ground. Such is tiie procedure adopted by AISP/GRU to optimize the use of 
the existent infrastructure. 

The airport has 2 terminal building (finger concept), with eleven parking positions each one. 
Only thiee positions at the fin^r extnranity can receive aircraft larger than Boeing 767. This feet 
is due tte initial proposal in which the airport would assist only South America flights. However, 
INFRAERO are already modifying the structiire of the aircrafts parking positions in the terminals 
and should conclude this stage until the end of 2003. 

The 2 terminals buildings have a planned capacity to receive, each one, up to 7.5 million 
PAX/year. 

5.2. Third Runway Model 

The cfevelopment of this model was directly linked to the new project that INFRAERO is 

developing. 



Santana, E. S. M, 
Muller, C. 



The master plan of the airport established a third runway with 2025m of extension. However to 
implement that, it was adopted its construction inside existent patrimonial area, imder the 
responsibility of INFRAERO as basic premise. According this, the solution was to locate the 
runway to 1462m from the runway 09L/27R axis, reducing the original length for 1800m. 

The most appropriate position for the implantation of the third runway inside of the patrimonial 
limits, it was then to 1462m of the axis of the current runway 09L/27R, resulting in the 
dimensions of 1800m X 45m, could be enlarged for the 2025m, predicted in the master plan, by 
incorporation of new areas to the airport. 

lAC (Civil Aviation Institute) in the report of viability of operation of the third runway of 
landings and takeoflfs (lAC, 1999) has verified that a runway of 1800m would allow the landing 
and takeoff, in most of the cases with 100% of the acceptable maximum weights. Besides, as for 
the positioning of the runway of parallel taxi-lanes, the criterion of FAA was adopted (Federal 
Aviation Administration), that it establishes a removal of 120m of the axis of the new runway, 
compatibb with aircrafts with wingspans minor or equal to the one of the Group IV (up to 52m 
of wingspan), that reaches Fokker 100, Boeing 737, 767, and others. The report attests that about 
80% of the programmed operations to the airport can be accomplished in the future third runway, 
excluding only the aircrafts that accomplish international flights. 

A factor that hamis the operation of the third runway lives in the accessibihty of the aircrafts for 
the accomplishment of the landing operations and takeoffs. Starting from that consideration the 
need was verified of doing an analysis among the values of delay and travel times. 

With the entrance of the operation of the third runway, LAC (LAC, 1999) in the same study it was 
presented the configuration of the new mix of aircrafts to the airport together with the percentile 
of use of the runways as for the operation (Table 1) that was used in the conception of the model. 

5.3. New Terminal of Passengers (TPS-3) Model 

As well as the development of the previous model, the implantation of a new terminal of 
passengers in the model considered the premises of the project bid by LNFRAERO, 
approximating to the reality the future operations. 

Starting from the knowledge of demand studies accomplished by LAC (Brazilian Ministry of 
Aeronautics, 1999), where it was pointed for the year 2017 a movement of passengers in the 
order of 39 million PAX/year, INFRAERO bid in 1996 the project for the construction of the 
third tenmnal. Their main characteristics were (INFRAERO, 2000): 

• Capacity for 12 million PAX/year; 

• Architectural concept should accommodate the largest possible number of aircrafts parking on 
nose in, with two positions dedicated to NLA's (new large aircraft) and the minimum of seven 
positions for aircrafts type Boeing 747-400; 



Santana, E. S, M. 
Muiler, C 



• Expansion Capability, making possible the implantation of the fourth tenninal with identical 

dimensions; 

• The access circulation to the bridges was divided in two levels to separate passengers* departure 

and arrival flows; 

• Mix of stands for aircrafts of great and medium loads; 

- Semi-automated System of dockages of aircrafts. 

- Constnictiosi of one more remote parking area for the aircrafts of great, mediimi and small load, 
access the tiiird runway and still a small inear terminal, that will absorb part of the domestic 
trafxic. 

5.3J. Operational strategies in the Third Passengers Terminal (TPS-S) 

As mentioned previously, the operation of the new terminal of passengers (TPS-3) of AISP/GRU 
is an unknown that reaches all of the users of the system. 

Once it is not the objective of the present work to point the best ways to obtain the best operation, 
it was cbcided to come iq> with 2 propositions of operations and to submit them to the simulation. 

The 2 prc5>ositions were based in: 

Percentile Division of the operations for each airline: 

• Alliances among ^ airlines; 

• Apron Cqiadiy - size of liw aircraft and tteir restrictions; 

- Fhght Plm Characteristics; 

- Operations balance wifldn the 3 terminals. 

The way 2 operational scenarios were generated where there was a proportionality' to the c^acitj' 
projected for the pas^ngers* processing, TPS-1 and TPS-2 with 7.5 milUon PAX^year, 
representing 27.8% each of the passengers' demand and TPS-3 witii 12 milhon PAX/year, 
representing 44.4% (Table 2). It is worth to remind that the scenarios were elaborated according 
to demand of March 2000 (INFRAERO, 2000). 

6. DEMAND 

Looking for to evaluate the study object better, and to produce results to generate significant 
analyses as for tiie edacity on the airport airside, it was verified that the variation of the demand 
associated to the simulations in the models would consist in the very inq>ortant step. 

The stucfy "Detailed Demand Analysis of Brazilian Airports" (Brazilian Ministry of Aeronautics, 
1999) indicated that a 30% demand growth on Ae mmiber of operations was expected in the 
AISP/GRU in a 5 year-horizon. The adoption of a range of increasing demand values was 
influeiK:ed by tiie lack of an iq>dated forecast model after the countless problems caused in the 
world and BraziUan aviation by the attacks in the USA, that caused the observed break and 
decrease of fUghts. 



Santana, E. S. M. 
Muller, C 



Therefore, models representing 16 demand levels, ranging from 0% to 30% demand increment, 
were simulated and the for each one the level of traffic and the respective delays and travel times 
were obtained. 

The results allowed identifying the potential behavior of the average delay and average travel 
time, evidencing, this way, the appropriate instant to intervene in the installed infrastructure. 

7. RESULTS AND ANALYSES 

To generate results with statistical significance, although the experience suggests 5 iterations as 
enough to ensure statistically valid resuks (TRANSSOLUTIONS, 2000), 20 iterations of each 
analyzed model were performed. 

After executing the simulation, the first concern was that regarding the validation of the model, 
turning it capable to accomplish the other proposed analyses. The vaUdation stage was 
accomplished starting from the comparison of the results of the simulated system with the real 
system, testing logically and numerical the model. 

The vahdation followed the criteria: 

• Representative day = peak day of the data base supplied (March of 2000) 

• Total Period used in the simulation = 1 day of airport operations = complete cycle 

• Validation Process and verification = comparison of the generated results (Figure 2) 

Besides the number of operations, the process of validation of the model of the current situation 
consisted of the close verification of the input data from the report of the International 
Consultancy MITRE Co. (MITRE, 2001) contracted by the Civil Aviation Authority (DAC). In 
this report, MITRE Co. mentions that the maximum capacity is between 46 and 49 operations, 
varying that number according to the operation type, departure or arrival. The model developed 
m SIMMOD obeyed exactly to these numbers, arriving in the maximimi number of 50 operations 
with an increase in the demand of 30% reaching its capacity limit. 

In Figure 2 the adaptation of the "flotation" of the number of operations was observed, besides 
with the proximity of the "peaks" and "valleys" operational between the real values and the 
simulated model. 

Other metric verified was that related to the number of aircrafts in the takeoff line that coincided 
with the numbers of INFRAERO. In the peak hour, these values reached 9 aircrafts in the waiting 
line on the ground, both in the developed model and in the real operation. 

7.L Analysis of the Results 

The analysis of the simulated operational models is linked to the demand and offer of the system, 
besides the installed capacity of the airport. 



Santana, E. S. M, 
MuUer, C. 



Observing Figure 3 it is noticed Ae influence of the third runway of landings and takeoffs firstly 
in the decrease of delays. It is clear the existence of the division, among the models, in 2 groups 
with seaned characteristics as for the variation in the evolution of the demand. They are 
differentiated amongst themselves just by the inclusion of the third runway in their scenarios. 

It is also demon^iated in the same Figure 3, that the percentile growth of the delays doesn't 
depend a lot on the choice of the operational strategy for the third terminal of passengers (TPS-3). 
Tte behavior is i^ntical consictering or not the existence of the third runway of landings and 
takeoffs. However, in both cases, th^- accompanied the tax of percentile gro\\'th of the delays. 

As expected, the results generated starting firom the simulation of the models 3 and 4 those that 
presented ihe minor values were when £^hed to the increase of 30% in the demand of aircrafts. 
Almost 75% of delays increase in the inodels 3 and 4, against 495% in the model 1 (current 
situation of infrastructure). 

Conqjleting the anal3^ses of the study is necessary to confront the delays against the travel times 
accomplished by the aircrafts in their courses in the groxmd. 

The fliird runway of landings and takeoffs possesses a small problem as for its operation already 
described pre\'iously. The aircrafts to re^ch the "new" runway threshold 09 will face a long taxi 
distance increasing tierefore the travel time on the ground. 

The aircraft would be le^ subject to the delays (happened in the gates, in the taxi-lanes and 
aprons), once the itineraries would not conflict with the existent procedures in the current taxi- 
lanes. However, the traveled time to the runway threshold 09 "new", in the case of takeoffs, 
would be high. Ihe time spent for the arrival procedures would not be so accentuated, once the 
liu^cat Juuculiy wuulu be to reach the nmway threshold 09 "new" for takeoff. 

The results were practically the same, where the differential once again was the inclusion of the 
third runway of landings and takeoffs. The demand usually varied and the travel time also in the 
same way, in accordance with the same growth taxes. 

On Figure 4, it is verified that the tax of growth of the travel time proceeded exactly to the tax of 
growth of tiie demand, generating a constant graph for Aeir average values. Once constant, the 
difference can be measured by the average time the aircraft spent in the models where there is the 
inclusion of a new rum^^y of landings and takeoffs and h the models in that the operations 
follow aD for tbe same runway. 

This time observed in Figure 4 is of 2 minutes. In other words, in the models where the new 
nmway is considered, the operations were added by 2 minutes in relation to those that diddt 
considered the third runway. 

According to Table 2, where the models differ for the existence of the new runway of landings 
and takeoffs, and of the new temainal of passengers, it can be verified that starting from an 



Santana, E. S. M. 10 

MMer, C 



increase of 14% in the demand of the operations in the Airport, the difference among their 
average delays passes 2 minutes. This value had been mentioned as the reference pattern on the 
average time of trip spent by the aircrafts that use the new runway of landings and takeoffs. 

On Figure 5, it is noticed that when increasing 30% in the demand of the movement of aircrafts, 
the difference among the average values of delay for operation passes 5 minutes. 

However, for demand values up to an increase of 14%, the delay among the models 2 and 5, and 
2 and 6, tums smaller difference than 2 minutes, resulting non effective the construction of the 
third runway of landings and takeoffs, once the average travel time stayed constant in 2 minutes 
for any demand increase. 

The installed capacity was not completely used for any of the mentioned cases. However there is 
a great tendency, starting from 30% in the increase of the demand of the movement of the 
aircrafts, that AISP/GRU, reach its operation limit quickly, above all in the model 1 that 
represents the operation situation lived in the days today, with many points of operational 
conflict. 

However, one of the great problems visualized now at the airport regards the concentration of 
flights in certain schedules, causing excessive delays in certain hoiu^ of the day. The existence of 
idleness during other hours of the day made possible the operations in the simulations, although 
with many delays. 

8. CONCLUSIONS 

The viability of use of the methodology as an aid to the decision- making in airport planning is 

observed. 

The study suggests that the construction pf a third runway of landings and takeoffs, along with 
taxi- lanes, would bring more benefits in the long run (30% demand increase) of the operation of 
the airside as opposed to the solely construction of a new terminal of passengers (TPS-3). 

However, the best option would be the construction of the 2 facilities, the new terminal of 
passengers and the new runway of landings and takeoffs. For an increment of 30% in the 
demand, the impact on delays would be very small, once the average delays would reach values 
close to the ones observed in 2000. 

The new terminal of passengers alone would be operationally more effective than the third 
runway of landings and takeoffs up to 14% increase in the demand of the Airport. 



Santana, E. S.M. 11 

Mullen C 



ACKNOWLEDGEMENTS 

To CAPES, for the financial support dwing the MSc program; 
To Instituto de Protegao ao Voo - IFVJor the use ofSIMMOD. 

BIBLIOGRAPHICAL REFERENCES 

Barros, A. G. (1994) Analise de Capacidade do Lado Aereo do Aeroporto Intemacional de Sao Paulo/Guandhos, 

Master Tli^s^ instituto Tecnoldgico de Aeronautica, Sao Jose dos Campos- SP- Brazil. 
COMAKDO DA AERONAUTICA (2000) AIP-MAP, Dirctoria de Elctronica e Protc^ao ao Voo, Rio de Jai^iro- RJ 

-BiasiL 
Coscarelii, C. H. (2001) Estudos de Capacidade da Infra -Estrutura Aeroportudria Brasileira: Andiise do Emprego 

de Novas Tecnologias de Simulate, Master Thesis, COPPE/Universidade Federal do Rio de Janeiro, Rio de 

Janeiro - RJ- Brazil. 
Deicaire, B., Feron, E. (1997) Development of an on-site Ground Operations Model for Logan International Airport. 

FAA Air Transportation Center of Excellence in Operations Research, Research Report RR-97-09, MIT, 

Cambridge. 
HoronjefF, R., McKclvey, F. X. (1994) Planning and Design of Airports^ McGraw-Hill. 
INFRAERO (2002) Boletim de Informagoes Gerenciais, Dirctoria de Opcra^ocs, Brasilia - DF - Brazil. 
INFRAERO (2000) Relatorio Tecnico sobre a Implementagdo da Terceira Pista de Pousos e Decolagens e Terceiro 

Terminal de Passageiros (TPS-S), Centro de Neg6cios Sao Paulo, Sao Paulo- SP - Brazil. 
Instituto de Avia^ Civil (1999^ Parecer Tecnico n ° 401/DPT-4/99, Dqjartamento de Avia?ao Civil, Rio de Janeiro 

-RJ-BraziL 
Law, A. M, Kelton, W. D. (1991) Simulation Modeling & Analysis, McGrall Hill. 
MiNlb i tRiO DA AERONAUTICA (1987) IMA 100-12 - Regtas do Ar e serri^os de Trdfcgo Acrcc. Diretoria ds 

Elctrd3iica e Piote9^ ao Voo, Rio de Jaimo— RJ- Brazil. 
MINISTfeRIO DA AERONAUTICA (1993) Carta de Planejamento de Voo e Rotas Aereas, Dirctoria de Eletr&nica 

e Prote^ 80 Voo, Rio de Janeiro- RJ- Brajdl. 
MINISTfeRIO DA AERONAUTICA (1999) Hardrio de Transporte - H07RAN, Departamento de Avia^ao Civil, 

Rio dc Janeiro- RJ - Brazil. 
MINIST6R10 da AERONAUTICA (1999) Dewumda Detalhada dos Aeroportos Brasileiros. Departamento de 

Aviapao Civil, Rio de Janeiro- RJ - Brazil. 



IvUTRE Co. (ZGOl) Ai^yx.,^ ^u^ Air^p^c Sy:;tc7r. Plzr^tir- fcr the Sac Pa^i!c »e^-- - '^^^^ '>' ^^^h^^i^ nfCurr^mt 

Operations, Departamento de Avia^ao Civil, Rio de Janeiro - RJ - Brazil. 
OACI (1999).4miaWo Estatistico , Montreal - Canada. 
Pereira, B. D.; Silva, C. M; Guedes, E. P. e Feitosa, M. V. M. (2000) Andiise de Capacidade da Area de Movimento 

do Aeroporto Santos Dumont via Modelo de Simulagdo por (hmputador. Anais do XIV Congresso dc 

Pesquisa e Ensino em Transportes , ANPET, Gramado- RS - Brazil, v. 1 , p. 3 1-39; 
Pereira, B. D.; Silva, C. M; Guedes, E. P. e Feitosa, M. V. M. (2001) Estudo dos Atrasos na Area Terminal Sao 

Paulo: Situagdo Atual e Futura via Modelo de Simulagdo, Anais do XV Congresso de Pesquisa e Ensino em 

Transportes, ANPET, Campinas-SP- Brazil, Vol. 3. p. 157-161. 
Santana, E. S. M. (2002) Andiise de Novos Cendrios Operacionais para o Aeroporto Intemacional de Sao Paulo / 

GuaruJhos, Master Thesis, Instituto Tecnologico de Aeronautica, Sao Jose dos Campos- SP- Brazil. 
Siewerdt, E, (2001) Gestdo de Aeroportos e Espagos Aereos no Limite de suas Capacidades^ U Simposio dc 

Transporte Aereo - SITRAER, Instituto Tecnol6gico de Aeronautica, Sao Jose dos Csmpos - SP - Brazil. 
Targa, D. (2(K)1) Uma Ferramenta Automatizada no Auxilio a Alocagdo de Slots para o Problema de Gerenciamento 

de Fbao de Trafego Aereo Brasileiro, Master Thesis, Instituto Tecnologico de Aeroniutica, Sao Jose dos 

Ckirq>os — SP — Brazil. 
Trani, A. A., Wing-Ho, F. {1991) Enhancements to SIMMOD: A Neural Network Post-processor to Estimate Aircraft 

Fuel Consumption^ NEXTOR Research Report RR-97-8, Department of Civil Engineering, Virginia Tech, 

Blackshurg. 
TRANSSOLUTIONS (2000) The Airport and Airspace Simulation Model, Basic Simmod Training Rio de Janeiro- 

RJ- Brazil. 



Santana, E. S. M. 
Muller, a 



12 



ANNEXES 



^ijmz*o WfWt 



C'AG 






50 
45 

40 



**^e-PTt?cfc^s^«' '<imc: ^pu 



Sir^1L-A~'iDf^ ENaiP^t; Ojtpu! ft«j:» ■ ^ ard tarz 



fo?tii.ife 






Figure 1. The Simulation Framework from SIMMOD Tool 
Source: Delcaire AFeron, 1997 

Operations (Real X Simulated) 






35 .:^- 



30 



2D - 

15 -■ 



ID -' 



5 -■ 



■■ * # ■% 



.^*> - .^^. 



■» " '»' ' T ■' t IP — ^^.i.-,,, 



0:00 2:00 4 00 $;00 e^OO 1000 12:00 14 OO 16:00 18:fX} 2000 220G 

Time (GMT) 



Figure 2. The Simulation Validation and Verification 



Santana, £. S. M. 
Muller, a 



13 



[>elay5 against CurrBnt Sitt^on Model Scenario Base 



^>:»% 4-- 






Ml ~ Currant smiation 
M2-RWY3 

M3-7PS34RWY3-1 
M4-*rPS3&RWY3- 
M5-7PS3-1 
M6-1F83-2 




16 1o 2C 22 



- 1 r; >^- 



Demand (%> 

FWgr? ?. Sii!!i!l«tloii llMnltx frnni 6 Models 

THAVB. HUES - ARR « 09 i AVBWSE) 



ie 



u 



13 4 



12 4 



10 



T * 




- - ^^^w^*V^ 


4- " 1 


•^2mimAes 


^ ^ jt ^1 


L*rtfc=-^rt A^ yv_^^^ 11%^^ A y» fs A — a-— ^ 


. , Jfr.\,*!t ?? ?f JT. . ,.^-„^;^. ,. -T. T". in r: r: «- .■*■ v- ^ 







-Ml 



.M2 — .^M5 



M4 



-M? 



34 26 28 30 

Demand (%) 



Figure 4- The Simulation Results - Travel Times (average) - 6 Models 



Santana, E. S. M. 
Muller, C. 



14 



DELAY - Min/OPER (ARR 4- DEP) 




-M: -*>;— M3 



It; 18 20 22 3^ 26 2i> 3.; 

Demand (7o) 



■ Mr:. -^--Me ■ 



Figure 5. The Simulation Results -Delays (average) - 6 Models 
Table 1. Aircraft's Operational Mix to AISP/GRU 



Runways Using Percentile 


Aircraft 


Third Runway 


09L/27R 


09R/27L 


B737 


100% arr 


35,13% dep 


64,87% dq) 


FlOO 


100% arr 
37,4% dep 


62,6% dep 


- 


El 20 


100% arr and dep 


- 


- 


F50 


100% arr and dep 


- 


- 


B767 


26,56% arr 


100% dep 


73,44% arr 


MDll 


- 


100% dep 


100% arr 


B747 
A300 


- 


100% dep 
100% dep 


100% an- 
100% arr 



Santana, E. S. M. 
Mailer, C. 



15 



Table 2. Operational Mix for Third Passengers Terminal at AISP/GRU 



Scenario 01 



Star Alliance and others 

TAUe Group 

Vasp e Group 



43.91% 
28.01% 
28.09% 



TPS3 
TPS5 
TPS-1 



Long Iteul Fiiglits and BrazHian airlines 
Scenario 02 Vaspe Group 

Domestic and tnt South America 



43.76% TPS^ 

28.15% TPS-2 

28.09% TPS-1 



Authors addresses: 

tria^ Smiamo Mmrtims Samittna 

Instituto de Protegdo ao Voo- JPV 
Pra^ Marecha! Eduardo Gomes, 50 
Sao Josi dos Campos SP SKiZH 
<erico@^.cSa.br> 

Carlos MMUer 

Instituto Tecnologico de Aeronautica- TTA 
Praga Marechal Eduardo Gomes, 50 
Sao Jose dos Campos -SP~ BRAZIL 

<muiier@infra.iULbr>