JOURNAL OF THE BRAZILIAN AIR TRANSPORTATION RESEARCH SOCIETY VOLUME 10, ISSUE 1, 2014 JOURNAL OF THE BRAZILIAN AIR TRANSPORTATION RESEARCH SOCIETY
FOREWORD
SBTA is an independent scientific society. Its goals are to stimulate and to advance academic and technical research applied to the air transportation industry in Brazil. Formally established in 2002, it is a nonprofit organization affiliated to the Air Transport Research Society (ATRS).
This issue of the SBTA Journal contains four articles. The four papers were selected from the many high-quality original works submitted by several researchers and professionals from the industry, the government and the academia. These four articles, authored by researchers from the Brazilian Aeronautics Institute of Technology (ITA), from University of São Paulo and from YTU Civil Engineering Dept. Transportation from Turkiye discuss topics of great importance for the air transportation industry in the world. These four articles undoubtedly represent a significant contribution to the air transportation field in world.
The first paper presents an estimate of a risk of two B737 aircraft wings colliding while parking in designated positions on the apron of the Congonhas Airport (São Paulo – Brazil). The second paper conducts a safety analysis at seven Brazilian airports through the application of a method of risk analysis, which considers the occurrence of the following events: landing overrun, takeoff overrun, landing undershoot, and veer-off on landing and takeoff operations. The third paper aims at simulating the long-term demand impacts of improving ground access to a secondary airport with low cost carrier operations in a multiple airport region. It’s analyzed the São Paulo multi airport system – Guarulhos, Congonhas and Viracopos – under a high-speed railway (TAV) scenario. The fourth paper aims to determine the relationship between occupancy times and level of service (LOS) at baggage claim area and offer a new gate assignment policy.
João Batista Camargo Jr. Editor-In-Chief
Guidelines for Submission of Manuscripts to the Journal of the Brazilian Air Transportation Research Society
This document summarizes the basic instructions to submit your article for publication in the Journal of the Brazilian Air Transportation Research Society.
Please go to http://www.sbta.org.br/jbats/ for details on manuscript preparation. On-line submission available at: http://www.sbta.org.br/jbats/
1) Guidelines for submission
a) Manuscript style
The Authors' full name and affiliation must appear on the title page under the article’s title. The authors’ affiliation, full address, telephone, email and fax number must also be on this page. The second page must contain the article’s title, an abstract of about 100 words, and 5 key words. Articles must be in English, 15-25 pages in length, paper size A4, letter size 12 – Times New
Roman, space between lines 1.0 and format Word, PS or PDF.
Authors must send all figures and graphics in separate files. The figures have to be sent in high-resolution 300 dpi and gray scale files. They must be numbered (Arabic) and accompanied by the corresponding captions. All figures and tables must be referenced in the text.
b) Recommendations and reference style
Articles should be as concise as possible and clearly present the abstract, introduction, objectives, justification and/or motivation, definition of the problem, methodology, analysis, results and
conclusion, plus references;
Authors should use the author-date system for text references and include the relevant page number if applicable.
A list of references is required. It must contain only the works cited in the text, in alphabetical order by author. This should include the author's name, initials, year of publication in round brackets, the place of publication and publisher. For journal articles, the journal title, volume number, either month or issue number and page numbers are required.
Copyright will belong to the Brazilian Air Transportation Research Society.
2) Article review
Your article will be reviewed by two members of the journal’s editorial board through a double-blind process.
A copy of the referees' comments will be sent to the authors for revision. For publication, the revised article must follow the guidelines above and must address the referee’s comments. Authors must state that their work has not been submitted or published elsewhere.
Yearly publication produced by SBTA – Sociedade Brasileira de Pesquisa em Transporte Aéreo EDITOR-IN-CHIEF
João Batista Camargo Jr. School of Engineering University of São Paulo - Brazil
EDITORIAL BOARD
Alexandre Gomes de Barros University of Calgary – Canada Elton Fernandes
Federal University of Rio de Janeiro – Brazil Felix Mora-Camino
ENAC – France Li Weigang
University of Brasília – Brazil
THE BRAZILIAN AIR TRANSPORTATION RESEARCH SOCIETY
Erivelton Pires Guedes - Ipea President
Li Weigang – University of Brasília Vice-president
Ítalo Romani– GE Counselor
João Batista Camargo Jr – POLI-USP Counselor
Ricardo A. V. Gimenes – POLI-USP Counselor
Dorieldo Luiz dos Prazeres - ANAC Member of Fiscal Board
Érico Santana Érico Santana – CCR SBTA List Moderator
PRODUCTION COORDINATION Ricardo Alexandre Veiga Gimenes Editorial Coordinator
Fernando de Oliveira Gil Cover Design
Alan Cleber Borim Cover Edition
Maria Cristina Vidal Borba Text Revision
Ana Maria de Castro Badiali Bibliographic Normalization
Papers
ANALYSIS OF WINGS COLLISION RISKS DURING THE OPERATION
OF B737W AIRCRAFT AT THE CONGONHAS AIRPORT APRON / SP -
BRAZIL
...pg 09Oswaldo Sansone Rodrigues Filho
Civil Engineering School of Mauá Institute of Technology (Instituto Mauá de Tecnologia) Planway Engineering, Architecture and Consulting LTDA.
Linda Lee Ho
José Alberto Quintanilha
Dept. of Transport Engineering – Polytechnic School of the University of São Paulo (Escola Politécnica da Universidade de São Paulo)
SAFETY EVALUATION OF RUNWAY SAFETY AREAS: CASE STUDY
AT MAJOR BRAZILIAN AIRPORTS
...pg 25Anderson Ribeiro Correia Júlio Alves Ribeiro Neto Aeronautics Institute of Technology
LONG-TERM DEMAND IMPACTS OF ENHANCED GROUND ACCESS
TO A SECONDARY AIRPORT WITH LOW COST CARRIER
OPERATIONS
...pg 37Renan Rios Diniz
Mayara Condé Rocha Murça Alessandro V. M. Oliveira Aeronautics Institute of Technology
A NEW GATE ASSIGNMENT POLICY AND THE EFFECT OF GATE
OCCUPANCY TIME ON THE LEVEL OF SERVICE AT AIRPORT
TERMINALS
...pg 53Dr. Güzin Akyildiz Alçura Dr. Mustafa Gürsoy
ANALYSIS OF WINGS COLLISION RISKS DURING THE OPERATION
OF B737W AIRCRAFT AT THE CONGONHAS AIRPORT APRON / SP -
BRAZIL
Oswaldo Sansone Rodrigues Filho
Civil Engineering School of Mauá Institute of Technology (Instituto Mauá de Tecnologia); Planway Engineering, Architecture and Consulting LTDA.
Linda Lee Ho
Dept. of Production Engineering – Polytechnic School of the University of São Paulo (Escola Politécnica da Universidade de São Paulo)
José Alberto Quintanilha
Dept. of Transport Engineering – Polytechnic School of the University of São Paulo (Escola Politécnica da Universidade de São Paulo) – 55 11 30915174 – [email protected]
Received: February 26, 2014 Accepted: August 22, 2014
ABSTRACT
This article presents an estimate of a risk of two B737 aircraft wings colliding while parking in designated positions on the apron of the Congonhas Airport (São Paulo – Brazil). The movement of the aircraft around the apron was observed and analyzed during docking. The final parking position when cleared for the boarding and disembarking of passengers and luggage was monitored both at the boarding bridges and at remote locations. Statistical analysis was performed, and the results reveal the feasibility of parking these aircraft under current conditions available at the airport. This study indicates that a smaller distance will not have an adverse effect on the safety of the operations.
Keywords: wings collision risk, airport operation, B737W aircraft, Congonhas Airport,
1. INTRODUCTION
The current configuration of the Congonhas Airport (São Paulo – Brazil) is a result of its development history, initiated in April of 1936, when the “Campo de Aviação da Companhia de Auto-Estradas” [“Aviation Field of the Highways Company”] as formerly nominated, received its first experimental flights.
Three transverse runways with length of 1,700; 880 and 1,040 meters were planned at the end of the 1940s. The first runway was paved at the end of 1950. Subsequently, a second provisional runway was built parallel to the first, which would later become an auxiliary runway. Studies performed in 1950 demonstrated that it was no longer necessary to build transverse runways. Thus, the airport is currently composed of two parallel runways. Over a long period, the operation at Congonhas Airport proceeded with the aircraft parking in the apron at remote positions, away from the terminal building, with the passengers transported by foot or bus.
With the expressive air traffic growth in the early 2000s the improvement of the Passenger Terminal was necessary as also the adoption of a new design for the parking apron to adequate the capacity of the installations to attend the volume of passengers and aircraft movements. The terminal building was expanded and twelve new boarding bridges were installed. Today the airport has a total of thirty parking positions for commercial aircraft, twelve of which have boarding bridges. The other eighteen positions are remote, where passengers are transported by bus.
With the implementation of the new apron (with new geometry and twelve boarding bridges) and the advent of B737 New Generation (with winglet and larger wingspan), the evaluation of the safety issues in the operation of the apron becomes necessary. This is especially true for the B737-700W and 800W as the design of new apron does not allow these aircraft to utilize
the boarding bridges, as the space between the wing tips of two aircraft parked next to one another is below the minimum safety distance set by the regulations.
The parking apron at Congonhas Airport can accommodate thirty commercial aircraft. Positions 1 through 12 are in front of the Passenger Terminal and have boarding bridges. The other 18 positions are remote, where passengers are transported by bus. Figure 1 shows the parking positions and separations. The separation between two adjacent positions [referred as aircraft stand lead-in lines by the International Civil Aviation Organization (ICAO)] is 39.22 m for positions 1 and 2. For position 2 through 11 the separation between two adjacent positions is 38.82 m, and for positions 11 and 12 the separation is 39.56 m. For the remote positions 13 through 23, the separation is 42 m between stand lead-in lines. Positions 25 to 30 are also remote with a separation of 35 m in the stopping lines. Positions 17 and 24 are at 45° in relation to the adjacent positions, without restrictions for any aircraft.
The largest aircraft in the apron maneuvering at Congonhas Airport are the Boeing 737 family and the Airbus 320, whose main physical characteristics are in Table 1. Figure 2 illustrates the geometric characteristics of the aircraft.
Currently, Congonhas Airport receives six types of aircraft: B737-700W; B737-800W; B737-700; B737-800; A320 and B737-300. The aircraft stand lead-in lines values (in meters) between the wing tips of two aircraft parked side-by-side are in Table 2. The most restrictive positions, that is, those that result for the smallest separations, are at boarding bridges number 2 to 11, with a separation between the aircraft stand lead-in lines of 38.82 m.
For safety reasons the ICAO (2005) recommends a separation of 4.5 m between the wing tips of two adjacent parking aircraft. Table 2 shows, for positions 2 through 11, that the side-by-side parking separations when B737-700W/800W and
A320-200 are parked, may result less than the 4.5 m required by the ICAO (2005). The critical parking situations are concentrated around the front of the terminal 2 to 11 and are indicated in Figure 3. The smallest distance of 3.02 m is between two B737-700W/800W aircraft.
The objective of this study is to develop a methodology for calculating the risk of wings collision between two aircraft parked in adjacent positions. Deviation measurements in the parking position of the aircraft are taken and a generalized linear model (GLM) is adjusted for the deviation measure to identify relevant explanatory variables to explain the variability of the deviation. Results of the models provide inputs to calculate the risk of two B737 aircraft wings colliding when parked at adjacent positions. The results indicate the feasibility of the operation of B737-700W/800W aircraft at any position regardless of whether the aircraft are parked on the left or the right position. This finding constitutes a contribution to the management and planning of airports.
ICAO (2005) specifies that it is permissible to use a smaller separation at an existing airport if an aeronautical study, such as this study, indicates that a smaller distance will not have an adverse effect on the safety of the operations.
Although generalized linear model (Montgomery 2005 and 1997) is not a new feature, the application in the current context is novelty in technical and academic literature related to airport safety. Moreover, this problem is common nowadays, since the disposal area to park in airports demands long time building work to change an existent infrastructure to attend the rules to accommodate new and bigger aircraft in specified different context.
To accommodate new large aircraft (NLA) operations Barros and Wirasinghe (2003) have analyzed the passenger terminal configurations. NLA are new aircraft developments larger than the Boeing 747,
performed individually for a single pier, several types of pier–satellites, and a set of remote parallel piers connected by an automated people mover (APM). In all cases, the best location for the NLA gate positions is sought, using analytical models. A design of experiments approach has been employed by Buxi and Hansen (2011) to determine profiles that minimize the total average costs. The average cost of the methodologies is evaluated against realized capacities to determine the benefit of the weather forecast, since the weather is a role in determining the capacity of an airport. Borille and Correia (2012) consider the explanatory variables as demand characteristics, terminal layout, the number and type of carousels, waiting time and space available to study their influence in the level of service of the operational arrival components at airports. The analysis combines user monitoring techniques, data collection, simulation models, design of experiments and linear regression. Five major international airports in Brazil are used as case studies. Atasoy et al. (2013) provide analytical evidence of the impact of a new innovative modular aircraft on the operations of an airline. The impact analysis is carried out with an integrated scheduled planning model that presents a combination of appropriate optimization and behavioral modeling methodologies. An experimental design was meant to minimize the impacts of the differences in size and to reveal to a larger extent the impact of modularity. Chiang (2011) used two-level full factorial design of experiments (DOE) to simulate the different scenarios to identify relevant factors and their interactions considered in the research to measure its impact on passenger corridor occupancy.
Other uses of statistical analysis in different problems related to airport operations are discussed in Xianfenga and Shengguoa (2012). Contributions related to airport operations can be mentioned: Ale and Piers (2000), Lee (2006), Yun (2003), Kai (2006), Gang and Jin-fu (2008 a, b), Gui-mei and Sheng-guo (2010). Hang and Gui-hong
(2009) investigated airport safety and used different data analysis methodologies (groupings, scores, fuzzy logic, AHP, etc.) to assess the risk of different events related to aviation and, in particular, to airports. A single study on the “length and reference codes of runways and taxiways; runway and taxiway strips; runway end safety areas; separation distances between runways and taxiways; and definition of obstacle limitation surfaces” was developed by Eddowes et al. (2001).
This paper is organized as follows: this introduction section; section 2, the elements
of the full database are described; the exploratory data analysis and generalized linear model are subject of section 3; the procedures for the determining the risk of two aircraft wings colliding when they are simultaneously boarding in neighboring boxes is the subject of the section 4; section 5 shows the final conclusions. All the technical terminologies related to the airport operations can be found in International Civil Aviation Organization – ICAO – (2013).
Table 1: Physical characteristics of critical aircraft (Boeing, 2001; Airbus, 2005)
Physical characteristics (m)/Aircraft B737-800W B737-800 A320-300
Wing span (P) 35.80 34.32 34.09
Length (L) 39.48 39.48 37.54
Main Gear Wheel Span (M) 7.00 7.00 8.97
Figure 1: Parking positions and separations at the Congonhas Airport – São Paulo - Brazil Source: SBTA's Homepage (http://www.sbta.org.br)
Figure 2: Geometric characteristics of the aircraft
Table 2: Wing tip separation (in meters) parking of the aircraft combinations at the Congonhas Airport for the
boarding bridges 2 to 11. Aircraft B737-700W B737-800W B737-700 B737-800 A320 B737-300 B737-700W 3.02 B737-800W 3.02 3.02 B737-700 3.76 3.76 4.50 B737-800 3.76 3.76 4.50 4.50 A320 3.88 3.88 4.62 4.62 B737-300 6.48 6.48 7.22 7.22 7.34 9.94
2. DATABASE
For the determination of the effective aircraft position in relation to the aircraft stand lead-in lines (painted on ground), three measurements are taken at different positions in the landing gear. These measurements determine accurately the position of the aircraft in relation to the stopping line painted on pavement, as well as the resulting separation from the other neighboring aircraft:
a) The deviation or relative distance from the axis of the nose gear and that of the stopping T position in the longitudinal direction to the movement;
b) The deviation or distance from the axis of the nose gear and that of the aircraft stand lead-in line position in the direction transverse to the movement;
c) The distance from the axis of one leg of the main landing gear and the aircraft stand lead-in line position.
The scheme provided in Figure 4 depicts the location of the measurements. Using these measurements, the deviation between the axis of the main landing gear and the parking position axis, and the deviations, longitudinal and transversal, of the nose gear are determined.
The deviation measurements are collected in seven positions with boarding bridge (positions from 6 to 12) and in eight remote positions (position from 15 to 23, except the position 17) around two weeks (from the last week of October to the first week of November in 2008). Due to the operational restrictions, the deviations are taken only for the aircraft 300, 700 and B737-800 totaling 196 observations.
Two measurements are obtained to determine the location of the nose gear. One is the deviation or distance from the landing gear to the stopping line (longitudinal direction of the aircraft). In addition to this measurement, the position of the nose gear
before (-) and after (+) the T-stopping line is also noted.
Deviations in relation to the “T” after the determination of the painted line are obtained with the metal bracket positioned on the axis of the nose gear. This position is marked on the ground and the difference between the mark and the painted line is recorded.
Deviations of the nose gear in relation to the aircraft stand lead-in line are obtained in the following manner:
- Determination of the center axis of the nose gear with the laser device;
- Recording of the nose gear axis position on the ground;
- Measurement of the distance between the nose gear axis position and the painted line. Both deviations are recorded to the left and to the right in the longitudinal direction. And additionally deviations in the main landing gear with respect to the aircraft stand lead-in line are obtained using the location of the painted line in the following manner: - Determination of the axis for a set of wheels in the main landing gear using a laser device
- Recording of the position of the main landing gear axis position on the ground; - Determination of the line perpendicular to the aircraft stand lead-in line position up to the axis of the main landing gear projection using an aluminum ruler;
- Measurement of the distance between the position of the axis of the main landing gear and the painted line.
Although many measurements are in database, in this paper the statistical analysis concerns on the deviations in the main landing gear with respect to the aircraft stand lead-in line as the risk of wings collisions depends on such deviation and this is the subject of the next sections.
Figure 4: The location of measurements
3. DATA ANALYSIS
In this section, the statistical data analysis is presented. Notations used hereon are introduced. Let be
D= the distance from the main landing gear to the aircraft stand lead-in line.
The explanatory variables considered to the statistical data analysis to explain the variability of the deviation measurement D are:
- Types of parked aircraft (“aircraft”): B737-300, B737-700 or B737-800; - Type of the localization (“type”)
boarding bridge or remote;
- The number of the position (“box”): the box number 6 to 12 if the position is a boarding bridge or the box number 15 to 23, except 17 if the position is remote; - The driven direction to put the aircraft
into the box (“direction”): the aircraft turns to the left or to the right hand. Table 3 presents some descriptive statistics of D (average, standard deviation - SD and median) by each of the explanatory variables. Tables 4-7 present the same
descriptive statistics but concerning to two explanatory variables at each time, specifically: Table 4: type x direction; Table 5: type x aircraft; Table 6: aircraft x direction; Table 7: type x box number. Note that the average values of D in any table are negative, representing an average deviation to the left side of the reference line. Analyzing the tables, some interesting observations can be noted. From Table 3, (absolute) larger values of D are observed for aircraft B737-300 and 800 but when analyzed together with type of position, the behavior differs a lot (see Table 5). For the boarding bridges, the larger (absolute) average differences are observed for the B737-300 aircraft and the smaller differences for the B737-800 aircraft but for the remote locations, the opposite trend is observed, that is, smaller differences for the B737-300 aircraft and largest differences for the B737-800 aircraft. Such fact suggests that an interaction effect of the type of aircraft and type of parked position is active. From Table 7, the average deviations depend on the box number. For the boarding bridges, the boxes located in the center position (9 and 10) exhibit the greater level of deviation. The same observation is valid
for the remotely located boxes, where the deviations are higher for the center boxes. These results demonstrate that the boxes cannot be considered homogeneous (in sense of equal behavior relative to distances). Note also the lowest average deviations are observed for the last box of each type of position, but also largest standard deviations
(box number 12 and 23, respectively for boarding bridge and remote position).
In order to identify which explanatory variables are more significant to explain the variability of D, a generalized linear model is proposed. The k-th observation of D can be expressed as:
(1) with
β= ( )
the vector of the parameters to be estimated: , a constant; , are the effects related to the different types of aircraft; is related to the direction which the aircraft takes to be parked in the box; , to the type of parked position (boarding bridge/remote); , the effects of the position of boxes at boarding bridges;
, the effects of the position of boxes when the aircraft parked in remote position; the interaction effects of the types of aircraft and type of parked position.
ek is the random error and assumes that follows a normal distribution (0; σ2). The explanatory variables used in these analyses assume the following values:
X11=1; X12=0 for the aircraft B737-300;
X11=0; X12=1 for the aircraft B737-700;
X11=-1; X12=-1 for the aircraft B737-800;
X2=1 for the aircraft parked in the box by
the right hand direction;
X2=-1 for the aircraft parked in the box by
the left hand direction;
X3=1 for the aircraft parked in boarding
bridges;
X3=-1 for the aircraft parked in remote
position.
As the number of the box is nested of type of parked position, so for the boarding bridges (X3=1) the explanatory variables X4j assume
the following values:
X41=1; X4j=0; 2≤ j ≤6 for box #6; X42=1; X4j=0; j ≠2; 1≤ j ≤6 for box #7; X43=1; X4j=0; j ≠3; 1≤ j ≤6 for box #8; X44=1; X4j=0; j ≠4; 1≤ j ≤6 for box #9; X45=1; X4j=0; j ≠5; 1≤ j ≤6 for box #10; X46=1; X4j=0; j ≠6; 1≤ j ≤6 for box #11; X4j=-1; 1≤ j ≤6 for box #12;
And for the remote positions (that is, X3= -1)
the values of variables X4j are:
X41=1 and X4j=0; 2≤ j ≤7 for box #15;
X42=1 and X4j=0; for j ≠2; 1≤ j ≤7 for box #16;
X43=1 and X4j=0; for j ≠3; 1≤ j ≤7 for box #18;
X45=1 and X4j=0; for j ≠5; 1≤ j ≤7 for box #20;
X46=1 and X4j=0; for j ≠6; 1≤ j ≤7 for box #21;
X47=1 and X4j=0; for j ≠7; 1≤ j ≤7 for box #22; and X4j=-1; 1≤ j ≤7 for box #23.
The null hypothesis (the
deviations D for the different types of aircraft are equal) is rejected as it yields a p-value of 0.05. The null hypothesis
(the deviations D of the aircraft parked in the box by the right hand and left are equal) is also rejected (p-value of 0.023). But the null hypothesis (the average deviations D of the aircraft parked at boarding bridges and remote position are equal) is not rejected (p-value of 0.23). This explanatory variable: type of localization (X3) is not active alone and usually it would
be discarded from the model. But in this case, it interacts with other explanatory variables producing active effects. Note that
the box number (X4i) depends on the type
position (X3) (if the aircraft is parked at
boarding bridges, the box numbers go from 6 to 12; if parked in the remote terminals, the box number are from 15 to 23, except for 17) and its respective coefficients are significant with a p-value of 0.002 by the rejection of
the null hypotheses:
Additionally, (null
interaction of the types of aircraft: X1 and
type of parked localization X3: boarding
bridge and remote position) is not true (p-value of 0.001). Due to these reasons, the explanatory variable X3 is kept in the final
model. Estimates of the coefficients of the model 1 are obtained by Minitab Statistical Software and put in Table 8.
Table 3: Descriptive Statistics of D by aircraft, type and direction Average SD Median
Aircraft B737-300 -32.48 20.43 -36.30
B737-700 -9.75 17.57 -11.00 B737-800 -26.87 30.03 -25.00 Type Boarding bridge -14.05 18.38 -13.90 Remote -17.41 28.40 -12.90
Direction Right -13.53 24.56 -9.40
Left -18.87 17.75 -18.90 Table 4: Descriptive Statistics of D by type and direction
Type Direction Average SD Median Boarding bridge Right -10,70 17.67 -10.00
Left -17.92 18.58 -16.30 Remote Right -16.41 29.86 -9.10
Table 5: Descriptive Statistics of D by aircraft and type
Type Aircraft Average SD Median Boarding bridge B737-300 -35.60 15.36 -36.30 B737-700 -11.55 16.11 -12.20 B737-800 -2.63 25.74 2.05 Remote B737-300 -9.10 46.24 -9.10 B737-700 -4.64 20.58 -3.90 B737-800 -31.02 29.01 -30.00
Table 6: Descriptive Statistics of D by aircraft and direction
Direction Aircraft Average SD Median Right B737-300 -29.23 23.64 -30.45 B737-700 -5.38 16.66 -5.30 B737-800 -27.21 31.00 -26.65 Left B737-300 -37.13 15.23 -36.30 B737-700 -16.34 16.99 -15.90 B737-800 -22.53 14.94 -23.80
Table 7: Descriptive Statistics of D by aircraft and type and box number
Type Box Number Average SD Median Boarding 6 -14.66 15.16 -12.80 bridge 7 -6.89 23.55 -7.40 8 -12.47 13.37 -11.70 9 -20.78 18.81 -16.00 10 -21.86 16.84 -19.55 11 -16.95 11.94 -15.90 12 -2.25 20.09 4.80 Remote 15 -24.14 14.71 -29.70 16 -15.84 12.49 -19.50 18 -31.35 34.37 -34.50 19 -16.12 36.04 -9.60 20 -16.74 28.93 -10.80 21 -7.07 14.86 -11.20 22 -13.26 31.47 -12.90 23 -2.22 29.03 8.70
Table 8: Descriptive Statistics of D by aircraft and direction
Direction Aircraft Average SD Median Right B737-300 -29.23 23.64 -30.45 B737-700 -5.38 16.66 -5.30 B737-800 -27.21 31.00 -26.65
The goodness of fit of the model (1) can be confirmed by residual analysis. Figure 5 shows four residuals plots. By these plots, the standardized residuals follow a normal
distribution as also a random pattern, showing no tendencies and few unusual residuals.
4. DETERMINATION OF RISK OF WINGS COLLISIONS
The procedures for the determination of the risk of two aircraft wings colliding when they are simultaneously boarding in neighboring boxes are the subject of this section. We are concerning to determine the risk only for box numbers with access by boarding bridges. For box number at the remote positions similar procedure can be adopted with few adjustments.
The maximum absolute allowable distance (W) is defined for situations when wings collision would be possible, using 38.8 m (this value is the most common among the positions with boarding bridges- boxes 2 through 11) as the distance between positions and the wing spans of each type of aircraft. It consists of two portions: D (the distance from the main landing gear to the aircraft stand lead-in line) and a clearance
value between the wing tips, depending on the type of aircraft. Considering two neighboring boxes, a wings collision will occur if simultaneously the aircraft in the box on the left moves to the right more than D + (19400-0.5 x 100P) cm (values of P, for the different types of aircraft, in Table 1) and the aircraft in the box on the right moves to the left more than (19400-0.5 x 100P) - D cm.
Figure 7 illustrates an example for two B737-300 aircraft. A wings collision will occur if simultaneously the aircraft in the box on the left moves to the right more than D + 500 cm and the aircraft in the box on the right moves to the left more than 500 - D cm. For two neighboring boxes with B737-700/800 aircraft, a wings collision will occur if the aircraft on the left moves to the right more than D + 224 cm and the aircraft on the right moves to the left a more than 224 - D cm.
Table 9: Estimates of the coefficient of the model (1)
Coefficient Estimate SE T P-value -16.039 3.133 -5.12 0 0.505 5.373 0.09 0.925 8.347 3.212 2.6 0.01 -3.711 1.616 -2.3 0.023 -3.697 3.07 -1.2 0.23 -1.962 5.37 -0.37 0.715 10.133 4.326 2.34 0.02 1.863 3.931 0.47 0.636 -7.799 4.092 -1.91 0.058 -7.516 3.861 -1.95 0.053 -3.292 4.277 -0.77 0.443 -20.803 8.463 -2.46 0.015 -7.295 6.583 -1.11 0.269 -14.303 5.046 -2.83 0.005 4.846 5.262 0.92 0.358 7.488 5.225 1.43 0.154 4.62 6.036 0.77 0.445 13.785 7.94 1.74 0.084 -16.578 5.38 -3.08 0.002 0.434 3.201 0.14 0.892 σ2 356.7
And similarly, for two neighboring boxes with B737-700W/800W aircraft, a wings collision will occur if the aircraft parked on the left moves to the right more than D+150
cm and the aircraft parked on the right moves to the left more than 150-D cm. Other combinations of two neighboring boxes can be considered in a similar manner.
Figure 6: Maximum allowable distance for B737-300 aircraft
The predicted values (PD) of D as a function of box number, aircraft and direction considering the coefficients are in Table 9. Consider two neighboring boxes: i and (i+1). A possible wings collision will occur if simultaneously the aircraft in box i parks at a distance greater than (PD+19400-50P) cm (displacement to the right hand) and the aircraft in box i+1 parks at a distance greater than (19400-50P-D) cm (displacement to the left hand).
For example, consider two neighboring boxes 6 and 7 with both B737-300 aircraft parked from the right-hand. The B737-300 aircraft in box 6 parks at an average distance of 34.06 cm from the left of the center of the box (hence the negative value in Table 8), whereas the B737-300 aircraft in box 7 parks at an average distance of 21.97 cm to the left of the center of the box (also negative value). If simultaneously the aircraft in box 6 parks at a distance greater than 534.06 cm
(displacement to the right hand) and the aircraft in box 7 parks at a distance greater than 478.04 cm (displacement to the left hand) there is a situation of a possible wings collision.
To calculate the probability, average value of the normal distribution equal to PD values in Table 9 is used and the standard deviation as 18.5565 (see Table 8). For this example, it is equal to P (Z > 28.28) x P (Z < -25.31) (see Table 9) which probabilities are respectively put in the last two columns of Table 9. Or alternatively if simultaneously the aircraft in box 6 parks at a distance greater than 465.94 cm (displacement to the left hand) and the aircraft in box 7 parks at a distance greater than 521.97 cm (displacement to the right hand). Other combinations of aircraft, direction and neighbored boxes can be evaluated in the same manner.
Table 9: Values obtained from the Adopted Model Maximum Distance
(cm)
Standardized maximum
distance (Z) Probability “position” Aircraft Box PD Right Left Right Left P(Z>z) Right P(Z<-z) Left
Rig h t 300 6 -34.06 534.06 465.94 28.28 24.67 0.00E+00 1.11E-134 7 -21.97 521.97 478.04 27.64 25.31 0.00E+00 1.21E-141 8 -30.24 530.24 469.77 28.07 24.87 0.00E+00 7.28E-137 9 -39.90 539.90 460.10 28.59 24.36 0.00E+00 2.19E-131 10 -39.61 539.61 460.39 28.57 24.38 0.00E+00 1.52E-131 11 -35.39 535.39 464.61 28.35 24.60 0.00E+00 6.30E-134 12 -23.53 523.53 476.48 27.72 25.23 0.00E+00 9.79E-141 Rig h t 700 6 -9.21 233.21 214.79 12.35 11.37 0.00E+00 2.85E-30 7 2.89 221.11 226.89 11.71 12.01 0.00E+00 1.51E-33 8 -5.38 229.38 218.62 12.15 11.58 0.00E+00 2.75E-31 9 -15.04 239.04 208.96 12.66 11.06 0.00E+00 9.40E-29 10 -14.76 238.76 209.24 12.64 11.08 0.00E+00 7.95E-29 11 -10.54 234.54 213.46 12.42 11.30 0.00E+00 6.38E-30 12 1.33 222.67 225.33 11.79 11.93 0.00E+00 4.09E-33 Rig h t 800 6 -10.70 160.7 139.31 8.6600 7.5073 2.07E-17 3.02E-14 7 1.40 148.6 151.4 8.0080 8.1589 4.74E-15 1.40E-15 8 -6.87 156.87 143.13 8.4536 7.7132 1.21E-16 6.11E-15 9 -16.53 166.53 133.47 8.9742 7.1926 1.30E-18 3.18E-13 10 -16.25 166.25 133.75 8.9591 7.2077 1.48E-18 2.84E-13 11 -12.03 162.03 137.98 8.7317 7.4357 1.11E-17 5.21E-14 12 -0.16 150.16 149.84 8.0920 8.0748 2.41E-15 2.77E-15 Left 300 6 -41.48 541.48 458.52 28.67 24.28 0.00E+00 1.69E-130 7 -29.39 529.39 470.61 28.03 24.92 0.00E+00 2.38E-137 8 -37.66 537.66 462.34 28.47 24.48 0.00E+00 1.20E-132 9 -47.32 547.32 452.68 28.98 23.97 0.00E+00 2.96E-127 10 -47.04 547.04 452.96 28.96 23.98 0.00E+00 2.07E-127 11 -42.81 542.81 457.19 28.74 24.21 0.00E+00 9.36E-130 12 -30.95 530.95 469.05 28.11 24.84 0.00E+00 1.86E-136 Left 700 6 -16.63 240.63 207.37 12.74 10.98 0.00E+00 2.39E-28 7 -4.53 228.53 219.47 12.10 11.62 0.00E+00 1.62E-31 8 -12.80 236.80 211.20 12.54 11.18 0.00E+00 2.49E-29 9 -22.47 246.47 201.54 13.05 10.67 0.00E+00 6.97E-27 10 -22.18 246.18 201.82 13.03 10.69 0.00E+00 5.93E-27 11 -17.96 241.96 206.04 12.81 10.91 0.00E+00 5.19E-28 12 -6.09 230.09 217.91 12.18 11.54 0.00E+00 4.26E-31 Left 800 6 -18.12 168.12 131.88 9.0599 7.1069 5.99E-19 5.93E-13 7 -6.02 156.02 143.98 8.4078 7.7590 1.78E-16 4.33E-15 8 -14.29 164.29 135.71 8.8535 7.3133 3.80E-18 1.30E-13 9 -23.95 173.95 126.05 9.3741 6.7928 3.31E-20 5.50E-12 10 -23.67 173.67 126.33 9.3590 6.8079 3.81E-20 4.95E-12 11 -19.45 169.45 130.55 9.1316 7.0353 3.12E-19 9.94E-13 12 -7.58 157.58 142.42 8.4919 7.6749 8.75E-17 8.33E-15 5. CONCLUSION
The statistical analysis conducted to evaluate the risk of two B737 aircraft wings colliding when parked at adjacent positions at the Congonhas Airport indicates the feasibility of the operation of B737-700W/800W aircraft at any position regardless of whether the aircraft are parked on the left or the right-hand. ICAO (2005) specifies that it is permissible to use a smaller separation at an
existing airport if an aeronautical study, such as this study, indicates that a smaller distance will not have an adverse effect on the safety of the operations.
The expected outcome of this work is the development of similar studies that will be performed under different conditions to improve the existing airport legislation and allow for better use of existing facilities. Most of the assumptions and hypothesis used
in the design of experiments are detailed constituting a valuable reference for flexible transportation systems to be designed in the future.
ACKNOWLEDGMENTS
The authors would like to thanks to Gol Companhia Aérea, CNPq (Conselho Nacional de Desenvolvimento Científico e Tecnológico – National Council for Scientific and Technological Development) and EPUSP (Escola Politécnica da Universidade de São Paulo – Engineering School of the University of São Paulo).
REFERENCES
Atasoy, B.; Salani, M.; Bierlaire, M.; Leonardi, C. (2013) Impact Analysis of a Flexible Air Transportation System. EJTIR 13(2), pp. 123-146 128.
Airbus (2005) A320: Aircraft Characteristics Airport and Maintenance Planning.
Ale, B. J. M. and M. Piers (2000) The Assessment and Management of Third Party Risk Around a Major Airport. Journal of Hazardous Materials, v. 71, n. 1-3, p. 1-16. Barros A. G., Wirasinghe, S.C. (2003) Optimal terminal configurations for new large aircraft operations. Transportation Research Part A 37, 315–331.
Boeing (2001) 737-600/700/800/900 Airplane Characteristics for Airport Planning.
Borille, G.M.R.; Correia, A. R. (2013) A method for evaluating the level of service arrival components at airports. Journal of Air Transport Management 27, 5-10. Buxi, G.; Hansen, M. (2011) Generating Probabilistic Capacity Profiles from weather forecast: A design-of-experiment approach. Ninth USA/Europe Air Traffic Management Research and Development Seminar (ATM2011)
Chiang. P-N (2011) A decision support system for assessing conveyance options and modeling passenger flow in airport terminals. A Dissertation presented to the Graduate School of Clemson University. All Dissertations. Paper 709. Eddowes, M.; J. Hancox and A. MacInnes (2001) Final Report on the Risk Analysis in Support of Aerodrome Design Rules. A report produced for the Norwegian Civil Aviation Authority. File reference RD02325. Report number AEAT/RAIR/RD02325/R/002.Report status Issue 1.
ICAO (2004) Annex 14: Aerodromes. Volume I: Aerodrome Design and Operations (4th ed).
ICAO (2005) Doc 9157-AN/901 Aerodrome Design Manual - Part 2: Taxiways, Aprons and Holding Bays (4th ed.).
Gang, L. and Z. Jin-fu (2008a) A Grey Hierarchy Risk Evaluation for Civil Airport Operation Security and Safety. Chinese Journal of Ergonomics, v. 14, n. 3, p. 1-4.
Gang, L. and Z. Jin-fu (2008b) Based on FTA and Dynamic Gray Correlation Method Analysis for Apron Human Factor Error, Chinese Journal of Ergonomics, v. 14, n. 2, p. 22-25. Gui-mei, X. and H. Sheng-guo (2010) Runway Incursion Risk Assessment Model Based on HRA. Science Technology and Engineering, v. 10, n. 19, p. 4715-4719. Hang, L. and Z. Gui-hong (2009) Early Warning Model of Airport Safety Based on Extension Theory. Computer Engineering and Applications, v. 45, n. 14, p. 238-244. 2009.
International Civil Aviation Organization – ICAO – (2013) Annex 14, Aerodromes — Volume I, Aerodrome Design and Operation. Sixth Edition. July 2013.
Kai, G. (2006) Research on a Forewarning Management Method for Civil Aviation Airport Disaster. Wuhan: College of Management, Wuhan. University of Technology. Lee, W-K. (2006) Risk Assessment Modeling in Aviation Safety Management. Journal of Air Transport Management, v. 12, n. 5, p. 267-273.
Montgomery, D. C. Design and analysis of experiments. 6a edition. Arizona: John Wiley & Sons, Inc., 2005.
Montgomery, D. C. Introduction to Statistical Quality Control, John Wiley & Sons, New York, 1997.
Xianfenga, L. and H. Shengguoa (2012) Airport Safety Risk Evaluation Based on Modification of Quantitative Safety Management Model. Procedia Engineering, v. 43, n. 2012, p. 238–244. Special Issue from International Symposium on Safety Science and Engineering in China, 2012 (ISSSE-2012).
Yun, P. (2003) Research on Safety Forewarning Management System for Civil Aviation Airport. Wuhan: College of Management, Wuhan University of Technology.
SAFETY EVALUATION OF RUNWAY SAFETY AREAS: CASE STUDY
AT MAJOR BRAZILIAN AIRPORTS
Anderson Ribeiro Correia
Aeronautics Institute of Technology
Praça Marechal Eduardo Gomes 50 – São José dos Campos/SP, +55 (12) 3947-6828, [email protected]
Júlio Alves Ribeiro Neto
Aeronautics Institute of Technology
Praça Marechal Eduardo Gomes 50 – São José dos Campos/SP, +55 (61) 8287-8383, [email protected]
Received: April 07, 2014 Accepted: August 22, 2014
ABSTRACT
This study conducts a safety analysis at seven Brazilian airports through the application of a method of risk analysis, which considers the occurrence of the following events: landing overrun, takeoff overrun, landing undershoot, and veer-off on landing and takeoff operations; these types of accidents accounts for most accidents that occur on or in the immediate vicinity of the runway. The models employed are based on evidence from worldwide accidents and incidents that occurred during the past 27 years and evaluate the likelihood and severity of event types due to engineering non-compliances at airports. For the purposes of the research provided in this paper, data was collected at seven major Brazilian airports, in order to obtain information on meteorology, operations, airport layout, and to identify main non-compliances that might compromise safety. Analysis for a given day and a critical design aircraft indicate that 10.84% of non-compliances in these airports should not be tolerable, demanding actions to solve these issues; 45.78% are tolerable, which risks could be managed with mitigating actions; 43.37% are acceptable, considering a given level of safety. Additionally, sensibility analysis was done to identify and quantify the impact of several variables on risk. The case-study and its results are useful for engineers and planners, in order to assess risk and to support planning and engineering decisions.
1. INTRODUCTION
Historical data collected between the years 1995 and 2012 show that 82% of accidents and incidents in the world with jet aircraft operations occurred during takeoff and landing; additionally, 50% of the occurrences on board and a 57% of fatalities in air transport is given in situations of takeoff and landing (Boeing, 2013). In such operations, the events with the aircraft are often related to the occurrence of phenomena Landing Undershoot (LDUS), which is set when the aircraft performs the landing before the end of the runway in operation, Landing Overrun (LDOR)¸ which happens when the aircraft lands after the end of the runway in operation, and Takeoff Overrun (TOOR), characterized when the plane can not successfully take off before the end of the runway length (Caves , 1996). In addition to the events mentioned above, there is also the occurrence of Veer Off (VOFF), which corresponds to the lateral deviation of the aircraft with respect to the axis of the runway and might eventually happen simultaneously with the phenomena mentioned above, resulting in Landing Veer Off (LDVOFF) or Takeoff Veer Off (TOVOFF). Figures 1 and 2 show the main indicators of incidents in these categories at the international level (Ayres et al., 2011).
Figure 2: Ocurrences by Event Type (Ayres et al., 2011)
Figure 3: Distribution by Event Type (Ayres et al., 2011)
In order to reduce the risk of occurrence of events in the three categories of events mentioned earlier, it is necessary to quantify the risk associated with takeoffs and landings, especially when there are obstacles located in safety areas. It is usual to consider an acceptable level of safety of 10-7 (one occurrence for every 10 million operations), as recommended by Annex 14 of ICAO (International Civil Aviation Organization), in order to evaluate risk (DAC, 2004). Besides evaluating risk of occurrence, it is convenient to evaluate probabilities of final location of aircrafts, in case of accidents on the former categories of events (Eddowes et al., 2001); these probability locations are also useful to estimate the necessary safety areas (Hall et al., 2008).
This study intends to perform the verification of the risks of accidents at seven major Brazilian airports, analyzing the probability of such events in Brazilian airport sites through the application of a quantitative method, as well as recommendations issued by ICAO - International Civil Aviation Organization.
2. METHODOLOGY
The method used for assessing the risk of accidents in this case study is the result of a research effort sponsored by the Transportation Research Board (TRB) and the Federal Aviation Administration (FAA), under the agreement ACRP - Airport Cooperative Research Program. In this concern, two studies relating to operational safety at airports have been developed, as it is among the few developed methods that evaluate data from aircraft accidents
worldwide. Project reports were entitled: "Analysis of Aircraft Overruns and Undershoots for Runway Safety - ACRP - REPORT 03" (Hall et al, 2008) and "Improved Models for Risk Assessment of Runway Safety Areas - ACRP - REPORT 50" (Ayres et al., 2011).
For the implementation of such methods of risk analysis, data was collected to obtain the characteristics of takeoff and landing operations at the case study airports, such as weather information obtained from data provided by the Meteorological Network of the Brazilian Air Force (REDEMET). The recommendations and guidelines of ICAO for the risk management of civil aviation (ICAO, 2009) were used to assess the severity of the consequences of a possible accident involving an aircraft in landing or takeoff operation. This methodology qualitatively defines the severity of certain events in order to identify non-compliances that result in the reduction of operational safety as well as guide the actions of such mitigating consequences of accidents of operations at airport sites.
The determination of the probability of occurrence of the events is based on a probabilistic model, which is dependent from a number of several variables; it quantifies the occurrence of a certain event type. From the occurrence of a certain event, it becomes necessary to evaluate the location of the aircraft beyond the limits for landing or takeoff. Thus, we use a probabilistic model that determines (for each airport analyzed) a probability distribution for the location of the aircraft at the airport site. Figures 3-5 present the concept used for modeling the events undershoot, overrun and veer-off.
Figure 3: Modeling concept for aircraft Overruns (Hall et al., 2008)
Figure 3: Modeling concept for aircraft Overruns (Hall et al., 2008)
Figure 5: Modeling concept for aircraft Veer-offs (Ayres et al., 2011)
The basic model structure selected is
presented at Equation 1 (Ayres et al., 2011):
.... 1 0 1 1 2 2 3 3 1 1 } _ { b bX bX bX e Ocurrence Accident P (1) where P{Accident_Occurrence} is probability (0–100%) of an accident type occurring given certain operational conditions; Xi are independent variables
(e.g., ceiling, visibility, crosswind, precipitation, aircraft type, criticality factor); and bi are regression coefficients.
Concerning the location probabilities, the conceptual model for longitudinal distribution is given at Equation 2 (Ayres et al., 2011):
P{Location
>
x}
=
e
-axn (2)where P{Location > x} is the probability of the overrun/undershoot distance along the runway centerline beyond end of runway is greater than x; x is a given location or distance beyond end of runway; and a, n are regression coefficients.
Similar models exist for other accident types (undershoot, veer off, etc). Having the data probability of accidents and the locations of the final positions of aircraft during accidents and incidents (obtained by the modeling previously described and defined based on the guidelines of Table 1) and estimates of severity (based on the guidelines set out in Table 2), it is possible to define an estimated matrix of probability and severity.
Table 10: Likelihood Levels (FAA, 2010)
General Airport Specific ATC Operational
Per facility NAS – wide Frequent
5
Probability of occurrence per operation is equal to or greater than 1x10-3
Expected to occur more than once per week or every 2500 departures (4x10-4), whichever occurs sooner
Expected to occur more than once per week Expected to occur every 1-2 days Probable 4 Probability of occurrence per operation is less than 1x10-3, but equal to or greater than 1x10-5
Expected to occur about once every month or 250,000 departures (4x10-6), whichever occurs sooner
Expected to occur about once every month
Expected to occur several times per month
Remote 3
Probability of occurrence per operation is less than 1x10-5 but equal to or greater than 1x10-7
Expected to occur about once every year or 2.5 million departures (4x10
-7
), whichever occurs sooner
Expected to occur about once every 1 -10 years
Expected to occur about once every few months
Extremely Remote
2
Probability of occurrence per operation is less than 1x10-7 but equal to or greater than 1x10-9
Expected to occur once every 10-100 years or 25 million departures (4x10
-8), whichever occurs
sooner
Expected to occur about once every 10-100 years
Expected to occur about once every 3 years
Extremely Improbable
1
Probability of occurrence per operation is less than 1x10-9
Expected to occur less than every 100 years
Expected to occur less than once every 100 years
Expected to occur less than once every 30 years
Table 11: Severity Definitions (FAA, 2010) Hazard Severity Classification Minimal E Minor D Major C Hazardous B Catastrophic A No damage to aircraft but minimal injury or discomfort of little consequence to passenger(s) or workers. Minimal damage to aircraft; Minor injury to passengers; Minimal unplanned airport operations limitations (i.e. taxiway closure); Minor incident involving the use of airport emergency procedures. Major damager to aircraft and/or minor injury to passenger(s)/ worker(s); Major unplanned disruption to airport operations; Serious incident; Deduction on the airport's ability to deal with adverse conditions. Severe damage to aircraft and/or serious injury to passenger(s)/ worker(s); Complete unplanned airport closure; Major unplanned operations limitations (i.e. runway closure);
Major airport damage to equipment and facilities. Complete loss of aircraft and/or facilities or fatal injury in passenger(s)/worker (s); Complete unplanned airport closure and destruction of critical facilities; Airport facilities and equipment destroyed.
According to the ICAO risk management manual (ICAO, 2009), the risk assessment for an airport is classified into three categories of management: intolerable, tolerable and acceptable. Such categories are determined from the indices of risk assessment, determined from the probability and severity ICAO matrix, as shown in Table 3.
According to Table 3, the events at "intolerable category" are not acceptable under the existing circumstances, the tolerable category defines the risks that are acceptable with mitigation of risk; finally, the events in the "acceptable category" are acceptable and do not require mitigation risks.
Table 12: Probability and Severity ICAO Matrix (ICAO, 2009)
Likelihood Severity Minimal E Minor D Major C Hazardous B Catastrophic A Frequent 5 5E 5D 5C 5B 5A Probable 4 4E 4D 4C 4B 4A Remote 3 3E 3D 3C 3B 3A Extremely Remote 2 2E 2D 2C 2B 2A Extremely Improbable 1 1E 1D 1C 1B 1A Intolerable category Tolerable category Acceptable category
3. DATA COLLECTION
Table 4 presents the necessary data to be used for analysis. These data were collected for the seven case-study airports.
Table 13: Items of Collected Data Aircraft types Thunderstorms Runway dimensions (m) Frozen precipitation Geographical coordinates Snow occurrence
Air humidity Crosswinds (knots) Altitude (m) Wind occurrence Aircraft class Origin/Destination Visibility (km) Criticality factors Temperature (0C) Wind direction
Fog occurrence Speed Wind (km/h) Ceiling (ft) Icing conditions
These categories of collected data, as proposed at ACRP 50 (Ayres et. al., 2011), are used to define the normal operations conditions in a specific situation in each Brazilian airport analyzed, ie, for a landing or takeoff on specific operational conditions. Regarding the weather information of airport sites analyzed, the collection was made for a specific situation, i.e. in a given time period. The analysis of this study considers operations that occurred on Feb, 10th, 2011. In this case, all analysis and risk evaluations should be considered regarding this limitation.
Table 5 presents the critical aircrafts for the case-study airports, as proposed by Fraga and Müller (2008). They represent the aircrafts that were used as design aircrafts for airport.
Table 5: Critical Aircraft Characteristics (Fraga and Müller, 2008)
Airport Location ICAO
Code Aircraft Wingspan (m)
Distance Between Landing Wheels (m) SBGR Guarulhos-SP 4E B747 - 400 64.92 11.00 SBEG Manaus - AM 4C B737 - 800 34.30 5.72 SBGL Rio de Janeiro - RJ 4E B747 - 300 59.64 11.00 SBBR Brasília - DF 4E A340 - 300 60.30 10.69 SBBE Belém - PA 4C B737 - 800 34.30 5.72
SBPA Porto Alegre - RS 4D MD - 11 51.97 10.70
SBRF Recife - PE 4E A330 - 200 60.30 10.69
4. RISK EVALUATION
Analysis was done using the models proposed at ACRP 50 (Ayres et. al., 2011), in addition to the collected data regarding operations, meteorological factors and airport layout/obstacles. In this case, Table 6 presents the obtained probabilities for events: Landing Overrun, Landing Undershoot, Landing Veer-off, Takeoff
Overrun, and Takeoff Veer-off. The results were obtained from models proposed at ACRP 50, that were developed without considering Brazilian data to determine the regression coefficients, which may bring inconsistencies in terms of engineering.
The Table 7 identifies the obstacles that are located at the vicinity of airport safety areas, which might eventually compromise safety operations.
Table 6: Probabilities of Occurrence for each Event Type Airport Landing Overrun Landing Undershoot Landing Veer-off Takeoff Overrun Takeoff Veer-off SBGR 1.640E-06 9.523E-08 3.983E-07 2.729E-07 4.101E-09 SBEG 3.916E-07 8.454E-08 4.135E-06 9.892E-08 1.490E-08 SBGL 3.494E-07 2.281E-08 8.382E-07 1.817E-07 2.485E-09 SBBR 2.119E-07 3.383E-08 2.463E-06 2.578E-07 3.484E-09 SBBE 6.463E-07 3.893E-07 5.563E-07 1.018E-07 1.043E-07 SBPA 3.029E-07 5.161E-07 5.880E-06 2.878E-07 1.103E-08 SBRF 5.984E-07 4.521E-08 5.455E-06 1.016E-07 2.113E-07
Table 7: Identification of Obstacle at the Vicinity of Airport Safety Areas
Airport Threshold Takeoff Overrun Landing Undershoot Landing Overrun
São Paulo
(Guarulhos) 09
Terrain (A) Minor Edification (E) Minor Edification (H) Vista do Paraíso Street (B) Hélio Smidt Road (F) Ditch (I)
Localizer (C) Sewage Treatment Station (G) Steel Company (J) Minor Edification (D) Rio de Janeiro (Galeão) 15
Localizer (A) Guanabara Bay (E) Localizer (A)
Galeão Road (B) Electrical Station (F) Galeão Road (B) Guanabara Bay (C)
Circulation Road (G) Guanabara Bay (C)
Maracujá Road (D) Maracujá Road (D)
Brasília 11
Service Road (A)
EPAR Road (D) Localizer (F)
Country Club (B) VI Comar (G)
Tree (C) Hangars (E) Minor Edification (H)
Porto Alegre 11
ILS Antenna (A) Localizer (E) ILS Antenna (A)
Vila Dique (B) Estados Avenue (F) Vila Dique (B)
Jardim Floresta (C)
Ditch (G) Jardim Floresta (C)
Ditch (D) Ditch (D)
Manaus 10 Tapajos Avenue (B) Turismo Avenue (A) Tapajos Avenue (B)
Belém 06
Bairro do Bengui (C) Guajará Bay (A) Bairro do Bengui (C) Jardim São Clemente (D) Artur Bernardes Road (B) Jardim São Clemente (D) Recife 16 Mascarenhas Avenue (C) Dom Camara Avenue (A) Mascarenhas Avenue (C)
Jardim Jordão (D) Recife Avenue (B) Jardim Jordão (D)
Defined the barriers within airport sites, the probability of an aircraft impact with each of the obstacles identified for each event type was determined, considering the location probabilities. The Tables 8 and 9 shows the values of the probability of an eventual impact between the aircraft, subject to certain type of event, and obstacles analyzed. Altogether, 83 evaluations of obstacles were performed using the proposed methodology. Generally speaking, most obstacles are categorized within the tolerable or acceptable region for risk management, meaning that either the situation is under
control, or mitigation actions should be proposed for that. The Tables 10 and 11 present ICAO Risk Categories for each obstacle. The Table 12 presents the distribution of obstacles according to the categories of ICAO risk matrix. Figure 6 illustrates the averages of risk values for each airport.
Table 8: Probability of Impact for each Obstacle in Veer-offs
Airports Landing Veer-off
Takeoff Veer-off
São Paulo / Guarulhos 7,58E-08 9,70E-10 7,87E-07 3,52E-09
Airports Landing Veer-off
Takeoff Veer-off
Rio de Janeiro / Galeão 1,60E-07 5,88E-10
Brasília 4,69E-07 8,24E-10
Belém 1,06E-07 2,47E-08
Airports Landing Veer-off
Takeoff Veer-off
Porto Alegre 1,12E-06 2,61E-09
Recife 1,04E-06 5,00E-08
Table 9: Probability of Impact for each Obstacle in Undershoots and Overruns
Airport Takeoff Overrun Landing Undershoot Landing Overrun
Obstacle Probability Obstacle Probability Obstacle Probability
São Paulo (Guarulhos)
Terrain 3,62E-07 Minor
Edification 4,15E-08 Minor Edification 4,08E-08 Vista do Paraíso Street 3,05E-07 Hélio Smidt
Road 1,05E-07 Ditch 2,29E-08
Localizer 3,85E-07
Sewage
Treatment Plant 4,81E-08 Steel Company 3,34E-09 Minor
Edification 1,32E-07 Rio de
Janeiro (Galeão)
Localizer 2,79E-07 Guanabara Bay 2,66E-08 Localizer 3,25E-08 Galeão Road 2,40E-07 Electrical
Station 6,73E-09 Galeão Road 9,87E-09 Guanabara Bay 2,21E-07 Circulation
Road 2,72E-08
Guanabara Bay 4,05E-09
Maracujá Road 1,19E-07 Maracujá Road 3,96E-08
Brasília
Service Road 3,06E-07 Epar Road
(DF-055) 3,44E-08
Localizer 1,39E-08
Country Club 3,23E-07 VI Comar 1,71E-08
Tree 3,25E-07 Hangars 6,40E-09 Minor
Edification 9,83E-08 Porto
Alegre
ILS Antenna 3,42E-07 Localizer 5,80E-07 ILS Antenna 3,19E-07 Vila Dique 3,22E-07 Estados Avenue 5,58E-07 Vila Dique 3,11E-07 Jardim Floresta 1,48E-07
Ditch 6,30E-07 Jardim Floresta 3,74E-07
Ditch 4,44E-08 Ditch 3,36E-07
Manaus Tapajos Avenue 3,77E-11 Turismo Avenue 2,66E-09 Tapajos
Avenue 2,58E-10 Recife Mascarenhas Avenue 1,83E-08 Dom Camara Avenue 8,94E-09 Mascarenhas Avenue 4,38E-10
Jardim Jordão 6,10E-10 Recife Avenue 3,00E-10 Jardim Jordão 9,27E-10
Belém
Bairro do
Bengui 3,22E-08 Guajará Bay 4,85E-09
Bairro do Bengui 5,25E-09 Jardim São Clemente 3,54E-08 Artur Bernardes Road 2,12E-09 Jardim São Clemente 5,09E-09
Table 10: Identification of Obstacle at the Vicinity of Airport Safety Areas Aeroportos Landing Veer-off
(Left Side)
Landing Veer-off (Right Side)
Takeoff Veer-off (Left Side)
Takeoff Veer-off (Right Side)
São Paulo (Guarulhos) 2E 2E 1C 1E
Manaus 3D 3D 2E 2E
Rio de Janeiro (Galeão) 3A 3D 1D 1A
Brasília 3C 3D 1D 1E
Belém 3C 3E 2D 2D
Porto Alegre 3D 3E 2D 2A
Table 11: ICAO Risk Categories for each Obstacle in Undershoots and Overruns
Airport Takeoff Overrun Landing Undershoot Landing Overrun
Obstacle Category Obstacle Category Obstacle Category
São Paulo (Guarulhos) Terrain 3C Minor Edification 2D Minor Edification 2B Vista do Paraíso Street 3B Hélio Smidt Road 3A Ditch 2D Localizer 3E Sewage Treatment Plant 2C Steel Company 2A Minor Edification 3C Rio de Janeiro (Galeão)
Localizer 3E Guanabara Bay 2A Localizer 2D
Galeão Road 3A Electrical
Station 2C Galeão Road 2A
Guanabara Bay 3A Circulation
Road 2D
Guanabara Bay 2A
Maracujá Road 3B Maracujá Road 2B
Brasília
Service Road 3D EPAR Road
(DF-055) 2B
Localizer 2E
Country Club 3C VI Comar 2B
Tree 3C Hangars 2B Minor
Edification 2C
Porto Alegre
Antena ILS 3E Localizer 3E Antena ILS 3E
Vila Dique 3A Estados
Avenue 3A Vila Dique 3A
Jardim Floresta 3A
Ditch 3D Jardim Floresta 3A
Ditch 2D Ditch 3D
Manaus Tapajos Avenue 1D Turismo
Avenue 2D Tapajos Avenue 1D
Recife Mascarenhas Avenue 2B Dom Camara Avenue 2B Mascarenhas Avenue 1B
Jardim Jordão 1B Recife Avenue 1A Jardim Jordão 1B
Belém Bairro do Bengui 2B Guajará 2A Bairro do Bengui 2B Jardim São Clemente 2B Artur Bernardes Road 2B Jardim São Clemente 2B
As it can be seen from Table 12 and Figure 6, Porto Alegre and Galeão (Rio de Janeiro) are the airports with the highest number of non-compliances, among the selected list of airports evaluated in this study. Observing Figure 7, we identify that proximity to the Guanabara Bay is one of the main concerns of the airport; there is little room for locating service roads and other elements of the airport between the runway
ends and the Bay, without compromising safety. Similar situation happens at Porto Alegre International Airport, where the major concern is the close proximity of urban areas at the vicinity of the airport. In this case, besides imposing a challenge for managing current operations, these obstacles might be a restriction for future airport expansions.
Table 12: Distribution of Obstacle Within ICAO Risk Categories
Airports Acceptable Tolerable Intolerable São Paulo
(Guarulhos) 50.00% 42.86% 7.14%
Rio de Janeiro
(Galeão) 33.33% 46.67% 20.00%
Brasília 25.00% 75.00% 0.00%
Airports Acceptable Tolerable Intolerable
Porto Alegre 40.00% 26.67% 33.33%
Manaus 71.43% 28.57% 0.00%
Recife 70.00% 30.00% 0.00%
Belém 30.00% 70.00% 0.00%
TOTAL 43.37% 45.78% 10.84%
Figure 6: Risk Averages for Each Airport
5. CONCLUSIONS
The presented analysis in this paper presents a proposed methodology for risk evaluation at airports, combining an existing quantitative method, in association with guidelines from FAA and ICAO. Data was collected to apply the methodology, and to develop a preliminary assessment of risk values at major Brazilian airports.
Future research could be developed in order to collect considerably more data, especially meteorological and operations data for a long period. A simulation could be performed to estimate risk for a long period (10/20 years), especially considering future airport expansions and estimated growing air transport demand in Brazil.
AKNOWLEDGEMENTS
The authors of this paper would like to thank Federal Agency for Research in Brazil for the support provided for this research.
REFERENCES
AYRES, M., Shirazi, H., Carvalho, R., Hall, J., Speir., R., Arambula, E., Robert, D., Wong, D. and Gadzinski, J. (2011), Improved Models for Risk Assessment of Runway Safety Areas (ACRP - Report 50), Transportation Research Board, Washington, D. C..
BOEING (2013), Statistical Summary of Commercial Jet Airplane Accidents - Worldwide Operations 1959-2012., Seattle, Washington.
CAVES, R. (1996), "Control of Risk Near Airports", Built Environment, vol. 22, n. 3, pp. 223-233.
DAC (2004), Reference Guidelines for an Aeronautical Study, Brazilian Department of Civil Aviation, Rio de Janeiro.
EDDOWES, M., Hancox, J. and MacInnes, A. (2001), Final Report on the Risk Analysis in Support of Aerodrome Design Rules, Norwegian Civil Aviation Authority, Norway.
FAA (2010), Order 5200.11, FAA Airports (ARP) Safety Management System, Washington, D. C.
FRAGA, R. and Müller, C. (2008), "Análise das Características Físicas da Área de Movimento dos Aeroportos Brasileiros", SITRAER Conference Proceedings, pp. 432-444.
HALL, J., Ayres, M., Wong D., Appleyard, A., Eddowes, M., Shirazi, H., Speir, R., Pitfield, D., Caves, R., Selezneva, O. and Puzin, T. (2008), Analysis of Aircraft Overruns and Undershoots for Runway Safety Areas (ACRP – Report 03. Transportation Research Board, Washington, D. C.. ICAO (2013), DOC 9859-AN/474, Safety Management Manual, International Civil Aviation Organization, Montreal.
