Fuel Consumption of the 50 Most Used Passenger Aircraft

• Explain the methods to calculate fuel consumption. • Introduce the data sources to determine fuel consumption directly (EEA, BADA) or the data sources that obtain the parameters to calculate fuel consumption (Aircraft Characteristics for Airport Planning, EASA Aeroplane CO2 Emissions Database) 5 • Analyze your fuel consumption data by looking at average values, the coefficient of variation, the coefficients of determination, the mean absolute percentage error, take your matrix of coefficients of determination as a Design Structure Matrix (DSM) and analyze the DSM. • Analyze and interpret the fuel consumption data for all your aircraft in a systematic way. • Describe your Excel tool. • List the fuel consumption data in Excel and HTML on the WWW. The report has to be written in English based on German or international standards on report writing. 6 Table of Contents Page List of Figures ............................................................................................................................ 8 List of Tables ............................................................................................................................ 9 List of Symbols ........................................................................................................................ 10 Greek Symbols ......................................................................................................................... 11 List of Abbreviations ................................................................................................................ 12 List of Definitions .................................................................................................................... 13 1 Introduction .................................................................................................... 14 1.1 Motivation ........................................................................................................ 14 1.2 Title Terminology ............................................................................................. 14 1.3 Objectives ......................................................................................................... 15 1.4 Previous Research............................................................................................. 15 1.5 Structure of the Work ....................................................................................... 16 2 Literature Review ........................................................................................... 17 2.1 Overview .......................................................................................................... 17 2.2 Bibliometric Analysis ....................................................................................... 17 2.3 Literature Review ............................................................................................. 21 3 Fuel Calculation .............................................................................................. 22 3.1 Overview .......................................................................................................... 22 3.2 SAR Method ..................................................................................................... 24 3.3 Extended Payload Method ................................................................................ 26 3.4 Bathtub Method ................................................................................................ 28 3.5 CO2 Method ..................................................................................................... 32 3.6 Handbook Method ............................................................................................ 34 4 Data Sources .................................................................................................... 37 4.1 Overview .......................................................................................................... 37 4.2 Passenger Aircraft Census ................................................................................ 37 4.3 Student Data ..................................................................................................... 39 4.4 BADA ............................................................................................................... 41 4.5 EEA .................................................................................................................. 43 4.6 EASA ................................................................................................................ 48 5 Data Analysis................................................................................................... 50 5.1 Overview .......................................................................................................... 50 5.2 Comparison of Methods ................................................................................... 51 5.2.1 Average Analysis .............................................................................................. 53 7 5.2.2 Coefficient of Variation Analysis ..................................................................... 53 5.2.3 Coefficients of Determination Analysis ........................................................... 55 5.2.4 Regression Coefficient Analysis as Design Matrix .......................................... 56 5.2.5 Mean Absolute Percentage Error (MAPE) Analysis ........................................ 57 5.3 Comparison of Data .......................................................................................... 58 6 Summary ......................................................................................................... 63 6.1 Summary Methods ............................................................................................ 63 6.2 Summary Fuel Consumption Aircraft .............................................................. 64 6.3 Importance of Fuel Consumption in Civil Aviation ......................................... 64 List of References ................................................................................................................... 66 8 List of Figures Figure 3.1 Summary of candidate metrics (Bonnefoy 2010) ............................................. 23 Figure 3.2 Simplified payload range-diagram (Van Endert 2017) .................................... 24 Figure 3.3 Statistic to predict the typical number of seats from the maximum seating capacity (Hurtecant, 2021)................................................................................ 25 Figure 3.4 Extended payload-range diagram (Young 2017) .............................................. 27 Figure 3.5 Example of an A350-900 bathtub curve (Burzlaff 2017) ................................. 28 Figure 3.6 Different phases during flight operation (Mac Donald 2012) .......................... 30 Figure 4.1 Extract from fuel calculator (Burzlaff 2017) .................................................... 40 Figure 4.2 Extract from BADA (EUROCONTROL 1998) ............................................... 42 Figure 4.3 Annual number of flights and payload-range diagram of an Airbus A320-200 (left) and a Boeing 737-800 (right) (Linke 2020)............................................. 44 Figure 4.4 Annual number of flights and payload range diagram of a Boeing 787-8 (Chai 2018) ................................................................................................................. 45 Figure 4.5 Extract of EEA Master Emission Calculator by example of an A320 ............. 47 Figure 5.1 Results of the average fuel consumption of all methods (left) and the ranking (right) ................................................................................................................ 53 Figure 5.2 Results of the standard deviation of all methods (left) and the ranking (right) 54 Figure 5.3 Results of the coefficient of variation of all methods (left) and the ranking (right) .......................................................................................................................... 54 Figure 5.4 7x7 Matrix of the coefficient of determination................................................. 55 Figure 5.5 Results of the average coefficient of variation of all methods (left) and the ranking (right) ................................................................................................... 56 Figure 5.6 Results of the values of the highest eigenvalue of all methods (left) and the ranking (right) ................................................................................................... 57 Figure 5.7 7x7 Matrix of MAPE ........................................................................................ 57 Figure 5.8 Results of MAPE of all methods (left) and the ranking (right) ........................ 58 Figure 5.9 Comparison of fuel consumption of turboprop aircraft .................................... 59 Figure 5.10 Comparison of fuel consumption of narrow body aircraft ............................... 60 Figure 5.11 Comparison of fuel consumption of wide body aircraft ................................... 61 Figure 5.12 Comparison of fuel consumption of A330-900 ................................................ 62 9 List of Tables Table 2.1 Boolean Operators and keywords for systematic literature review .................. 18 Table 2.2 Findings ............................................................................................................ 19 Table 3.1 Overview flight phases ..................................................................................... 30 Table 4.1 List of the 50 most used passenger aircraft (World Airliner Census 2020) ..... 38 Table 4.2 Extended Payload Range Diagram vs. EEA Master Emission Calculator ....... 46 Table 5.1 Overview of Excel tabs with explanation ......................................................... 51 10 List of Symbols A Aspect Wing Ratio a0 Speed of sound at sea level a11km Speed of sound in 11km B Breguet Range Factor bW Wingspan C Fuel consumption c Specific fuel consumption, thrust specific (for jets) cP Specific fuel consumption, thrust specific (for turboprop) dF Fuselage diameter E Aerodynamic efficiency Emax Max. Aerodynamic efficiency g Gravitational acceleration h Altitude ke Glide ratio factor lA/C Aircraft length lCabin Cabin length m Mass Mcr Cruise Mach number Mff Fuel fraction mF Fuel weight mMZF Maximum zero fuel mass mOE Operating empty mass mPL Payload mass mTO Takeoff mass nmax max. number of seats nOEM Number of seats from Original Equipment Manufacturer OPR Overall Pressure Ratio Pssl.eq Take-off power (equivalent) at static sea level R Range r Radius SW Wing Area Swet Wetted Area s Chord TET Turbine entry temperature T0 Reference temperature at sea level T11km Reference temperature in 11km V Velocity 11 Greek Symbols Δ Difference ηP Propeller Efficiency l Bypass ratio 12 List of Abbreviations AEED Aircraft Engine Emissions Databank AEM Advanced Emission Model ATM Air Traffic Management BADA Base of Aircraft Data BFFM2 Boeing Fuel Flow Method 2 CAS Calibrated Air Speed CCD Climb, Cruise, Descent EEA European Environment Agency EEC EUROCONTROL Experimental Centre EASA European Union Aviation Safety Agency FAR Federal Aviation Regulations FEIS Fuel Burn and Emissions Inventory System FL Flight Level ICAO International Civil Aviation Organization ISA International Standard Atmosphere LCR Long Range Cruise LTO Landing and Take-Off cycle MCR Maximum Range Cruise MEW Manufacturer Empty Weight MTOW Maximum Take-Off Weight MZFW Minimum Zero Fuel Weight MV Metric Value OEM Original Equipment Manufacturer OEW Operating Empty Weight OML Outer Mould Line OPERA Optimization in Preliminary Aircraft Design RGF Reference Geometric Factor R&D Research & Development SAR Specific Air Range TOW Take-Off Weight UID Unique Identification Number 13 List of Definitions Breguet Louis Charles Breguet was a 1880 born aircraft designer, who is falsely considered as the originator of the “Breguet Range Equation”. Originally, this equation was introduced in 1920 by J. G. Coffin in his NACA Report (NACA 1969). Since this equation is known as the Breguet Range Equation, it will be called in this project Breguet Range Equation as well (Burzlaff 2017) Bathtub curve The Bathtub Curve is a visualization of fuel consumption per passenger and 100 km flight distance over the flown distance. With this diagram, the range on which an aircraft can be operated most efficiently can be shown. The course of this curve conforms figurative to the profile of a bathtub, where the name originates. (Burzlaff 2017) Cruise flight Portion of a flight from the point where the aircraft has leveled off following a climb to its initial cruising altitude until the point where it commences its descent. (Young 2017) Fuel consumption Indicates the amount of fuel consumed per kilogram per kilometer per seat. All phases of the flight are considered. Fuel reserves The reserves are additional fuel carried on every flight to be prepared for unscheduled occurrences. The size of reserves depends on various factors, such as distance to alternate or weather conditions (Burzlaff 2017). Specific air range The still air distance traveled per unit of fuel mass consumed. (Young 2017) 14 1 Introduction 1.1 Motivation Fuel consumption of cars is known for each model, and it is easily comparable because every car manufacturer is obliged to refer its fuel consumption to the Worldwide Harmonized Light Vehicles Test Procedure (WLTP) standard. But what about the fuel consumption of passenger aircraft? A standard is missing, the variety of configurations in different market segments are more difficult to compare (as for cars) and aircraft manufacturers do not entirely publish the fuel consumptions of their models. But there are ways to get more transparency on that. This paper provides a variety of calculations for the 50 most used passenger aircraft in the world. The different methods will be explained, analysed and compared so that the most rele- vant data is then taken into account for the ranking the most efficient passenger aircraft. The evaluation will be discussed in detail, and it reveals transparency for older aircraft versus newer ones, turboprop and turbojet aircraft and single aisle as well as wide body aircraft. Have the aircraft manufacturers already reached a certain optimum? Some technological advances are indicated at the end. The motivation of aircraft manufacturers to increase fuel efficiency is twofold – to increase the economy of operations and to reduce the environmental footprint. 1.2 Title Terminology “Fuel Consumption of the 50 Most Used Passenger Aircraft” Fuel A substance that is used to provide heat or power, usually by being burned (Cambridge Dictionary). Consumption An amount of something that is used, or the process of using something, especially so that there is less of it (Cambridge Dictionary). Passenger 15 A person who is travelling in a vehicle bus is not driving it, flying it, or working on it (Cambridge Dictionary). Aircraft Any machine that can derive support in the atmosphere from the reaction of the air (ICAO 2005). Aircraft category Classification of aircraft according to specified basic characteristics, e.g., airplane, glider, rotorcraft, free balloon (ICAO 2020b). Each aircraft is divided into categories by certain certification rules, which are included in the type certificate. All aircraft mentioned in the project are certified according to FAR Part 25, Transport Category Airplanes (USA) or EASA CS-25, Large Airplanes (Europe). In order to clearly distinguish them from cargo aircraft, the word aircraft has been replaced by passenger aircraft in the title. 1.3 Objectives The difficulty lies in the fact that fuel consumption for individual aircraft types is not suffi- ciently described in the literature. Therefore, the first step is to conduct a systematic literature review on the topic in order to reflect the current state of the art. In addition, the work is to inform about the different definitions of fuel consumption and show which calculation methods for fuel consumption are available. The main objective of the project is to create an Excel file that allows the user to view all the consumptions of the most commonly used aircraft models (World Airliner Census 2020), based on the different calculation methods. The selected meth- ods will be compared and analyzed with each other and the aircraft will be sorted according to their consumptions. 1.4 Previous Research This project is based on the master thesis of Daan Hurtecant (2021) and the project work of Marcus Burzlaff (2017). In his master's thesis, Daan Hurtecant (2021) dealt with the most important environmental im- pacts of air traffic and how these can be evaluated from publicly accessible databases using an 16 Ecolabel. His considerations and calculations for the fuel consumption of certain aircraft types are meaningful. He also uses the World Airliner Census 2020 as a basis for this. In addition, he discusses three calculation methods, all of which are an integral part of this project work and are compared with each other. His Excel file for the Ecolabel is very helpful, as it also contains, among other things, many essential key figures for the calculation of fuel consumption. The project work of Marcus Burzlaff (2017) deals with the calculation of fuel consumption and the implementation of an Excel file in order to be able to determine the fuel requirements of each aircraft from the publicly available documents "Aircraft characteristics for Airport plan- ning". This so-called bathtub curve method is used as another possibility for calculating fuel consumption. 1.5 Structure of the Work This project consists of 6 main chapters. The structure of this work is as follows: Chapter 2 The project starts with a systematic literature review on fuel consumption re- lated to aircraft of general aviation. The current state of the art is summarized here. Chapter 3 This chapter explains the different methods of fuel consumption calculations with the underlying mathematical relations. Chapter 4 All necessary data sources and databases are listed and explained here. Chapter 5 This chapter explains the Excel file, compares all calculation methods, lists their variance and attempts to explain them. Chapter 6 This chapter provides the summary of the project. 17 2 Literature Review 2.1 Overview In science, a Systematic Literature Review (SLR) is a procedure or method that involves iden- tifying, collecting, analyzing, and critically evaluating relevant literature. In doing so, it is im- portant to focus exclusively on articles that meet the specified inclusion criteria in an unbiased manner in order to approach a specific research question or hypothesis. If all relevant literature is available, a conclusion can be drawn that will influence further work and its results. Thus, the systematic literature review shows the state of the art for a specific topic. (Snyder 2019) 2.2 Bibliometric Analysis In this chapter, a systematic literature search will show what expert literature already exists about fuel consumption of aircraft. Let’s start with the master thesis of Daan Hurtecant (2021). Using the corresponding bibliography, further sources can be found to which this work refers. As a next step the search was extended to the data beacons of libraries. Elsevier's Scopus and Google Scholar, which were also used for the preparation of this project work, are particularly suitable for this purpose. Other catalogs can be found in the listing of Table 2.1. All entries in the library databases can be concretised or combined with Boolean operators. Their meaning is described below: • AND means that both entered terms are included in the results. With each further "AND" the search is supplemented by the further search term. However, this limits the number of search hits. • OR is used if at least one of the terms is to be contained in the search hits. Each additional "OR" extends the search radius by the described term and the number of hits. • NOT excludes the following term. • Parentheses are used to link several search terms by different Boolean operators. • Truncations are a great help when a word family is searched for, an exact spelling is not known, or several spellings should be allowed. Often "*" or "?" are used for this purpose. (Kache 2005) Table 2.1 now shows specifically which keywords and operators were used to try to specify the literature search. 18 Table 2.1 Boolean Operators and keywords for systematic literature review Keyword 1 Operator Keyword 2 ((Fuel) OR (Fuel Consumption) OR (Fuel Burn)) AND ((Aircraft) OR (Aviation) OR (Airplane)) On Google Scholar, well over 100,000 hits were obtained using the above keywords. A detailed and differentiated analysis is not possible due to the high number of hits. Therefore, the search had to be narrowed down further. Thus, the focus was placed on sources that contain one or more of the keywords in the title. Then the search was preceded by an "allintitle:" to narrow down the search. This reduced the number of relevant literatures to 44. After a review of the documents, apparently only one document is really useful: The aforementioned project paper by Marcus Burzlaff (2017) entitled "Aircraft Fuel Consump- tion - Estimation and Visualization". The focus was more on emissions in flight rather than fuel consumption in general. The same keywords were used to continue the search at Elsevier's Scopus. However, Elsevier's Scopus has an extended search mask. So it is possible to limit the search with the keywords to the abstract of a document, to the title, to the abstract or to keywords defined by the author. This reduces the number of hits. However, almost 50,000 entries were still found. Different restrictions reduced the entries, but very few were helpful. Rather, fuel consumption was de- scribed in general, some cited and explained Breguet's range formula, others mentioned pay- load-range diagrams besides. In general, it can be stated that most sources only deal with different approaches to fuel calcu- lation. In almost every source, Breguet's range equation is listed, but extended correlations are not mentioned. Furthermore, there is no literature or work that explicitly lists the fuel consumption of individual aircraft. Scholz's (2021) "Hints for Project and Thesis Writing" provides additional links to libraries and catalogs. Further search queries can be made in them. An overview of libraries used and the results can be seen in Table 2.2. 19 Table 2.2 Findings Title Author Year http://reports-at-aero.ProfScholz.de Conditions for Passenger Aircraft Minimum Fuel Consumption, Direct Operating Costs and Envi- ronmental Impact Caers, B.; Scholz, D. 2020 Fuel Costs due to Aircraft Systems Scholz, D. 2007 Der spezifische Kraftstoffverbrauch von Flug- triebwerken (TSFC und PSFC) Scholz, D. 2023 Kraftstoffverbräuche von Turbofan, Propfan und Turboprop im Vergleich Koppe, M. 2012 Aircraft Preliminary Sizing with PreSTo - Re-De- sign of the Boeing B777-200LR Seeckt, K. 2008 An Ecolabel for Aircraft Scholz, D. 2017 http://library.ProfScholz.de Direct Operating Costs, Fuel Consumption, and Layout of the Airbus A321LR Fonseca, D. 2021 Flugzeugentwurf am Beispiel eines Embraer ERJ 145 Lee, S. 2001 Flugzeugentwurf am Beispiel einer Fokker 100 Kohrt, M. 2001 Flugzeugentwurf am Beispiel einer Boeing B737- 300 Wolgast, P. 2002 Flugzeugentwurf am Beispiel einer Boeing 757- 200 Fricke, A. 2004 Aircraft Design Studies Based on the ATR 72 Niţă, M. 2008 Assessment of Numerical Models for Thrust and Specific Fuel Consumption for Turbofan Engines Schulz, O. 2007 Identifying CO2 Reducing Aircraft Technologies and Estimating their Impact on Global Emissions Apffelstaedt, A. 2009 Background to the 3-Liter-Aircraft - How Clean is Aviation? Tan, Y.L. 2011 Developing an Ecolabel for Aircraft Haß, T. 2015 A General View on Fuel Efficiency in Commercial Aviation MacDonald, A. 2012 Launch of an Ecolabel for Passenger Aircraft Hurtecant, D. 2021 Definition of an Ecolabel for Aircraft Van Endert, L. 2017 Aircraft Fuel Consumption- Estimation and Visu- alization Burtzlaff, M. 2017 Conditions for Passenger Aircraft Minimum Fuel Consumption, Direct Operating Costs and Envi- ron- mental Impact Brecht, C. 2019 http://paper.ProfScholz.de Fuel Consumption due to Shaft Power Off-Takes from the Engine Scholz, D.; Seresinhe, R.; Staack, I.; Lawson, C. 2013 https://katalog.haw-hamburg.de 20 Impacts of COVID-19 on aircraft usage and fuel consumption: A case study on four Chinese in- ternational airports Dabin Xue, Zhizhao Liu, Bing Wang, Jian Yang 2021 Fuel consumption optimization in air transport: a review, classification, critique, simple meta-analysis, and future research implications Singh, V.; Sharma, S.K. 2015 Future directions of fuel efficiency in aviation in- dustry Mrazova, M. 2013 Factors Affecting the Rate of Fuel Consumption in Aircraft Hassan, T.H.; Sobaih, A.E.E; Salem, A.E. 2021 https://dnb.de Characteristics of the Specific Fuel Consumption for Jet Engines Bensel, A. 2018 https://catalog.loc.gov General aviation statistics [Washington] Dept. of Transportation, Federal Aviation Administration, Office of Management Systems, Information and Statistics Division. https://www.worldcat.org Minimization of aircraft fuel consumption sub- jected to arrival time constraint Gilani, M.; Körpe, S.K. 2020 https://www.scopus.com/home.uri Fuel Estimation in Air Transportation: Modeling global fuel consumption for commercial aviation Seymour, K.; Held, M.; Georges, G.; Boulouchos, K. 2020 Improving aircraft approach operations taking into account noise and fuel consumption A. Rodríguez-Díaz, A.; Adenso-Díaz, B.; González-Torre, P.L. 2019 Fuel efficiency of commercial aircraft An overview of historical and future trends Peeters, P.M.; Middel, J.; Hoolhorst, A. 2005 Measurement of Aircraft Fuel Efficiency Lim, D.; Kirby, M.; Nam, T. 2014 A Neural Network Model to Estimate Aircraft Fuel Consumption Trani, A.A.; Wing-Ho, F.C.; Schilling, G.; H. Baik, H.; Seshadri A. 2004 Fuel Consumption Model of the Climbing Phase of Departure Aircraft Based on Flight Data Anal- ysis Zhang, M.; Huang, Q.; Liu, S.; Zhang, Y. 2019 Multipoint optimization on fuel efficiency in con- ceptual design of wide-body aircraft Chai, X.; YU, X.; WANG, Y. 2017 21 2.3 Literature Review In the course of the literature research, in addition to the different calculation methods, an ex- plicit search was made for fuel consumption of aircraft. In Wikipedia there is a listing of different airplanes and their consumptions referenced to a certain flight route. The task was to find out, view and analyze the original referenced sources. With the help of the Wayback Machine even deleted documents or web pages could be recov- ered. Most of the documents are information brochures of the manufacturers or articles that compare some aircraft models and their consumption. In most cases, the consumption figures are refer- enced to actual routes flown. It should be noted that not every airline has the same seat config- uration, especially not on long-haul routes. Therefore, the number of seats should be checked with the help of Seatguru and, if necessary, the typical seat configuration should be selected in case of missing information. It should be noted that consumption figures are usually given in different units. These need to be converted to a comparable unit. A detailed listing of the aircraft with the respective links can be found in the appendix. The consumptions can be found in the Excel file in the tab "Literature review". 22 3 Fuel Calculation 3.1 Overview Before the individual calculation methods are presented, it is important to clarify which key figures describe the fuel consumption of an aircraft. Besides, the efficiency of an aircraft is determined by additional metrics that matters for airlines. When measuring fuel consumption, a distinction must be made between: • Full Mission Performance: The fuel consumption of the aircraft is considered during the entire mission. From leaving the gate at the departure airport until arrival at the destination gate. • Instantaneous Performance: The fuel consumption of the aircraft is measured in steady-state cruise flight at an arbitrary point in time. The full mission performance data, ideally being standardized and therefore comparable, is hardly published. Hence, in the following calculation of fuel consumption a selective perfor- mance indicator for steady-state cruise flight based on the Specific Air Range (SAR) is pro- posed for further considerations. SAR describes the distance an aircraft will travel on the next incremental amount of fuel burned (Bonnefoy 2010). SAR is not a certified metric, but the key advantage is that only a few parameters such as alti- tude, speed, weight and atmospheric conditions are required for that. For traceability reasons it is necessary to include only publicly available parameters in the calculation. Equation (3.1) defines the SAR and shows that the Breguet factor (B) depends on the aerody- namic efficiency (E), the velocity (V), the gravitational acceleration (g) and the specific fuel consumption (c). The exact determination of the parameters is complex, e.g., the aerodynamic efficiency can only be determined by (wind tunnel) measurements or Computational Fluid Dy- namics calculations, a simplification is necessary here. (Chapter 3.2) SAR = − 𝑑𝑅 𝑑𝑚 = 𝑉 ∙ 𝐸 𝑐 ∙ 𝑔 ∙ 1 𝑚!"#$!%# = 𝐵 ∙ 1 𝑚!"#$!%# (3.1) Furthermore, the efficiency or "Air Transportation Output" includes the following parameters to the comparison (Bonnefoy 2010). 23 • „Measure of distance travel“ - Range (i.e. distance) • “Measure of speed“ - Maximum Range Cruise (MCR), Long Range Cruise (LCR) • “Measure of time” - Time (e.g. Block time, Air time) • “Measure of transportation” - Payload (Maximum Zero Fuel Weight (MZFW) - Operating Empty Weight (OEW) or Maximum Zero Fuel Weight (MZFW) – Manufacturer Empty Weight (MEW)) - Useful load (Maximum Takeoff Weight (MTOW) – Operating Empty Weight (OEW)) - Maximum Takeoff Weight (MTOW) - Floor Area - Number of Available Seats In order to be able to give a practical value the number of available seats was chosen for the "Air Transportation Output". It is important that all the methods described below have the same unit so they can be compared with each other. A graphical summary of the parameters men- tioned and the variant selected for the further course can be found in Figure 3.1. Figure 3.1 Summary of candidate metrics (Bonnefoy 2010) 24 3.2 SAR Method The specific flight range can be determined only with the help of the payload range diagram according to the definition and formula of SAR. Already known is: SAR = − 𝑑𝑅 𝑑𝑚 (3.2) Figure 3.2 shows a simplified payload-range diagram. Here, R1 is the maximum range with the maximum payload m1 (also called harmonic range). Reducing the payload and increasing the fuel leads to the longer range R2 as the maximum range with maximum fuel carrying the asso- ciated reduced payload m2. These payload-range diagrams can be found in the publicly available documents for airport planning. They are published by the respective aircraft manufacturer, who provides information about various characteristics, dimensions and key figures of the aircraft, which are indispen- sable for airport and maintenance planners. Figure 3.2 Simplified payload range-diagram (Van Endert 2017) Therefore, considering the payload-range diagram SAR can alternatively be described as fol- lows. SAR = 𝑅& − 𝑅' 𝑚' − 𝑚& [km/kg] (3.3) However, SAR is expressed in kilometers per kilogram according to Equation (3.3). Since this is not a typical unit, the reciprocal of SAR results in an expression commonly used in the transport sector. 25 C = 1 𝑆𝐴𝑅 = 𝑚' − 𝑚& 𝑅& − 𝑅' [km/kg] (3.4) This results in an absolute value for the average fuel consumption of an aircraft. However, this value cannot yet be compared with other aircraft consumption figures. Naturally smaller air- craft, which have a shorter range and therefore a lower mass, consume less fuel than larger aircraft. Therefore, as already mentioned in Chapter 3.1, the capacity of the aircraft in the form of the number of available seats installed by the original equipment manufacturer (OEM) needs to be included in the calculation. This results in the inverse specific range being divided by the typical number (n) of seats on an aircraft. 𝐶()* = 1 𝑆𝐴𝑅 ∙ 𝑛()* [km/kg/seat] (3.5) The typical seating is often found in the documents for airport planning. It may happen that this figure cannot be found and only the maximum seating capacity is given. Using the maximum seating capacity to normalize fuel consumption would cause the rating scale to be overly strict because most airliners are fitted with a number of seats that is lower than the maxi- mum number (Hurtecant 2021) Therefore, Hurtecant (2021) plotted the typical seating of 75 reference aircraft against the max- imum seating capacity to estimate the typical number of seats for aircraft using a linear regres- sion. The accuracy of the straight-line equation is 97% and is part of Figure 3.3. Figure 3.3 Statistic to predict the typical number of seats from the maximum seating capacity (Hur- tecant, 2021) 26 Hurtecant (2021) concludes that the number of seats can be calculated as follows. 𝑛()* = 0.6696 ∙ 𝑛+!, + 22.858 (3.6) In a nutshell, the SAR method is a point-by-point consideration of fuel consumption in cruise flight. Therefore, SAR cannot be used if the amount of fuel consumed is to be considered over the entire flight. However, the inherent simplicity of a point performance metric makes it attractive for this research (Bonnefoy 2010). As the fuel consumption for every aircraft is calculated similarly, it allows for easy comparison between aircraft. The proposed equation can thus be used as a reliable and standardized indicator for the average fuel performance (Hurtecant 2021). 3.3 Extended Payload Method Chapter 3.2 describes a possible method for calculating aircraft fuel consumption from the pay- load-range diagram using only four known, publicly available parameters. Another method re- sults from extending the payload-range diagram by the following relations (Young 2018). 𝑚*-./ = 𝑚()/ + 𝑚01 (3.7) 𝑚*2(/ = 𝑚*-./ + 𝑚. (3.8) This can be shown graphically using an alternative format (see Figure 3.4), which has the aircraft’s weight as the ordinate and the payload–range envelope shown within the diagram (shaded). From point A to point B, the payload is constant, but the TOW increases as additional fuel is required for the increasing range. From point B to point C, payload is traded for fuel (here the TOW is equal to the MTOW). This progressive increase in range, resulting from the increase in fuel, is possible until the maximum fuel limit (i.e., fuel capacity limit) is reached. From point C to point D, the fuel is limited by the size of the fuel tanks, but the payload (and hence the TOW) reduces as the range increases (Young 2018). 27 Figure 3.4 Extended payload-range diagram (Young 2017) In addition, the two reference values R1 and m1 are shown in blue in Figure 3.4. Both values describe a point in the payload-range diagram, which is also called the design point of an air- craft. As already used in the SAR method, only the punctual fuel consumption in cruise flight is considered here. Under this assumption and using Figure 3.4 in combination with Equation (3.7), the fuel con- sumption in kilograms per kilometer can now be calculated. C = 𝑚. 𝑅' = 𝑚*2(/ − 𝑚*-./ 𝑅' [kg/km] (3.9) The fuel mass mF considered here contains an additional increment to the regular fuel: the mass of the reserve fuel. Therefore, the fuel consumption calculated with Equation (3.8) is higher than the actual value. In order to be able to compare the consumption of the respective aircraft, the consumption is normalized with the typical seat configuration of the manufacturer. 𝐶()* = 𝑚. 𝑅' ∙ 𝑛()* = 𝑚*2(/ − 𝑚*-./ 𝑅' ∙ 𝑛()* [kg/km/seat] (3.10) The decisive advantage of this Extended Payload Method over the SAR method is that only the parameter R1 has to be read from the payload-range diagram for the calculation. The other two parameters, MTOW and MZFW, are always known from the certification documents of the aircraft. The extended payload-range diagram method eliminates the need to study every payload range di- agram in depth, and therefore any inaccuracies in reading the diagram are also avoided. In conclu- sion, the proposed equation can be used as a reliable and standardized indicator for the average 28 fuel performance and as a simplified alternative to the SAR-based fuel consumption (Hurtecant, 2021). 3.4 Bathtub Method In the previous chapters, two methods of calculating fuel consumption were explained which are only applicable to a flight at cruising altitude. In the following, a method will be presented in which the consumption is considered over the entire flight, including all flight phases. In his project work, Burzlaff 2017 developed an Excel tool that calculates the fuel consumption of an aircraft for an exact range and generates a graph in form of a so-called bathtub curve. This curve describes the fuel consumption per 100 km and per seat and shows the area of least con- sumption. Figure 3.5 shows the bathtub curve using the A350-900 as an example: Figure 3.5 Example of an A350-900 bathtub curve (Burzlaff 2017) It can be clearly seen that consumption jumps at the edges of the bathtub curve, but in the range from 2800 km to 11000 km it remains in a relatively narrow range of 2.8…2.1 kg/100km/pax. To better understand the composition of the bathtub curve, some mathematical relationships are explained: Already known is the Breguet factor B. B = 𝑉 ∙ 𝐸 𝑐 ∙ 𝑔 (3.11) 0,00 1,00 2,00 3,00 4,00 5,00 6,00 7,00 8,00 9,00 0 2000 4000 6000 8000 10000 12000 14000 16000 18000 Fuel per Range & per Pax [kg/100km] Range [km] Fuel per Range & Pax versus Range 29 This forms the Breguet Range Equation. 𝑅 = 𝐵 ∙ ln 𝑚' 𝑚& (3.12) Transforming Equation (3.11). 𝐵 = 𝑅 ln 𝑚' 𝑚& (3.13) Equation (3.12) only applies to the horizontal flight. However, since the bathtub method should take into account the entire mission of the flight (including takeoff, climb, cruise descent, loiter and landing), the individual fuel fractions must also be taken into account. A Fuel Fraction is a relation between the mass m2 at the end of a phase of flight and the mass m1 at the beginning of this phase of flight (Scholz 2015) 𝑀33 = 𝑚& 𝑚' (3.14) Considering the flight phases, the result for Mff: 𝑀!! = 𝑚"#$% '!! 𝑚()*+,*- ∙ 𝑚()*+,*- 𝑚(.,%/0 ∙ 𝑚(.,%/0 𝑚1/23/*% ∙ 𝑚1/23/*% 𝑚4/2/05/ ∙ 𝑚4/2/05/ 𝑚67,89 ∙ 𝑚67,89 𝑚1/23/*% ∙ 𝑚1/23/*% 𝑚60$,2/ ∙ 𝑚60$,2/ 𝑚67,89 ∙ 𝑚67,89 𝑚:);/ '!! (3.15) Simplified, it can be written: 𝑀33 = 𝑀33,15% ∙ 𝑀33,167 ∙ 𝑀33,8#9 ∙ 𝑀33,:#9 ∙ 𝑀33,;<= ∙ 𝑀33,8#9 ∙ 𝑀33,;$9 ∙ 𝑀33,;<= ∙ 𝑀33,26 (3.16) 30 Figure 3.6 shows an overview of the flight phases accompanying Equations (3.14) and (3.15): Figure 3.6 Different phases during flight operation (Mac Donald 2012) The individual flight phases can be divided into two different groups. The Horizontal Flight and the Non-Horizontal Flight. Table 3.1 shows besides the two categories also the classifica- tion of the formula expressions according to Equation (3.14). Table 3.1 Overview flight phases Flight phase Non-Horizontal flight Horizontal flight Fuel Fraction Mff,TO; Mff,Clb; Mff,Des; Mff,Ldg Mff,Crs; Mff,Res; Mff,Loi All fuel fractions describing the non-horizontal flight can be described simplified with Mff,LTO. All other components are described simplified with Mff,Cr-Res-Loi The fuel mass fractions for each stage of the LTO cycle were calculated using an aircraft design optimization tool OPERA. While it is recognized that the fraction fuel burn for each stage of the LTO cycle are not equal, it was found during the analysis that the errors produced through estimat- ing each stage fuel fraction as 0.994 was not significant and provided better results over estimating individual stage fractions (MacDonald 2012). The fuel fraction for the LTO cycle is hence given as per Equation (3.16). 𝑀33,12( = 0.95929 (3.17) This value is used for the following calculation of the Breguet factor. Using Equation (3.12) and simplifying Table 3.1, the following relationship is obtained: 𝑚' 𝑚& = 1 𝑀33 = 1 𝑀33,;$>:#9>167 ∙ 𝑀33,12( (3.18) In order to cover the entire flight, the mass ratio is adjusted to rely on the horizontal flight mass ratio. 31 𝑀33,12( ∙ 𝑚' 𝑚& = 1 𝑀33,;$>:#9>167 (3.19) Considering Equations (3.18) and (3.12): 𝐵 = 𝑅 ln B 𝑀33,12( ∙ 𝑚' 𝑚& C (3.20) The Breguet factor is used in the calculation of fuel consumption (in kg). It still applies: 𝑚. = 𝑚' − 𝑚& (3.21) Where m1 is the mass before takeoff and m2 is the mass after landing. The difference between the two parameters represents the actual amount (in kg) of fuel consumed. Taking Equation (3.11) and (3.20) into account, the result is. 𝑅 = 𝐵 ∙ ln D𝑚' + 𝑚. 𝑚& E (3.22) The conversion to mF results in: 𝑚.(𝑅) = 𝑚& ∙ D𝑒: ? − 1E (3.23) In summary: 𝐶()* = 𝑚. 𝑅' ∙ 𝑛()* [kg/km/seat] (3.24) A detailed description of Burzlaff's Excel tool is given in Chapter 4.3. In conclusion, the Bathtub Method generates a more accurate value than the methods presented in Chapters 3.2 and 3.3. It takes into account all flight phases during a mission. But this can also be a disadvantage, because the result depends on the mission. 32 3.5 CO2 Method So far, methods have been presented that require the payload-range diagrams are taken from Airport Planning Documents in order to calculate the fuel consumption. In this chapter, another method will be presented, which includes the CO2 emissions evaluation metric value. How this is composed and on which data basis it is based is described in Chap- ter 4.6. In this chapter, which describes the mathematical basis, the value is taken as given. It applies (ICAO Annex 16 2017). 𝐶𝑂2 𝑒𝑚𝑖𝑠𝑠𝑖𝑜𝑛𝑠 𝑒𝑣𝑎𝑙𝑢𝑎𝑡𝑖𝑜𝑛 𝑚𝑒𝑡𝑟𝑖𝑐 𝑣𝑎𝑙𝑢𝑒 (𝑀𝑉) = 1 𝑆𝐴𝑅 𝑅𝐺𝐹@.&B (3.25) Transforming the equation yields. 𝐶 = 1 𝑆𝐴𝑅 = 𝑀𝑉 ∙ 𝑅𝐺𝐹@.&B (3.26) The reference geometric factor (RGF) is an adjustment factor based on the aircraft fuselage size derived from a two-dimensional projection of the fuselage. It is dimensionless and standardized with regard to 1 m2. In addition, it is defined as follows (ICAO Annex 16 2017): • for aeroplanes with a single deck determine the area of a surface (expressed in m2) bounded by the maximum width of the fuselage outer mould line (OML) projected to a flat plane parallel with the main deck floor; and • for aeroplanes with an upper deck determine the sum of the area of a surface (expressed in m2) bounded by the maximum width of the fuselage outer mould line (OML) projected to a flat plane parallel with the main deck floor, and the area of a surface bounded by the maximum width of the fuselage OML at or above the upper deck floor projected to a flat plane parallel with the upper deck floor is determined; and • determine the non-dimensional RGF by dividing the areas defined in 1(a) or 1(b) by 1 m2. Furthermore, note that the RGF includes all pressurized spaces including corridors, auxiliary spaces, stairwells, passageways, and areas that can accommodate cargo. The cockpit crew area or crew rest/work areas are excluded. The pressure bulkhead is assumed to be the aft boundary. This results in four possibilities how RGF can be determined. All variants differ in their accu- racy and are arranged in such a way that they start with the least accurate variant and end with the most accurate variant. 33 Option 1: RGF = 𝑙C/; ∙ 𝑑. (3.27) Option 2: RGF = 𝑙;!=7E ∙ 𝑑. (3.28) Option 3: RGF = 𝑙C/; ∙ 𝑠 (3.29) Option 4: RGF = 𝑙;!=7E ∙ 𝑠 (3.30) According to the Aircraft Design Chapter 6 (Scholz 2015): 𝑙;!=7E = 𝑙C/; − 1.6 ∙ 𝑑. − 4m (3.31) And: 𝑠 = 2 ∙ X2 ∙ 𝑟 ∙ ℎ − ℎ& (3.32) Whereas: ℎ = 𝑟 − 1m (3.33) And: 𝑟 = 𝑑. 2 (3.34) In order to compare the specific fuel consumption with the other methods, this still needs to be divided by the typical seat configuration. In summary, this method does not require many pa- rameters, only the dimensions of the aircraft and the metric value. 34 3.6 Handbook Method The handbook method goes back to the equation of Breguet's range formula and allows an estimation of the variables. Already known is. B = 𝑉 ∙ 𝐸+!, 𝑐;: ∙ 𝑔 (3.35) And further: 𝐶 = 1 𝑆𝐴𝑅 = 𝑐 ∙ 𝑔 ∙ 𝑚 𝑉 ∙ 𝐸+!, ∙ 𝑛()* [kg/km/seat] (3.36) According to the International Standard Atmosphere (ISA) of the International Civil Aviation Organization (ICAO), the following conditions apply at mean sea level: 𝑇@ = 288.15 K (3.37) 𝑎@ = 340.294 m/s (3.38) Moreover, in the troposphere (at an altitude of 11km) there is a constant temperature of 𝑇''F+ = 216.65 K (3.39) From this, the speed of sound a at an altitude of 11 km can now be calculated: 𝑎''F+ = 𝑎@ ∙ \𝑇''F+ 𝑇@ = 295.07 m/s (3.40) Using Equation (3.38) and the Mach number Mcr of an aircraft, V can now be calculated: 𝑉 = 𝑀;: ∙ 𝑎''F+ (3.41) In the following, an estimation for E will be considered. The following relationship applies here (Scholz 2017): 𝐸+!, = 𝑘# ∙ _ 𝐴 𝑆G#H 𝑆/ [−] (3.42) 35 Where A is the wing aspect ratio of the wingspan to the wing area (Schlüter 2006): A = 𝑏/7E%& 𝑆/ (3.43) ke is composed as follows (Scholz 2023): 𝑘# = D3.229 ∙ 10>B 1 𝑘𝑚 ∙ 𝑅'E + 12.18 (3.44) The following factor is assumed for the ratio of 𝑆𝑤𝑒𝑡 𝑆𝑊 . 𝑆G#H 𝑆/ ≈ 6.1 (3.45) The following relationship for m holds: 𝑚 = 𝑚*2(/ + 𝑚*-./ 2 [kg] (3.46) The following factor applies to g: 𝑔 = 9.81 m/s^2 (3.47) Next cCR should be estimated. For this purpose, an equation from Hammami (2021) is used: 𝑐;: = 3.735 ∙ 10>J ∙ 𝜆>&.'&∙'@%& ∙ 𝑉 + 1.65 ∙ 10>L ∙ 𝜆>@.B [kg/sN] (3.48) Where 𝜆 is the bypass ratio of the respective aircraft and V is the cruise speed. However, Equa- tion (3.48) only applies to turbofan aircraft. To determine the fuel consumption of turboprop aircraft, another equation, based on the evalu- ations of Koppe (2012), should be added. 𝑐0 = 3.25369 ∙ 10>M ∙ ln (𝑃99<.#N ∙ 𝑂𝑃𝑅 ∙ 𝑇)2) ∙ 1.00060 ∙ 10>J [kg/sW] (3.49) Fuel consumption for turboprop can be calculated with an adjustment of Equation (3.36) as follows. 36 𝐶 = 1 𝑆𝐴𝑅 = 𝑐0 ∙ 𝑔 ∙ 𝑚 𝜂0 ∙ 𝐸+!, ∙ 𝑛()* [kg/km/seat] (3.50) The following factor applies to ηP. 𝜂0 = 0.82 (3.51) In conclusion, the fuel consumption of an aircraft can be determined using the handbook meth- ods. For this, seven parameters of each aircraft must be known: MTOW, MZFW, V, R1, wing- span bWing wing area SW and the bypass ratio λ. Since the max. glide ratio drops at high Mach numbers (not considered here), this method is only applicable for "common" cruise Mach num- bers (up to 0.89 or up to 263 m/s). In addition, as with the SAR method, the fuel reserve is assumed to be consumed. 37 4 Data Sources 4.1 Overview Chapter 3 presented various methods that can be used to calculate fuel consumption for aircraft in general. The mathematical relationships were explained, a reference unit was defined that allows the consumption figures to be compared with each other. In this chapter, it is explained how to obtain the input parameters to the methods given in Chap- ter 3. Furthermore, databases are presented that can be used to determine the fuel consumption of a specific aircraft. 4.2 Passenger Aircraft Census First the Top 50 Most Used Passenger Aircraft need to be identified. This definition refers to all commercial aircraft in operation. The World Airliner Census of 2020 provides an excellent database for this purpose. The World Airliner Census covers all commercial aircraft, both jet-powered and turboprop pow- ered, with a capacity of more than 14 passengers, that were in service in August 2020 (Hurtecant 2017). In April 2020, aviation came to an almost complete standstill due to the Corona pandemic and many aircraft were grounded at the time of the survey. However, since air traffic also rebounded significantly last year due to the lifting of Corona restrictions, the selection of aircraft assumes a full recovery. This means that the stored aircraft have rejoined the airline's fleet and are in service. However, the total number of aircraft considered in each case represent an inaccuracy. In the course of the Corona pandemic, some older aircraft models that were most unprofitable for the airlines, such as the Boeing 747-400, were flown out of service sooner than originally planned. This practice was seen at many of the world's airlines. The World Airliner Census does not take this into account, and it is difficult to conduct even a data collection of how many aircraft have actually left the fleets, as some decisions have not yet been finalized. Therefore, the following procedure was chosen using World Airliner Census data: • A full recovery and thus a full return of all aircraft was assumed. 38 • Due to increasing deliveries of certain aircraft types recently and increasing importance (i.e. growing deliveries) in aviation, the following aircraft were included: Airbus A220-300, Embraer E195-E2, Boeing 737 MAX9, Bombardier CRJ1000. • Accordingly, some other aircraft types have been downgraded, either because of smaller deliveries or missing data: Fairchild Metro/Merlin, Beechcraft 1900C, Beechcraft B99, BAe Jetstream 31. This now results in a listing that contains the 50 Most Used Passenger Aircraft based on the World Airliner Census figures. This represents 92.78 % of all aircraft as measured by the World Airliner Census. Table 4.1 List of the 50 most used passenger aircraft (World Airliner Census 2020) Aircraft Typ Total passenger Aircraft Ranking Boeing 737-800 4788 1 Airbus A320 4132 2 Airbus A321 1637 3 Airbus A319 1243 4 Airbus A320neo 1009 5 Boeing 737-700 979 6 Boeing 777-300ER 805 7 ATR 72 795 8 Airbus A330-300 707 9 Embraer E175 624 10 Bombardier CRJ100 601 11 Bombardier CRJ200 12 Boeing 737-900 556 13 Boeing 787-9 540 14 Airbus A330-200 502 15 Embraer E190 501 16 Embraer ERJ-145 479 17 Bombardier CRJ900 471 18 De Havilland Canada Dash 8 Q400 462 19 Boeing 777-200 391 20 Boeing 777-200ER 21 Boeing 767-300 365 22 Boeing 787-8 363 23 Airbus A321neo 355 24 Boeing 737 Max 8 347 25 Airbus A350-900 321 26 De Havilland Canada Twin Otter 315 27 Boeing 757-200 302 28 39 Bombardier CRJ700 291 29 Airbus A380-800 237 30 Boeing MD-80 232 31 Beechcraft 1900D 220 32 Boeing 737-300 214 33 ATR 42 208 34 Airbus A220-300 200 35 Saab 340 188 36 Boeing 737-500 161 37 Embraer E195 161 38 De Havilland Canada Dash 8 Q300 157 39 Embraer E170 157 40 De Havilland Canada Dash 8 Q100 152 41 Boeing 717-200 145 42 Boeing 747-400 142 43 Boeing 737-400 141 44 Sukhoi Superjet 100 131 45 Embraer EMB-120 Brasilia 127 46 Fokker 100 109 47 Bombardier CRJ1000 67 48 Embraer E195-E2 40 49 Boeing 737 MAX 9 38 50 TOTAL 27108 TOTAL World Airliner Census 29217 Average 92.78 % 4.3 Student Data As described at the beginning, the project work is based on the master thesis of Hurtecant (2021) and the project work of Burtlaff (2017). Hurtecant created an Excel spreadsheet that already contains important information for the fuel calculation. All of the top 50 most used passenger aircraft can be found here. In addition to the typical seat distribution, the Maximum Take-Off Weight or Minimum Zero Fuel Weight, the important parameters of the payload-range diagram (R1, m1, R2, m2) are also listed. This facili- tated the work during the project. This table was taken over completely and can be found in the Excel file in the tab "Hurtecant". 40 In his project work, Burzlaff developed an Excel tool for precise calculation of fuel consump- tion over all flight phases. In addition to the already known parameters from the payload-range diagram (R1, m1, R2, m2), the number of seats, Maximum Take-Off Weight (MTOW) and Max- imum Zero Fuel Weight (MZFW) for each of the 50 aircraft are also required. All parameters are part of the documents for airport planning. The weight per passenger is assumed to be 100 kg. The Mach number in flight is assumed to be 0.82 for widebody aircraft and 0.76 for short/medium aircraft. The range (for single calculation) corresponds to R1 (harmonic range). Figure 4.1 shows an excerpt from the Excel file, which illustrates the calculation using the A350-900 as an example: Figure 4.1 Extract from fuel calculator (Burzlaff 2017) The blue fields show the fuel required for the specified route (including reserve fuel) and the consumption in kg/100km per pax. The output has to be converted to kg/km/seat accordingly. 41 4.4 BADA The Base of Data (BADA) is a collection of ASCII files which specifies operation performance pa- rameters, airline procedure parameters and performance summary tables for 438 aircraft types. This information is designed for use in trajectory simulation and prediction algorithms within the domain of Air Traffic Management (ATM). All files are maintained within a configuration manage- ment system at the EUROCONTROL Validation Infrastructure Centre of Expertise located at the EUROCONTROL Experimental Centre (EEC) in Brétigny-sur-Orge, France (EUROCONTROL 2014). Although detailed data on aircraft performance are confidential and commercially sensitive, BADA transforms it such that aircraft models can be made available to a wide audience. The use of BADA is not limited to EUROCONTROL. The BADA family 3 model is provided to the ATM community worldwide and is used by R&D organizations, academic institutions, ANSP’s and ATM support in- dustry. Its use is free of charge, but regulated through a licence agreement that safeguards the in- terests of the aircraft manufacturers who are the principal aircraft performance reference data pro- viders for BADA (Nuic 2010) Unfortunately, it was not possible to obtain such access to view the data. Apart from that this database is a source that is not publicly accessible and therefore has been excluded from this survey. However, a document could be found on BADA, which represents a set of performance summary tables for 67 aircraft types which are modelled by BADA Revision 3.0 (EUROCONTROL 1998). This file contains a table of values which specifies the true air speed, rate of climb or descent and fuel flow for conditions of climb, cruise and descent at various flight levels. After the re- view of the document, it became apparent that only 19 of 67 aircraft types are relevant for this project, because only these are also in the list from Chapter 4.2. All other models mentioned are very outdated and do not belong to the 50 most used passenger aircraft. 42 Below there is an excerpt of the document depicting the Airbus A320: Figure 4.2 Extract from BADA (EUROCONTROL 1998) It shows a table of performance data with 13 columns. Each of these columns is explained below: Column 1 Flight Level (FL) Column 2 cruise True Air Speed (TAS) (nominal mass) in kg/min Column 3 cruise fuel consumption (low mass) in kg/min Column 4 cruise fuel consumption (nominal mass) in kg/min Column 5 cruise fuel consumption (high mass) in kg/min Column 6 climb True Air Speed (nominal mass) in knots Column 7 rate of climb with reduced power (low mass) in fpm Column 8 rate of climb with reduced power (nominal mass) in fpm Column 9 rate of climb with reduced power (high mass) in fpm Column 10 climb fuel consumption in kg/min Column 11 descent True Air Speed (nominal mass) in knots Column 12 rate of descent (nominal mass) in fpm Column 13 descent fuel (nominal mass) consumption in fpm Line 3 shows the type of aircraft. 43 Further explanatory notes on the data presented in the performance tables are given below (EUROCONTROL 1998): (a) Cruise data is only specified for flight levels greater than or equal to 30. (b) Performance data is specified up to a maximum flight level of 400 or to highest level for which a positive rate of climb can be achieved at the low mass. This maximum level is not necessarily the same as the maximum level specified in the OPF file and given in the PTF header. (c) True Air Speed for climb, cruise and descent is determined based on the speed schedules speci- fied in Sections 4.1, 4.2 and 4.3 respectively of the BADA User Manual [RD1]. (d) Rates of climb are calculated at each flight level assuming the energy share factors associated with constant CAS or constant Mach speed laws and using the reduced power corrections as given in Section 3.8 of RD1. (e) The fuel consumption in climb is independent of the aircraft mass and thus only one value is given. There are three different climb rates however corresponding to low, nominal and high mass conditions. (f) The rate of descent and fuel consumption in descent is calculated assuming the nominal mass. Values for other mass conditions are not given. (g) Discontinuities in climb rate can occur for the following reasons: change in speed between flight levels (e.g. removal of 250 knot restriction above FL100) transition from constant CAS to constant Mach (typically around FL300) transition through the tropopause (FL360 for ISA) (h) Discontinuities in descent rate can occur for the following reasons: transition through tropopause (FL360 for ISA) transition from constant Mach to constant CAS change in assumed descent thrust (specified by the BADA hdes parameter) change in speed between flight levels (e.g. application of 250 knot limit below FL100) For fuel consumption, FL360 is assumed for the A320 as an example. In addition, cruise fuel consumption (nominal mass) in kg/min (column 4) is considered. In Figure 4.2, this selection is marked with a red box. Subsequently, this value must also be converted to the unit kg/km/seat as follows in order to be able to compare the results. The values from BADA only take into account the fuel consumed in flight. 𝐶 = 𝐶𝑅𝑈𝐼𝑆𝐸E6+ ∙ 50 𝑉 ∙ 3 ∙ 𝑛()* [km/kg/seat] (3.34) 4.5 EEA The European Environment Agency (EEA) provides and manages another database that illus- trates the fuel consumption of aircraft. This is known as the Master Emission Calculator. All the information it contains is based on Eurocontrol's Fuel Burn and Emissions Inventory System (FEIS). This system estimates the total mass of fuel from all aircraft from the previous year that departed from or arrived at an airport in a relevant part of the territory of one of the 28 EU 44 Member States, or both (see EEA 2019). The FEIS uses Eurocontrol's Advanced Emission Model (AEM), which estimates fuel based on aircraft movements and, associated with that, the emissions that result from the combustion of the fuel (EEA 2019). The aim of this database is to allow the user to display the fuel consumed and the resulting pollutants emitted by entering a specific aircraft type and a specified route length. It has to be discussed which route length should be selected for the respective aircraft. For this it is important to understand that the stage length described by the EEA Master Calculator cor- responds to the length of the cruise phase. In order to be able to determine a suitable reference stage length for the respective aircraft, it is worth taking a look at the payload-range diagram of the aircraft. However, airlines operate their aircraft on routes that are sometimes well below the maximum mission range. This is also illustrated in Figures 4.3 and 4.4. Figure 4.3 presents a payload-range diagram for an Airbus A320-200 and a Boeing 737-800 com- bined with the number of flights the airplanes executed with a specific payload-range combination (Hurtecant 2021). Figure 4.4 presents a payload range diagram of a Boeing 787-8 combined with the number of flights the airplanes executed with a specific payload-range combination. It can be observed that almost no flight was operated with maximum payload or rather the maximum number of passengers. In addition, most of the time, the aircraft are used for relatively short routes of about 1000 NM, which deviates massively from the design range (Husemann 2018). Figure 4.3 Annual number of flights and payload-range diagram of an Airbus A320-200 (left) and a Boeing 737-800 (right) (Linke 2020) 45 Figure 4.4 Annual number of flights and payload range diagram of a Boeing 787-8 (Chai 2018) Hurtecant (2017) concludes that the selected reference route length is half of Harmonic Range (R1). He then compared the fuel consumption with that of the extended payload method. For him it was found that the fuel consumption for an entire flight is equal to the fuel consumption calculated with the extended payload-range diagram method multiplied by a factor of 0.84 with a standard deviation of 0.09. The factor 0.84 means in this case that the fuel consumption is smaller than the consumption calculated with Equation (3.9). This is due to the fact that Equa- tion (3.9) assumes the reserve fuel to be consumed, and also because within the range R1 the aircraft is not operated with maximum payload. In the table below, the values of Hurtecant are now to be verified, taking into account the 50 most used passenger aircraft. Since some aircraft types are unfortunately not included in the Master Emission Calculator, the number of aircraft considered is reduced to 36. 46 Table 4.2 Extended Payload Range Diagram vs. EEA Master Emission Calculator Aircraft Type Fuel Consumption, Extended Payload- Range Diagram (kg/km/seat) Fuel Consump- tion, EEA Master Emission Calcu- lator (kg/km/seat) Ratio of fuel con- sumption EEA/fuel consumption Pay- load-Range Dia- gram Airbus A319 0.02740 0.02160 0.78815 Airbus A320 0.02662 0.02174 0.81666 Airbus A321 0.02013 0.02177 1.08121 Airbus A330-200 0.03410 0.02749 0.80620 Airbus A330-300 0.02892 0.02172 0.75115 Airbus A350-900 0.02479 0.01880 0.75836 Airbus A380-800 0.02953 0.02595 0.87872 ATR 42 0.04085 0.03387 0.82916 ATR 72 0.03176 0.02778 0.87460 Boeing 717-200 0.04015 0.03103 0.77294 Boeing 737 Max 8 0.02128 0.02048 0.96239 Boeing 737 MAX 9 0.02079 0.01940 0.93319 Boeing 737-300 0.02983 0.02791 0.93539 Boeing 737-400 0.03126 0.02601 0.83228 Boeing 737-500 0.04393 0.03106 0.70693 Boeing 737-700 0.02946 0.02482 0.84230 Boeing 737-800 0.02714 0.02163 0.79705 Boeing 737-900 0.01778 0.02047 1.15127 Boeing 747-400 0.03430 0.02676 0.78018 Boeing 757-200 0.02939 0.02428 0.82598 Boeing 767-300 0.02937 0.02425 0.82555 Boeing 777-200 0.02798 0.02491 0.89013 Boeing 777-200ER 0.02416 0.02503 1.03589 Boeing 777-300ER 0.02961 0.02636 0.89025 Boeing 787-8 0.02714 0.02291 0.84402 Boeing 787-9 0.02576 0.02056 0.79789 Boeing MD-80 0.03389 0.03067 0.90512 Bombardier CRJ900 0.03794 0.02859 0.75350 Embraer E170 0.04440 0.03264 0.73521 Embraer E175 0.04097 0.03166 0.77279 Embraer E190 0.03881 0.03225 0.83099 Embraer E195 0.03423 0.02790 0.81514 Embraer E195-E2 0.02637 0.02563 0.97190 Embraer EMB-120 Brasilia 0.06207 0.05823 0.93802 Embraer ERJ-145 0.03980 0.03660 0.91983 Fokker 100 0.03902 0.03393 0.86943 Average 0.86 47 The calculation yields a factor of 0.86 with a standard deviation of 0.1. This proves Hurtecant's reasoning. Figure 4.5 shows a section of the Master Emission Calculator using the A320 as an example at a route length of 1048 NM (1941km) which corresponds to half of R1 (harmonic range) of the A320. Figure 4.5 Extract of EEA Master Emission Calculator by example of an A320 A distinction is made between two different flight phases. The first phase describes the landing and take-off phases and ends at 3000 feet. Below that high, because information about flight profiles is not sufficiently accurate, the AEM cal- culates the amount of fuel burnt according to the ICAO LTO cycle methodology, which was defined by the ICAO Engine Certification specifications, and models flight movements (below 3 000 ft) as a series of defined thrust levels for defined lengths of time associated with each LTO cycle flight phase. The fuel burn is calculated thanks to the ICAO Aircraft Engine Emissions Data Bank (AEED), which provides EIs and fuel flow for a very large number of aircraft engines. As EUROCONTROL has developed a table that lists a large range of aircraft models and the engines with which they are generally equipped, the AEM can link each flight movement processed by the AEM to a specific engine as identified by its Unique Identification Number (UID), as listed in the ICAO Engine Ex- haust Emissions Databank. (Eurocontrol 2019). The second phase includes climb, cruise and descent (CCD) and is above 3000 feet. Above that high, this profile is described in terms of a sequence of straight-line segments that are retrieved from the updated flight plan data managed by the EUROCONTROL Network Manager Operations Centre or partly or completely generated for flight movements outside of the EUROCONTROL zone of coverage. The fuel burn is calculated for each segment of a flight profile thanks to the aircraft performance information provided by EUROCONTROL’s BADA (EUROCONTROL, 2016b). This database provides altitude- and attitude-dependent performance and fuel burn data for more than 200 aircraft types. Once the amount of fuel burnt is calculated for 48 each segment, the Boeing Fuel Flow Method 2 (BFFM2) is used to correct the amount of fuel that is burnt before multiplying by the EFs (Eurocontrol 2019). Thus, looking at Figure 4.5, the fuel burn (6329.08 kg) can be determined (marked red). In order to be able to compare the consumption figures with each other, this key figure should also be converted: 𝐶()* = 𝑚.O#< =O$E 𝑅' ∙ 𝑛()* [km/kg/seat] (3.9) In summary, the EEA Master Emission Calculator offers a very good possibility to display the consumption and also the emissions clearly and easily. The consumption calculated based on EEA data includes the LTO cycle and the cruise flight and can therefore be considered as real- istic. 4.6 EASA Finally, and for the sake of completeness, the EASA Aeroplane CO2 Emissions Database is presented. Here, all CO2 emissions of aircraft certified by EASA are collected. For this purpose, a so-called CO2 metric value is determined. It is a specific air range (SAR)-based metric adjusted to take into account fuselage size. The CO2 metric value aims at measuring the technology performance of an aircraft type with respect to its fuel efficiency. During certification, it is compared to the applicable CO2 limit as defined in ICAO Annex 16 Volume III (EASA 2023). This database is to be expanded step by step over the next few years and provide a transparent insight into the CO2 emissions of aviation, in particular for specific aircraft. Therefore, all jet aircraft with a maximum take-off mass of more than 5.7 t and propeller aircraft with a take-off mass of more than 8.6 t must be certified according to the ICAO CO2 standard. From 2028, this is to apply to all aircraft in production. So far, data availability is still scarce. EASA data only show the A330-800 and A330-900 cer- tified to the new standard. In the attached Excel file a fuel calculation was performed using these two aircraft types, taking into account the metric value and Equations (3.25), (3.26) and Equations (3.30) through (3.34). Since the metric value has only been specified for two aircraft models to date, this method is calculated as an example but not compared with the other methods. In the coming years and 49 with the increasing number of ICAO-certified aircraft, this calculation may become more im- portant. 50 5 Data Analysis In order to apply the presented methods with information from the different databases for the calculation of fuel consumption to the 50 most used passenger aircraft, an Excel file has been generated. With a normalised unit each consumption can be compared. After a short introduction, the structure of the Excel file will be explained, and the different interpretation methods will be compared. 5.1 Overview The Excel file contains 8 Methods that can be used to determine the fuel consumption for each of the 50 aircraft: Method 1 Fuel Consumption, SAR Method 2 Fuel Consumption, Extended Payload-Range Diagram Method 3 Fuel Consumption Bathtub Curve (Harmonic Range) Method 4 Fuel Consumption, EEA Master Emission Calculator Method 5 Fuel Consumption, BADA Method 6 Fuel Consumption, Handbook Methods Method 7 Fuel Consumption, Literature Review Method 8 Fuel Consumption, Metric Value As data for the fuel consumption based on metric value is only available for 2 aircraft it’s listed separately in tab "Fuel Consumption CO2 MV A330neo". Furthermore, the Excel file can be divided into 13 tabs. All tabs are described in more detail in Table 5.1. 51 Table 5.1 Overview of Excel tabs with explanation Tab Explanation Overview At a glance, all 50 aircraft and their consumptions are listed using the above calculation methods. In addition, the average, standard deviation and the coefficient of variation for each calculation are given and shown graphically. SAR Each tab lists the fuel consumption and the respective input data. Values taken out of the databases are marked in green. Extended Payload Diagram Bathtub Curve EEA Master Emission Calculator BADA Handbook Method Literatur review Fuel Consumption CO2 MV A330neo Fuel consumptions for A330-800 and A330-900 calculated using the Metric Value with different RGF and Maximum Take-Off Weights. Deviation Differences of fuel consumption for each aircraft typ and method are listed. In addition, the coefficient of determination and the mean absolute percentage error (MAPE) are determined. Matrix The results from the tab "Deviation" are shown here in the form of a matrix. Hurtecant Most of the metrics relevant to the calculations are based on Hur- tecant's Excel file. All necessary information was taken from this table and applied to the respective aircraft types. TOP 50 Most Used Aircraft In this table, you can find a ranking of which aircraft are most com- monly used in aviation. 5.2 Comparison of Methods The methods described in Section 5.1 differ in their approach to determining fuel consumption, this is why the determined fuel consumption also differs. Method 1 The fuel consumption is obtained from the slope in the payload range diagram for a given distance in cruise flight. The additional fuel consumption caused by takeoff, climb, descent, landing is not considered. Hence, the consumption per km is lower than an average consumption of a flight. 52 Method 2 In this method, the fuel reserves are included so that the consumption in kg/km/seat is higher than an average consumption of a flight. Method 3 The method considers the flight phases takeoff, climb, descent, landing with mission segment mass fractions, why realistic values can be expected. The selected range (Harmonic Range; R1), i.e., the "range at maximum payload", is comparatively generous for consideration. In practice, the range is still be- low the Harmonic Range. Method 4 The consumption calculated via EEA includes the LTO cycle (landing, take- off phase) as well as the cruise flight. The consumption can be regarded as very realistic and practical. Method 5 The data from BADA only takes into account the cruise flight. The resulting consumption does not cover the LTO cycle. Method 6 Consumption was determined using Handbook methods, with the Breguet factor and SAR. Mission Segment Mass Fractions were not considered here. It is therefore a consumption per km in cruise flight. Method 7 The researched consumption figures are based on trade magazines or the au- thor's own observations. Since in most cases a reference route is explicitly indicated or the consumption actually refers to flown routes, it can be con- cluded that the indicated consumption includes not the fuel reserves. It can also be assumed that all flight phases are included here. Method 8 The RGF and the Metric Value are needed to calculate the fuel consumption. It is not clear from the sources whether the consumption includes the LTO cycle or only refers to the cruise flight. It is also unclear whether the con- sumption includes fuel reserves. Nonetheless, this method may prove very useful in future years when multiple aircraft have been recorded. Comparing the different methods there are at least 2 categories of methods: Method 1, 2, 5 and 6 determine the fuel consumption only for the cruise flight, whereas method 3, 4 and 7 refer to the full mission cycle incl. LTO providing basically more realistic data. Apart from that method 4,5 and 7 include real flight data. The following subsections will now compare the different results of each calculation method. 53 5.2.1 Average Analysis The mean values of the consumption data of all methods over all selected aircraft differ. This can be explained by the different assumptions listed in Chapter 5.2. Figure 5.1 now shows the average consumptions of each calculation method. Each figure corresponds to the procedure listed in Chapter 5.1. Figure 5.1 Results of the average fuel consumption of all methods (left) and the ranking (right) It is particularly striking that the calculation according to Extended Payload-Range Diagram has the highest average fuel consumption. However, this can be explained by the complete consumption of the reserve fuel. The consumption in kg/km/seat is thus higher than an average consumption in flight. It is important to mention that the methods cannot be compared with each other, because the average refers to all recorded aircraft of a calculation. However, it is quite possible that in one method there is no information about a certain aircraft type, but in another there is. In the end, only aircraft types and thus their consumption can be compared if they occur in both types of calculation. 5.2.2 Coefficient of Variation Analysis First, a standard deviation should be determined for all 7 variants. The standard deviation de- pends on the mean value. Large values (in absolute terms) have a larger standard deviation than smaller values. Therefore, a relative standard deviation must be used for a comparative evalu- ation. The so-called coefficient of variation. 54 The coefficient of variation depends on the selection of aircraft. The list of the top 50 most used passenger aircraft covers all sorts of aircraft types (e.g., short-haul, long-haul, fuel-efficient, inefficient aircraft). The coefficient of variation for each method refers to almost the same types of aircraft. Therefore, the coefficients are comparable. They should be similar among the vari- ants. Large deviations are to be regarded as flaws. Figure 5.2 first shows the distribution of the standard deviation for all methods. This is followed by the distribution of the coefficient of variation in Graph 5.3. Figure 5.2 Results of the standard deviation of all methods (left) and the ranking (right) Figure 5.3 Results of the coefficient of variation of all methods (left) and the ranking (right) It can be seen that all methods except Extended Payload-Range Diagram consumption have a very similar coefficient of variation between 0.204 and 0.307. On average, this is a value of 0.26 among these methods. Only Method 2 shows a much higher coefficient of variation of 1.86. Thus, the values of Method 2 scatter much more. As already mentioned, Method 2 has a higher average consumption. The standard deviation is also significantly higher for Method 2 than for the others (see Figure 5.3). While the standard deviation for the other 6 methods is in 55 the range of thousandths (0.0061 on average), the standard deviation for Method 2 is in the range of hundredths (0.082). 5.2.3 Coefficients of Determination Analysis The coefficient of variation does not yet indicate in which direction the deviations from the average fuel consumption of the aircraft move. If in Method A an aircraft has a consumption value significantly above the mean value and this is also the case in Method B, then both meth- ods show the same behavior. Based on this, each method was compared to each other, and their coefficient of determination was determined. The results can be seen in a matrix of size 7x7. The diagonal of the matrix is by definition assigning the value 1, because the coefficient of a vector with itself is 1 (complete agreement). The matrix is symmetrical with respect to the diagonals. The coefficient of determination should be as high as possible (ideally 1.0). Therefore, it is good if the mean of all coefficients (of a method with all other methods and with itself) is as high as possible. Figure 5.4 shows the described 7x7 matrix and its mean values: Figure 5.4 7x7 Matrix of the coefficient of determination The mean values of these coefficients vary between 0.237 (low) and 0.446 (acceptable). The mean value of these mean values is 0.33. The mean values of the determination coefficients are listed again in the following diagram: Coefficient of determination Fuel Consumption SAR (kg/km/seat) Fuel Consumption Extended Payload Range Diagram (kg/km/seat) Fuel Consumption Bathtub Curve (Harmonic Range) (kg/km/seat) Fuel Consumption EEA Master Emission Calculator (kg/km/seat) Fuel Consumption BADA (kg/km/seat) Fuel Consumption, Handbook Methods (kg/km/seat) Fuel Consumption Literature review (kg/km/seat) Average Fuel Consumption SAR (kg/km/seat) 1 0.61879 0.04668 0.39468 0.02287 0.10377 0.52371 0.38721 Fuel Consumption Extended Payload Range Diagram (kg/km/seat) 0.61879 1 0.00754 0.85450 0.02415 0.04278 0.57103 0.44554 Fuel Consumption Bathtub Curve (Harmonic Range) (kg/km/seat) 0.04668 0.00754 1 0.00134 0.40219 0.27233 0.00004 0.24716 Fuel Consumption EEA Master Emission Calculator (kg/km/seat) 0.39468 0.85450 0.00134 1 0.05636 0.00440 0.35259 0.38055 Fuel Consumption BADA (kg/km/seat) 0.02287 0.02415 0.40219 0.05636 1 0.20456 0.03728 0.24963 Fuel Consumption, Handbook Methods (kg/km/seat) 0.10377 0.04278 0.27233 0.00440 0.20456 1 3.412E-02 0.23742 Fuel Consumption Literature review (kg/km/seat) 0.52371 0.57103 0.00004 0.35259 0.03728 0.03412 1 0.35982 Average 0.38721 0.44554 0.24716 0.38055 0.24963 0.23742 0.35982 Average of Average: 0.32962 56 Figure 5.5 Results of the average coefficient of variation of all methods (left) and the ranking (right) Method 2 thus has the highest and thus best value. This is in contradiction to the evaluation of the coefficient of variation. 5.2.4 Regression Coefficient Analysis as Design Matrix The matrix of regression coefficients can also be understood as a Design Structure Matrix (DSM). The design structure matrix (DSM) is useful in identifying where iteration is necessary (Smith 1995). The philosophy of the DSM method is that the design project is divided into individual tasks, and the relationships among these tasks can be analyzed to identify the underlying structure of the pro- ject. (Niţă 2010) One type of evaluation is the "Eigenstructure Analysis". The eigenvalues and eigenvectors determine the nature of the convergence of the design process in a similar way with the aircraft dynamics: • the eigenvalues give information about the rate of convergence, • the eigenvectors give information about the shape of the natural motion (Niţă 2010). Therefore, also here, with the help of the 7x7 matrix from Figure 5.4, we search for the highest eigenvalue and the corresponding eigenvector. The highest eigenvalue is 2.69 and the eigen- vector has the values for the 7 methods: 57 Figure 5.6 Results of the values of the highest eigenvalue of all methods (left) and the ranking (right) The figure shows that the ranking of the methods is different to the evaluation of the regression coefficients. 5.2.5 Mean Absolute Percentage Error (MAPE) Analysis Even though MAPE is defined as "absolute", MAPE is actually a relative value because the amount of deviation is divided by the (correct) value. As with the regression coefficient, a ma- trix is generated. On the diagonal, the value is 0 because the deviation with itself is zero. The MAPE should be as small as possible (ideally zero). The aim is to use the mean value of all MAPE (one method compared to all other methods and to itself) as small as possible. Figure 5.7 shows the 7x7 matrix for the mean absolute percentage error and its mean value. Figure 5.7 7x7 Matrix of MAPE MAPE Fuel Consumption SAR (kg/km/seat) Fuel Consumption Extended Payload Range Diagram (kg/km/seat) Fuel Consumption Bathtub Curve (Harmonic Range) (kg/km/seat) Fuel Consumption EEA Master Emission Calculator (kg/km/seat) Fuel Consumption BADA (kg/km/seat) Fuel Consumption, Handbook Methods (kg/km/seat) Fuel Consumption Literature review (kg/km/seat) Average Fuel Consumption SAR (kg/km/seat) 0 0.82384 0.22078 0.32704 0.26581 0.18693 0.29879 0.35387 Fuel Consumption Extended Payload Range Diagram (kg/km/seat) 0.82384 0 0.30445 0.15603 0.39199 0.32374 0.20579 0.31512 Fuel Consumption Bathtub Curve (Harmonic Range) (kg/km/seat) 0.22078 0.30445 0 0.17058 0.28160 0.17034 0.21715 0.19499 Fuel Consumption EEA Master Emission Calculator (kg/km/seat) 0.32704 0.15603 0.17058 0 0.24604 0.22915 0.12927 0.17973 Fuel Consumption BADA (kg/km/seat) 0.26581 0.39199 0.28160 0.24604 0 0.24235 0.52979 0.27965 Fuel Consumption, Handbook Methods (kg/km/seat) 0.18693 0.32374 0.17034 0.22915 0.24235 0 0.33025 0.21182 Fuel Consumption Literature review (kg/km/seat) 0.29879 0.20579 0.21715 0.12927 0.52979 0.33025 0 0.24443 Average 0.35387 0.31512 0.19499 0.17973 0.27965 0.21182 0.24443 Average of Average: 0.25423 58 The mean values of MAPE vary between 0.18 and 0.354. The mean value of these averages is 0.254. The following figure shows the ranking of the mean values of the methods. Figure 5.8 Results of MAPE of all methods (left) and the ranking (right) It can be seen that the EEA Master Emission Calculator method performs best here. 5.3 Comparison of Data In Chapter 5.2, the methods were compared with each other. What is missing is a comparison that includes the top 50 most used aircraft. First, the 50 aircraft are divided into three categories. • Turboprop Aircraft • Narrow Body Aircraft • Wide Body Aircraft In the following three graphs, consumption is plotted against the individual aircraft in the group. Each curve represents a method. Due to erroneous data, the De Havilland Canada Twin Otter was not included in any of the following evaluations. 59 Figure 5.9 Comparison of fuel consumption of turboprop aircraft The consumption figures for turboprop aircraft are widely scattered, see Figure 5.9. For the EMB-120 the max. value is roughly 400 % of the minimum value, for the De Havilland Canada Dash 8 Q400 it is 500%. This is probably due to an imprecise data basis. On the contrary data for the Saab 340 and ATR42 are closer to each other, an indication of more mature data. Given these scattered data it’s difficult to determine the turboprop with the best fuel consumption. In order to better compare data an average fuel consumption out of SAR, Extended Payload Range Diagram and Bathtub Curve has been calculated for all aircraft. The aircraft types are ordered in Fig.5.9 starting with the lowest average fuel consumption first. Accordingly, the results are given for narrow and wide body aircraft as well. 0.00000 0.01000 0.02000 0.03000 0.04000 0.05000 0.06000 0.07000 Saab 340 De Havilland Canada Dash 8 Q400 ATR 72 De Havilland Canada Dash 8 Q300 De Havilland Canada Dash 8 Q100 Embraer EMB- 120 Brasilia ATR 42 Beechcraft 1900D SAR (kg/km/seat) Extended Payload-Range Diagram (kg/km/seat) Bathtub Curve (Harmonic Range) (kg/km/seat) Average SAR, Extended Payload-Range Diagram, Bathtub Curve (Harmonic Range) (kg/km/seat) EEA Master Emission Calculator (kg/km/seat) BADA (kg/km/seat) Handbook Methods (kg/km/seat) Literature Review (kg/km/seat) 60 Figure 5.10 Comparison of fuel consumption of narrow body aircraft First, the fuel consumption of different methods is less scattered compared to turboprop aircraft, see fig. 5.10. The extended payload range diagram method reveals substantial peaks for some aged aircraft (e.g., Boeing 717-200, Boeing 737-500, Bombardier CRJ100, Embraer E175/ERJ145) being conservative for all aircraft types. As MTOW and MZFW are precisely known from the certification documents the peaks could only originate from imprecise design point data (m1, R1). Following the principle comparison in Chapter 5.2 Method 3 and 4 (i.e. Bathtub Curve, EEA) give the more realistic values. They are quite close to each other except for the Embraer aircraft. Unfortunately, some values are missing for Method 4. Apparently, the values for CRJ100/200 scatter widely and cannot be assessed here further. The comparison of the most used passenger aircraft A320 and B737-800 shows an equal fuel consumption (difference is less than 0.5% according to EEA) although the B737-800 (1st flight 1997) was introduced ca. 10 years after the A320. Then, Airbus developed the A320neo (1st 0.00000 0.00500 0.01000 0.01500 0.02000 0.02500 0.03000 0.03500 0.04000 0.04500 0.05000 Boeing 737-900 Airbus A321neo Boeing 737 MAX 9 Airbus A320neo Airbus A321 Boeing 737 Max 8 Airbus A220-300 Bombardier CRJ1000 Embraer E195-E2 Airbus A320 Airbus A319 Boeing 737-800 Embraer E195 Boeing 737-700 Bombardier CRJ700 Boeing 737-300 Boeing 757-200 Boeing MD-80 Boeing 737-400 Sukhoi Superjet 100 Bombardier CRJ100 Bombardier CRJ900 Bombardier CRJ200 Embraer E190 Fokker 100 Embraer E175 Boeing 717-200 Embraer ERJ-145 Embraer E170 Boeing 737-500 SAR (kg/km/seat) Extended Payload-Range Diagram (kg/km/seat) Bathtub Curve (Harmonic Range) (kg/km/seat) Average SAR, Extended Payload-Range Diagram, Bathtub Curve (Harmonic Range) (kg/km/seat) EEA Master Emission Calculator (kg/km/seat) BADA (kg/km/seat) Handbook Methods (kg/km/seat) Literature Review (kg/km/seat) 61 flight 2014) with ca. 15% less fuel consumption compared to the A320 according to the bathtub figures (A320: 0.02093 kg/km/seat; A320neo: 0.01758 kg/km/seat). The better efficiency of the A320neo has been achieved by aerodynamic measures (e.g., winglets), engine improve- ments and an extension of the seat capacity from 150 to 165. Boeing answered the A320neo market success with the B737 MAX 8 (1st flight 2016) as the direct competitor. Both the bathtub method and the literature review show very similar fuel consumptions (EEA data is missing for the A320neo). But the B737 MAX 8 suffered then from the 2 accidents and a long-lasting pro- cess to regain certification. Further improvements are presently being achieved with configura- tion changes, i.e., with stretched versions, thus entering another market segment (B737- MAX9). The B737 MAX9 and the A321neo (with similar seat capacities) have similar average values, whereas the average value of the B737-900 is the lowest. As can be seen fuel efficiency is an important driver for the aircraft market that have been improved by configuration (e.g., aerodynamics, seats) and engine changes. Figure 5.11 Comparison of fuel consumption of wide body aircraft First, fuel consumption of wide body aircraft is also quite scattered for many aircraft types (e.g., A330-300: 100% increase from min. to max. fuel consumption), see Figure 5.11. Again, the Extended Payload-Range method shows very conservative, the SAR method very optimistic (i.e. low) fuel consumptions for the cruise flight. As only the Extended Payload-Range method includes the fuel reserve the consumption is significantly higher. The comparison of the A350- 900 and B747-400 yields a higher fuel consumption of the Boeing aircraft. The reason is an 0.00000 0.00500 0.01000 0.01500 0.02000 0.02500 0.03000 0.03500 0.04000 Airbus A330-900 Airbus A350-900 Boeing 777-200ER Boeing 787-9 Airbus A330-300 Boeing 787-8 Boeing 767-300 Boeing 777-200 Airbus A380-800 Boeing 777-300ER Airbus A330-200 Boeing 747-400 SAR (kg/km/seat) Extended Payload-Range Diagram (kg/km/seat) Bathtub Curve (Harmonic Range) (kg/km/seat) Average SAR, Extended Payload-Range Diagram, Bathtub Curve (Harmonic Range) (kg/km/seat) EEA Master Emission Calculator (kg/km/seat) BADA (kg/km/seat) Handbook Methods (kg/km/seat) Literature Review (kg/km/seat) 62 aged B747 with 4 engines is compared with the newer A350 with 2 engines. The direct com- petitors A350-900 and B787-8 show less fuel consumption for the Airbus aircraft. Overall, the Airbus A330-900(neo) together with A350-900 and B777-200ER offer the lowest fuel con- sumption. It has to be noted that the SAR value of the A330-900 is significantly lower than the others, thus lowering the average value. Overall, a conclusion on absolute figures is difficult to make because of the scattered data. It would have been helpful to get access to OEM data. It can be seen that the consumption figures for turboprop aircraft are more widely scattered than those for wide-body aircraft, for example. This is due to the data basis and the more difficult calculation methods (bypass ratio for propellers difficult to determine). However, it can be stated for all three diagrams that the Extended Payload-Range Diagram method is probably the most conservative method. The method according to EEA hardly represents the turboprop aircraft, but it lies in the middle of the other curves for narrow body aircraft as well as for wide body aircraft and has a good data coverage. Figure 5.12 shows the consumption figures for the A330-900. Here, Method 8 (CO2 MV A330neo) is to be compared with the other methods. As mentioned in Chapter 4.6, this calcu- lation method is based on only one aircraft type (A330neo in two configurations). It can be seen that the average value of method 1-3 is very close to the value of method 8 and thus has the lowest value of deviation. Method 2, on the other hand, has the largest deviation. Figure 5.12 Comparison of fuel consumption of A330-900 A330-900 Analysis Method 1 (SAR) Method 2 (Extended Payload- Range Diagram) Method 3 (Bathtub Curve (Harmonic Range)) Average Method 1,2,3 (SAR, Extended Payload-Range Diagram, Bathtub Curve (Harmonic Range)) Method 4 (EEA Master Emission Calculator) Method 5 (BADA) Method 6 (Handbook Methods) Method 7 (Literature Review) Method 8 (CO2 MV A330neo) Fuel Consumption (kg/km/seat) 0.01410 0.02633 0.02160 0.02068 no data avaiable no data avaibale 0.02361 no data avaibale 0.01897 Deviation to Method 8 (CO2 MV A330neo) -0.25690 0.38764 0.13843 0.08972 no data avaiable no data avaibale 0.24457 no data avaibale 63 6 Summary 6.1 Summary Methods A total of 8 calculation methods have been presented. For the Metric Value method 8 values are only available for one single aircraft type (Airbus A330neo in two configurations). For this reason, this method has not been considered further in the statistical evaluation. For Method 1 (Fuel Consumption, SAR) four parameters (R1, m1, R2, m2) are required. These are parameters that directly describe the consumption in cruise flight. Usually the parameters are determined by a payload range chart. A payload range chart is not publicly available for every aircraft. Method 2 (Fuel Consumption Extended Payload-Range Diagram) shows consumption values that are clearly too high. A general correction of the values with a factor (of 0.86 as calculated here) should be avoided. However, Method 2 has the advantage that only three input parameters are required. Possibly, the method can be used to compare aircraft among each other. Method 3 (Bathtub Curve) requires, like Method 1, the four parameters R1, m1, R2, m2 as well as MTOW and MZFW. The method considers the flight phases takeoff, climb, descent, landing with mission segment mass fractions, therefore the values are to be considered realistic. In the comparisons under Chapter 5.2, Method 3 does not stand out particularly. Overall, Method 4 (EEA) delivers very good values. However, data could only be provided for aircraft that are contained in the database. For Method 5 (BADA) only a few (older) values were publicly available. BADA requires a license and is therefore not publicly available. It is not allowed to publish consumption values of individual aircraft. Method 6 (Fuel Consumption, Handbook Methods) provi