Calculation of in-flight cruise performance for integration in an

EFB System

Tiago Torretiago.torre@tecnico.ulisboa.pt

Instituto Superior Técnico, Universidade de Lisboa, Portugal

December 2016

Abstract

Information technology is slowly but surely revolutionizing commercial aviation operations, inand out of the cockpit. Manual processes and paper documentation are gradually being replaced bycomputerized systems, thanks to the increasing computational power and portability of electronicdevices. Flight Crew members use portable electronic devices, known as Electronic Flight Bags (EFB),which unlock a higher operational safety and efficiency. The use of EFB systems by airline companieshas been growing significantly in recent years. It unleashes the potential of aircraft performance toolsand critical documentation by making it portable and accessible to flight crew members, without theneed to carry large or heavy bags. The present work focuses on the development of a computationaltool that allows flight crew members to compute emergency descent profiles in real time, taking themost recent meteorological information into consideration. Atmospheric information released before aflight is prone to change. This implies that safer emergency routes are likely to arise as atmosphericdata gets updated. This work helps the airline take advantage of those atmospheric updates, andmake flight operations safer. The application allows the user to compute emergency descent profiles fordepressurization and engine failure scenarios. By comparing the computed profile against minimumflyable altitudes, the application can rapidly verify if the calculated flight path satisfies regulatoryrequirements for the targeted emergency procedure. Integrated in TAP’s EFB solution, this tool couldprove to be a game changer in how escape routes are managed.Keywords: Electronic Flight Bag, Emergency Descent, Aircraft Performance, Engine Failure,Depressurization

1. IntroductionThere’s a need for a more dynamic processing ofatmospheric data. As weather updates get releasedduring the course of the flight, it would be benefi-cial to verify if the initially planned escape routesremain the best option. Flight crew would greatlybenefit from a tool with which they can automat-ically scan the remaining route for the best emer-gency descent procedures. The development of sucha tool and the computational methodology thatsupports it is the main motivation for this work.

1.1. Electronic Flight BagAccording to European Aviation Safety Agency’s(EASA) AMC 20-25, EFB is defined as ”An in-formation system for flight deck crew memberswhich allows storing, updating, delivering, display-ing, and/or computing digital data to support flightoperations or duties” [1]. It seeks to replace theoriginal flight bag, a heavy device that carriesprinted documentation that pilots need while theyoperate an aircraft. Flight bags at TAP used to

weigh around 20kg before being gradually replacedby its digital counterpart.

The first significant advantage of the EFB isthe significant reduction of paper-based documen-tation, allowing it to be accessed digitally instead.The second one is the potential for increased opera-tional efficiency and safety, through the deploymentof custom computational applications.

EFBs have evolved from installed Global Posi-tioning System (GPS) aircraft devices with addi-tional features built into them, to powerful and all-round portable devices like tablets, available today.The current versatility and potential of EFB sys-tems is very significant, and translates into a pos-itive impact on the airline’s balance sheets at theend of the year, as on its environmental footprintand operational safety [2].

2. Emergency Descent OperationsTwo types of cruise failure scenarios will be consid-ered in the present work. The first is the failure ofaircraft pressurization systems, also called depres-

1

surization. The second is engine failure, leading toa drift down procedure.

2.1. Leveled Flight MechanicsThe lift and drag equations for level flight are asfollows [3]:

L = W = m · g = 12ρ (TAS)2S CL (1)

D = T =1

2ρ (TAS)2S CD (2)

Where L is the lift force, W is the weight, m isthe mass, g is the gravitational constant, ρ is theair density, TAS is the True Air Speed, S is thearea of the lifting surface, CL is the lift coefficient,D is the drag force, T is the engine thrust, and CDis the drag coefficient.

2.2. Descent and Drift Down PerformanceWhen an aircraft is cruising and at least one of itsengines becomes inoperative (OEI), it must descendto a new altitude. This is called a drift down ma-neuver. A drift down is an un-accelerated descentwhich occurs while an airplane descends from itsall-engines operational altitude to a lower altitudewhere the available thrust from its remaining en-gine(s) is enough for level-flight [3]. Figure 1 showsthe acting forces in a drift down flight condition.

Figure 1: Forces and angles - Drift Down Descent[3]

Where Xs is the stability X-axis which pointsthrough the center of gravity (CG) and along asteady-state velocity vector; Xb is the body fixedX-axis which points through the CG and along anarbitrary line; α is the angle of attack defined as theangle between Xb and Xs; θ is the airplane pitch at-titude angle, defined as the angle between Xb andthe horizon; γ is the airplane flight path angle, alsocalled climb angle, defined as the angle between Xsand the horizon; φT is the thrust inclination angle,defined as the angle between the resulting thrustdirection and Xb.

The equations of motion along the flight path andperpendicular to the flight path can be taken fromFigure 1 and are as follows [3]:

T cos (α+ φT )−D +W sin (γ̄) = 0 (3)

T sin (α+ φT ) + L+W cos (γ̄) = 0 (4)

Most descents are performed with small descentgradients. Let the thrust inclination, φT , the angleof attack, α, and the flight path angle, γ, be suf-ficiently small so that following simplifications arepossible [3]:

sin (α+ φT ) ≈ 0 cos (α+ φT ) ≈ 1sin(γ) ≈ γ cos(γ) ≈ 1

(5)

With these assumptions, Eqs. (3) and (4) be-come:

T −D +W · γ̄ = 0 (6)

L = W (7)

The expression for the descent angle is, from Eq.(6):

γ̄ =D − TW

(8)

In a drift down maneuver, the remaining en-gine(s) are kept at Maximum Continuous Thrust(MCT). This is the maximum thrust that can beused unlimitedly in flight. For this case, Eq. (8)becomes:

γ̄ =DOEI − TOEI

W=TreqOEI − TavOEI

W(9)

Where TreqOEI corresponds to the thrust requiredto overcome the Drag force acting on the aircraft,and TavOEI is the available thrust of the remainingengine(s) at MCT setting.

The rate of descent, RD, is defined as:

RD = −dhdt

= V · sin γ̄ ≈ V · γ̄ (10)

Inserting this relation in Eq. (8) yields:

RD =(D − T ) · V

W(11)

For drift down condition, Eq. (11) becomes:

RD =(TreqOEI − TavOEI ) · V

W(12)

A drift down descent is carried on until the air-craft has reached the so called drift down ceiling.Initially, when the drift down begins, the availablethrust is not enough to balance the Drag force of theaircraft (TreqOEI > TavOEI ). Inserting this result inEqs. (9) and (12), one can see that the aircraftstarts descending:

γ̄ > 0 and RD > 0

2

Recalling Eqs. (2) and (9) one can write:

TreqOEI =1

2ρ (TAS)2S CD (13)

Due to the lack of thrust, the Drag force causes theaircraft decelerate, and to reduce its TAS. FromEq. (13), TreqOEI also diminishes. Eventuallythe required thrust balances the available thrust(TreqOEI = TavOEI ). From Eqs. (9) and (12):

γ̄ = 0 and RD = 0

When this happens, the aircraft has reached thedrift down ceiling, and stops descending further.The drift down ceiling is the maximum altitude thatcan be flown in level flight, at green dot speed . TheGreen Dot Speed is the speed for which the lift-to-drag ratio (L/D) is maximum, and corresponds tothe minimum descent angle [4].

2.3. Depressurization ProcedurePressurization systems ensure that the pressure in-side the cabin is high enough for the air to bebreathable. When a depressurization occurs, it isassumed that the pressure inside the aircraft is thesame as the atmospheric pressure outside. Usuallythe altitude at which airplanes cruise doesn’t con-tain sufficient oxygen concentration to allow a hu-man to breathe. Therefore, supplemental oxygensystems have to be installed, which are activated inthese situations. An important aspect to keep inmind is that the oxygen supply of these systems islimited. This means that the aircraft has a limitedtime to reach 10000ft, where atmospheric pressureis considered to be high enough to allow safe breath-ing [4].

While descending, the aircraft has to follow a pro-file that satisfies the constraints published by themanufacturer in the Flight Crew Operating Man-ual (FCOM). An example is illustrated in Figure2, where the red line corresponds to the FCOM’sdepressurization profile. The aircraft has to be ator below this profile when descending along the de-pressurization flight path.

Figure 2: Oxygen vs. Performance Profile of A319- 21min oxygen system [4]

Albeit serving as a guide, the above profile de-picted in Figure 2 doesn’t take the performance

limitations of the aircraft into account. It plots thealtitude in relation to the elapsed time. This is onlydependent on the oxygen system’s capability. Onlyby first converting it into a distance-based profilecan one depict the aircraft’s descent capability.

This means that the real profile will look morelike the one in green, depicted in Figure 2. Theaircraft may have to start descending earlier thenpredicted in the FCOM profile, to achieve the vari-ous time constraints.

To ensure that the aircraft remains as high aspossible for the longest possible distance, two mea-sures are usually applied: Descent branches are per-formed at MMO/VMO, with extended airbrakes, foran increased rate of descent; Cruise branches areperformed at MMO/VMO. This also ensures thatthe number of possible escape route is maximized.MMO/VMO is the maximum operating Mach

number/Speed, which may not be exceeded in anyregime of flight.

2.4. Drift Down ProcedureWhen at least one engine fails during normal op-eration, an emergency procedure is conducted withthe goal of safely landing the aircraft at a near aero-drome. The thrust of the remaining engine(s), whenone (or more) of them become inoperative, is set toMaximum Continuous Thrust (MCT) setting. Theavailable thrust at MCT is not enough to matchthe opposing drag force and to maintain the designcruise scenario. The aircraft must reduce its speedand descend to a new altitude (drift down ceiling)where level flight is possible.

For conservative purposes, the flight path consid-ered in aviation regulation for engine failure situa-tions doesn’t correspond to the path actually flownby the aircraft. The path flown by the aircraft, orgross flight path, has to be penalized by a givengradient penalty. This yields the so called net flightpath, which is the one used for obstacle clearanceverification purposes. An example can be seen inFigure 3.

Figure 3: Gross and Net Drift Down Descent FlightPath [4]

3

The net descent gradient is obtained as follows:

γ̄net = γ̄gross − γ̄penalty (14)

Where the gradient penalty depends on the numberof operational engines and the number of installedengines. This information is presented in Table 1and can be found in [5].

Table 1: Gradient Penalties applied to Drift DownGross Flight Paths

γ̄penaltyTwo engines Four Engines

One engine out 1.1% 1.6%Two engines out - 0.5%

2.5. Route StudyGenerally speaking, both engine failure and depres-surization must always be expected to occur at themost critical points of the planned route. Moreover,since their descent profiles differ, the critical pointsmay differ between the two failure cases. These arepoints that separate two segments with differing es-cape strategies. An example is shown in Figure 4,where two critical points A and B are depicted.

Figure 4: Critical Points [4]

It is important to notice that regulations dontrequire to consider performance to cope with bothfailures simultaneously. The disadvantage of deal-ing with both failure cases separately, is that thenumber of critical points and specific escape routesincreases. This increases the workload of flightcrews and thereby the risk of errors.

For this reason, whenever possible, it is preferredto define identical critical points and escape routesfor all failure cases. This reduces the reaction timeand the risk of committing mistakes. In such a case,the route study should be based on the most penal-izing descent profile. Figure 5 illustrates both fail-ure descent profiles for a A319 over a mountainousarea.

Figure 5: Obstacle Clearance Profiles Engine Fail-ure and Depressurization[4]

3. ImplementationAn Emergency Profile Application (EPA) was de-veloped in order to develop the computational toolsand methodologies which seek to achieve the ob-jectives of the present work. The main purpose ofEPA is to verify if, at any point along a given route,an aircraft can execute an emergency descent safelyalong that same route.

EPA allows the computation and verificationof emergency flight profiles for depressurizationand engine failure scenarios. After computing theflight profile, it verifies if the profile can be flown,given current aircraft location and navigational con-straints. For its computations, EPA considers themost updated weather information, stored in theEFB’s database (DB). Flight performance calcula-tions are executed by Airbus’ Performance Engi-neer’s Program (PEP) software.

3.1. PEP integrationPEP is the computational motor of EPA. It is anAirbus software capable of performing every aircraftperformance calculations needed for EPA purposes.PEP computations should be as seamlessly inte-grated into EPA as possible. Ideally, EPA wouldonly be required to execute a batch file which wouldread the input .DAT file and return a .PRN file withthe computation results. This would allow PEP’scomputations to translate into a few lines of code inEPA, and to run in a very time efficient manner, re-quiring no interaction from the user. This solutionexists and can be distributed by Airbus.

The main PEP component used by EPA is theIn Flight Performance (IFP) module. IFP allowsto execute various computations regarding cruise,climb and descent scenarios. It is used by EPA inorder to compute depressurization profiles, branchby branch, and to get drift down gross profiles. Un-fortunately, as of the time at which this work waswritten, no solution was made available by Airbusthat allows PEP IFP computations to be executedwithout having to use the PEP standalone desktopapplication. This means that PEP’s input files, pro-duced by EPA, have to be manually imported intoPEP by EPA’s user, and manually run. This applies

4

to the computation of depressurization profiles andgross drift down profiles.

On the other hand, the net drift down profilescan already be computed automatically. PEP’sAPCMTP module, which includes flight planningsoftware and can simulate drift down emergen-cies, runs independently from PEP. APCMTP getscalled from EPA and doesn’t require any interac-tion from the user. It allows EPA to compute thenet drift down profile path which is to be verifiedagainst Minimum Altitude (MA) constraints.

3.2. Flowchart SymbologyThe flowcharts used in the following Subsections re-spect the symbology presented in Figure 6. Theyare based on Unified Modeling Language (UML)guidelines, and will be used to explain different al-gorithms and processes executed by EPA.

Figure 6: Flowchart Symbols Legend

3.3. EPA StructureEPA’s computational structure can be divided intofour main steps:

1. Configuration of Input Data

2. Computation of Flight Profile

3. Verification of MA constraints

4. Report Result to User

The order of these steps is as presented above,and they can be organized as shown in Figure 7.

Figure 7: EPA Basic Structure

3.4. Input DataIn order to compute an adequate flight profile, EPArequires data regarding the route, the aircraft and

atmospheric conditions. This data is required inorder to write the input files that are to be readby PEP, or to compute values for these input files.The way the data is gathered by EPA is illustratedin the flowchart of Figure 8.

Figure 8: EPA Data Retrieval Algorithm

EPA first requires the user to enter a valid FlightID, which is a unique flight identification code.Without entering the Flight ID, EPA doesn’t al-low the user to advance any further in the appli-cation. The Flight ID associates different types ofdata on the DB to the corresponding flight. Thisincludes the aircraft model and registration, routeand weather information, among other data. It istherefore an essential value.

As illustrated in Figure 8, the required data isdivided into two categories:

1. Configurable Data

2. Programmatically Defined Data

Configurable Data is suggested to the user basedon the flight context and the computation mode se-lected. The user can then customize this data. Forexample, if the user selects Depressurization Mode(DM), EPA will suggest that all engines are operat-ing. However, this can also not be the case, so thisoption can be customized.

Programmatically Defined Data includes all thevalues and information that are stored in the DB, orare established by regulation authorities, and thatare not to be changed by the user. This also in-cludes values computed by EPA and not open touser interaction. An example is the atmosphericconditions along the route, which are retrieved byEPA based on the current aircraft location, and arenot to be changed by the user.

3.5. Flight Profile ComputationThe computation of a descent flight profile is thecore component of the EPA. It calculates depressur-ization profiles branch by branch, following FCOM

5

and regulatory requirements. Atmospheric condi-tions are computed based on information availablein the DB, and appended to PEP’s input files. Co-ordinate tools have been developed to calculate dis-tances, heading angles, and other quantities relatedto global coordinates.

There are two computation modes configured intoEPA. The first one is Depressurization Mode (DM).As the name points out, this mode calculates a de-pressurization descent profile, tailored according tothe oxygen system installed in the aircraft. Thesecond mode is Engine Failure Mode (EFM), whichis used in case of drift down calculations followingengine failure.

3.6. Depressurization Mode (DM)DM allows the user to compute a depressurizationprofile along the default flight route, using the mostupdated meteorological conditions. Using the oxy-gen profile information (see Figure 2), EPA buildsa depressurization profile tailored to the FCOM’sconstraints. It follows the algorithm depicted inFigure 9.

Figure 9: EPA Computation Algorithm for DM

EPA starts by gathering all the required infor-mation provided by the DB and the user, whichtogether form the input data cluster.

To compute the full profile, EPA calculates andhandles each profile branch individually. Thismeans that each cruise and descent branch is com-puted individually, and verified against the FCOMoxygen profile requirements. This process is there-fore iterative. Once a given branch computationand verification is complete, EPA advances to thenext one. This proceeds until every branch has beensuccessfully computed and satisfies the FCOM’sconstraints.

One important problem to keep in mind is thatonly cruise branches are adjustable. This meansthat if a given descent branch turns out to end laterthan imposed by the FCOM profile, the precedentcruise branch has to be shortened. The opposite isalso true. If a given branch ends sooner than re-quired, the necessary adjustments are made to en-

sure that the aircraft spends the highest amount oftime, and thus covers the longest distance at thehighest altitude possible. Figure 10 illustrates theadjustable branches for a depressurization profilecorresponding to an aircraft cruising at an initial al-titude of 27300ft, with an installed oxygen system of12 minutes. The profile is composed of 6 branches,half of which are adjustable cruise branches. Theseare the branches that EPA can extend or shortenin order to achieve the FCOM time constraints de-picted in the time axis (x-Axis) of Figure 10.

Figure 10: Fixed and Adjustable Branches for 12minutes Oxygen Profile

The difference between the 1st and the 2nd PEPsimulation/computation for each branch is that thelatter includes the required weather information.Since weather information varies between every twoWaypoints (WPT) of the route, it has to be aver-aged per altitude, and per branch. It can only beincluded after the branch length is known, since thelength is a required value to average the weather in-formation of each branch.

As soon as the full profile has been successfullycomputed and satisfies all FCOM requirements, itis stored so it can be later verified against MA con-straints.

3.7. Engine Failure Mode (EFM)EFM allows the user to compute a drift down profilealong the default flight route, using the most up-dated meteorological conditions. EFM is executedaccording to the algorithm illustrated in Figure 11.

Figure 11: EPA Profile Computation Algorithm forEngine Failure Mode

6

Similarly to what happens in DM, EPA starts bygathering all the required information provided bythe DB and the user, which together form the inputdata cluster.

A drift down profile corresponds to a singlebranch or segment, contrasting with the multibranch depressurization profile. This means thata drift down profile can be calculated by PEP in asingle computation, and it simplifies the computa-tion process for EFM mode immensely.

The fact that EFM requires two computationsis because, similarly to what happened with otherbranches in DM, the latter computation includesweather information and yields the final results.Since weather information varies between every twoWaypoints (WPT) of the route, is has to be aver-aged per altitude, for the entire drift down profile.It also can only be included after the branch lengthis known, since the length is a required value toaverage the weather information of the profile.

EFM requires an additional step, not necessaryin DM. After the gross flight profile has been com-puted by PEP and the results have been stored,EPA penalizes the descent gradient in order to ob-tain the net flight path. It follows the guidelinespresented in [5], and uses a dedicated PEP modulefor it. This module obtains the net profile directly,given the initial descent conditions. This way, EPAnow has a gross and a net drift down profile avail-able as well.

3.8. Verification of Minimum Altitude ConstraintsAfter having computed the descent flight profile,EPA needs to verify if the computed profile can beflown safely, satisfying the obstacle clearance rulesimposed by regulations found in [5] and [6]. Formore detail on obstacle clearance rules please con-sult the main Thesis document.

Wether coming from DM or EFM, EPA now holdsthe full descent profile in a list of the type List(OfProfileCOORD), where ProfileCOORD.vb is an ob-ject class created for storing each individual pro-file point. A profile point corresponds to a step-or sub-computation executed by PEP. For exam-ple, each time PEP computes a drift down profile,it performs several sub-computations between theinitial and the final profile point. For each one ofthese sub-computations, PEP presents the elapsedtime, distance covered, fuel burned, and other in-formation. These values are each stored in a cor-responding ProfileCOORD.vb object, to build theentire flight profile.

For each profile point, EPA calculates two Mini-mum Altitude (MA) constraints:

• MSA - Minimum Safe Altitude• MORA - Minimum Off-route Altitude

The MSA is retrievable from the DB, and is re-leased together with the flight’s OFP. It clears allterrain and man-made structures by 2000ft [7]. Itis usually available for every cruise segment of theroute, and therefore establishes the minimum flightaltitude between two route WPTs. EPA extractsthe MSA of a given profile point by crossing the in-formation of the profile’s coordinates with the cor-responding route segment.

The MORA, on the other hand, defines a MAfor a certain rectangular area. Therefore, MA con-straints imposed by MORA are less restrictive thanthe ones that derive from MSA. It clears all terrainand man-made structures by 1000ft in areas at orbelow an elevation of 5000ft, and by 2000ft in areasabove an elevation of 5000ft [7]. MORA informa-tion is listed in Aeronautical Information Publica-tions (AIP), released according to the AeronauticalInformation Regulation And Control (AIRAC) cy-cle. EPA uses Honeywell’s AIP, which is subscribedby TAP, and loads the MORA information onto theDB. It then uses specifically developed methods todetermine the MORA at a given geographic coor-dinate.

After EPA has gathered available informationabout the MSA and the MORA at the coordinatesof each profile point, it just compares the altitude ofeach point against those MA conditions. The MSAis less restrictive than the MORA, and thereforethe altitude of each profile point is first comparedagainst the MSA. Moreover, the availability of anMSA value depends on wether it was filled in theOFP. If by any reason the MSA value is not avail-able or retrievable from the DB, EPA compares thealtitude of that point against the MORA.

The procedure is simple: If EPA finds that anyof the profile points has an altitude at or below theMSA/MORA condition at that location, the profilecannot be flown, as it doesn’t meet regulatory ap-proval. If, on the other hand, all point altitudes areabove their MSA/MORA requirement, the profileis safe.

3.9. Report Result to UserAfter establishing if the profile satisfies all MA con-straints and meets regulatory approval (see Subsec-tion 3.8), the result is communicated to the userwith a simple message. This message indicateswhether the profile satisfies all MA constraints, orif it violates MA requirements at one or more of itspoints.

4. ResultsSome sample flight routes were provided by TAPin order to test EPA’s functionalities. All flightsprovided are long haul flights over the ocean. Thismeans that the corresponding MA requirements atthe chosen locations are at their minimum. Sam-

7

ple computations were executed with EPA, and theresults obtained are thereby presented.

Some data tables of the DB are mentioned in thenext paragraphs, as well as auxiliary tools devel-oped for specific coordinate calculations. For moreinformation on the DB and these tools, please con-sult the main Thesis document.

Also note that some International Civil AviationOrganization (ICAO) codes will be used in orderto refer to specific global locations like airportsand aerial WPTs. These include LPPT, KMIA,EMAKO and TASNI.

4.1. Flight ScenarioA specific flight scenario was chosen in order todemonstrate EPA’s functionality. The same flightscenario will be used for both DM and EFM testing.

Flight Nr. 3051 is going to be utilized in order totest EPA’s DM and EFM. Flight 3051 departs fromLisbon Airport (LPPT) and its destination is Mi-ami’s International Airport (KMIA). The aircraftflying this route is the Airbus A330-223, with a 12minute oxygen system on board.

The flight information table (OFP RouteInfo)was filled until the WPT EMAKO, simulating thusa depressurization or and engine failure that hap-pened after the aircraft passed EMAKO and be-fore it reached the next WPT (TASNI). EMAKO istherefore the last WPT considered by EPA. Accord-ing to flight data records provided by TAP of thevery same flight, after WPT EMAKO the aircrafthad an initial weight of close to 170000kg, and wascruising at 38000ft, as predicted in the OFP. Thesewere therefore the initial conditions chosen for thiscomputation.

The WPT EMAKO has the following coordi-nates: (+31,397◦/-68,238◦) (Latitude/Longitude).TASNI has the coordinates (+30,9◦/-69,225◦). Thetwo WPTs are 58,84NM away from each other, andthe bearing when the aircraft flies over EMAKO is58,84◦.

A simple coordinate utility was developed to per-form calculations with coordinates. Using this util-ity, an arbitrary point after EMAKO was chosen.The initial coordinates considered for this calcu-lation are (+31,399◦/-68,575◦). These coordinatescorrespond to a point situated 20 NM after the air-craft passes EMAKO. It is situated above the At-lantic Ocean. This is assumed to be the point wherethe aircraft must start descending along its depres-surization or drift down profile.

4.2. Depressurization Mode ComputationEPA took a total of 22 PEP calculations to computethe full depressurization profile. The results of thefinal computation were extracted to a .CSV file andare illustrated in Figure 12.

The aircraft takes a total of around 42 minutes

and covers 286 NM before it reaches FL100.During its computations, EPA verified each

branch in order to adjust the cruise branches andmaximize the time and distance at each altitude. IfEPA didn’t have to verify and adjust the branches,it would only require 12 computations. This wouldcorrespond to two computations per branch (thefirst to compute the branch without atmosphericconditions, and the second already with this infor-mation).

Figure 12: Depressurization Profile computed byEPA

Please compare the results profile of Figure 12 tothe FCOM profile depicted in Figure 10.

One can easily see that the time limits at 4minutes and 11 minutes have been successfully re-spected. As soon as the aircraft reaches the corre-sponding profile points, it starts descending alongthe next branch. EPA successfully guarantees thatthe aircraft stays on each required altitude for themaximum amount of time and the for the longestdistance possible.

The only observed FCOM requisite that EPAcouldn’t keep was the last 30min time limit above10000ft, starting from the point where the air-craft reaches 14000ft. The duration of the last twobranches combined is 30, 02min = 30min 1 seg.This error is however negligible:

4 = 0,0230 ≈ 0, 067%

After finishing its computations, EPA displayed amessage informing that the profile satisfies all reg-ulatory requirements and doesn’t violate any MAconstraints. The MORA and the MSA values arekept at a minimum throughout the flight path, sincethe aircraft is supposedly flying over the sea. Thisresult was therefore expected.

4.3. Engine Failure ModeThe same flight scenario was used by EPA in orderto compute the drift down profile. The computa-tion of gross profile only required two PEP com-

8

putations, and only a small fraction of time whencompared to DM.

EPA then obtained the net profile path using theAPCMTP module. Both the gross and the net pro-files can be analyzed in Figure 14. Please note thatthe time axis corresponds to the gross profile path.

Figure 13: Drift Down Profile computed by EPA

The gross profile is computed until it reaches thelevel-off altitude, at 27511ft. The aircraft takes1h13min to reach this point, and covers a total dis-tance of 400,3NM. If the flight were to continue atthis point, the aircraft would start cruising again,and increase its altitude as it became lighter.

In contrast to this, the net profile is computedto and beyond its level-off point. Since the netlevel-off point is reached in a shorter distance thanthe gross one, the aircraft already started climbingwhen it reaches the gross level-off coordinates. Thenet level-off altitude is situated 266 NM ahead ofthe TOD. The net profile reaches the gross level-offaltitude, 27511ft, only 84 NM beyond passing theTOD. This corresponds to approximately a fifth ofthe distance required by the gross profile to achievethe gross level-off altitude.

The difference between gross and net level-off al-titudes is 3495 ft, which corresponds to 3495·10027511 % =12, 7% of the gross level-off altitude.

For the EFM, despite not having temporal con-straints to compare the results against, like in DM,the results are satisfactory. EPA successfully com-putes both the gross and the net profiles, and com-pares the latter against the corresponding MA con-straints. In an operational context, the gross profileis not required, since the net profile is the one to becompared against MAs. However, it was easily ob-tainable using the tools already developed for DM,and allows the user to easily compare both results.

4.4. Combined resultsFigure ?? shows the three resulting profiles over-lapped, in order to visually be able to comparethem. One can see that the depressurization profileis by far the most penalizing in terms of altitude.It maintains the initial 38000ft for just 1 minute

and 14 seconds, and afterwards starts descendingat a RD of over 11000 ftmin . After a ground dis-tance of 31 NM the aircraft is already cruising at18000ft, well below the 33254ft at which the aircraftfinds itself along the (net) drift down profile afteran equivalent ground distance from the TOD.

Figure 14: Combined Results - Depressurizationand Engine Failure

5. ConclusionsThis section presents an overall balance of the work,and future work.

5.1. BalanceThe overall balance of this work is positive. Themain objectives were successfully accomplished. Anapplication was developed, which contains the com-putational methodology to compute flight profilesfor emergency descent scenarios, for both depres-surization and engine failure scenarios. The appli-cation considers the most updated weather data,and also verifies the computed flight profiles againstthe appropriate altitude constraints, determiningwether the profile is valid or not.

However, to further expand and integrate thedeveloped methodology in TAP’s EFB solution,some developments are required. The fact thatPEP’s computations cannot currently be dissoci-ated from it’s desktop application, means that, fornow, computations have to be executed manually.This is a very time consuming process, as EPArequires a high number of computations to com-plete a depressurization profile. The user has toimport each file manually and execute the corre-sponding simulation. This issue should be solvedif EPA is to be integrated in more complex routeanalysis applications. It can only be solved if Air-bus releases a batch application to TAP, which canbe called to run PEP simulations independentlyfrom PEP’s desktop application, similarly to theAPCMTP module used for net drift down profilecalculation.

All-in-all, one can say that the work presentedachieved its proposed goals within the given con-

9

straints. EPA shouldn’t be viewed as a softwarewith low operational application, but rather asa computational foundation with high-value tools,which TAP’s EFB team can build upon.

5.2. Future WorkAs mentioned at the end of Subsection 5.1, EPAis a computational foundation upon which TAP’sEFB team can build upon. Its tools can be appliedto more robust and complex verification of escaperoutes, making this process run in a seamless way,throughout the flight.

Currently, EPA only considers the main route inits calculations, meaning that it is along that routethat it calculates the emergency descent profiles.It is also along this route that EPA verifies if theprofile is valid and safe.

In the future, parameters like the aircraft alti-tude and weight could be automatically sent fromthe aircraft to the EFB through a direct data link.Together with an independent execution of PEP’scomputation module (see 3.1), this would unlockthe full potential of EPA. The software can then beexpanded to analyze alternative emergency routes.The application could then evolve into scanning en-tire flight routes, identify critical points automat-ically and suggest appropriate escape routes, eachtime a new weather update becomes available.

Flight crews can then be sure that the proceduresthey follow in case of failure rely on the most re-cent meteorological conditions. This would trans-late into an improved operational safety in regardto cruise emergency procedures.

AcknowledgementsThe author would like to thank Prof. AntónioAguiar and Eng. Carlos Figueiredo for grantinghim the opportunity to participate in such an en-riching experience. Their guidance and availabilityat every stage were crucial for the success of thedeveloped work.

References[1] EASA. AMC 20-25: Airworthiness and Opera-

tional Consideration for Electronic Flight Bags(EFBs), February 2014.

[2] Dr. Addison Schonland. 2014 EFB Report, De-cember 2014.

[3] Dr. Chuan-Tau Edward Lan Dr. Jan Roskam.Airplane Aerodynamics and Performance.DARcorporation, 1997.

[4] Airbus. Getting to grips with aircraft perfor-mance, January 2002.

[5] European Comission. Comission Regulation(EU) No 965/2012, October 2012.

[6] EASA. Annex VIII the draft Commission Reg-ulation on ‘Air Operations - OPS’, April 2012.

[7] Jeppesen. Introduction to Jeppesen NavigationCharts, 2012.

10