Fuel Conservation Strategies

Through Flight Operation Optimization at PGA

Simao Trincheirassimaotrincheiras@tecnico.ulisboa.pt

Instituto Superior Tecnico, Lisboa, Portugal

November 2016

Abstract

Fuel conservation is an important topic for airlines, not only as a way of reducing operating andmaintenance costs, but also emissions. Even with the current expansion of the airline industry and lowfuel prices, it is estimated that a well implemented savings plan can help achieve 2% to 5% reducedfuel consumptions, which translates into higher profit margins. For this work, a Portuguese regionalairline – PGA - was used as a case study, providing data on its new fleet of nine Embraer 190. A setof five operating measures suggested by the manufacturer (like Single Engine Taxi and Idle Descent)were studied and analyzed using data recordings from sixty flights, over three frequent and differentlength city pairs, with two flight legs each (outbound and inbound), in order for the sample to berepresentative. The goal of this work was to study and implement this set of operational measuresin a software tool using MATLAB that was able to analyze the flights and estimate the potentialsavings if the measures had been fulfilled. The results of this work are very encouraging, and eventhough its full saving potential is unlikely to be achieved in a real-world scenario due to the natureand unpredictability of air traffic, it allows for PGA to understand and prioritize certain measures thatprovide greater return with a lower effort. The financial impact of these savings is also estimated, andcan reach up to around 875ke (3.62% of the fuel consumption) for PGA specific case.Keywords: Fuel Conservation, Regional Airline, Operational Measures, Flight Data Analysis.

1. Introduction

The 2008 economical crisis shook the world, and forthe aviation industry, this reflected in a lower de-mand and a rise in fuel prices, which resulted in anexcessive offer and the growth of low-cost airlines.To remain competitive, airlines were forced to lowertheir fares, however, with this decrease in revenue,a corresponding reduction in expenses was also due.In this sense, it became of the utmost importanceto develop sustainable policies for the airline oper-ation, not only to provide a more efficient and op-timized activity, but to also to make it more robustand less dependent on market fluctuations.

With this objective in mind, PGA - PortugaliaAirlines, a portuguese regional airline - started sev-eral years ago its first studies developing tools toidentify cost reduction strategies. This work illus-trates some of these efforts, applied to a recent fleet,and focuses on fuel conservation through the imple-mentation of a set of operational measures. Thefleet studied in this work consists of 9 Embraer E-190 aircraft that were bought and started flying onthe 23rd May 2016.

1.1. Motivation

For a small airline like PGA, operating a fleet ofregional jets on short and medium-haul routes, the

operational conditions present different challengeswhen compared to larger airlines. This is due toshorter flight distances while still focusing on ma-jor urban airports, but also due to the inherentlyhigher flight cycle to hour ratio, meaning highermaintenance costs, that can even reach about 20%of total direct operating costs.

When flying a particular airplane, airlines de-fine a Cash Operating Cost (COC) relating to thatoperation, which for a determined route involvesthree main components: Fixed Costs, Time Re-lated Costs and Fuel Related Costs. Since the year2000, jet fuel has faced tremendous variations inprice (Figure 1). On the one hand, political andeconomic factors, as well as wars, all influence thecrude price, which can be highly unpredictable andvolatile. On the other hand, Fuel Related Costs canreach up to 40% of the total COC of a flight, andare usually where the most flagrant saving oppor-tunities are identified.

As a consequence, the implementation of fuel con-servation optimization becomes one of the mostimportant tools when it comes to cost reductionstrategies. Even with jet fuel prices considerablylower compared to 2008, and the sector giving thefirst real signs of recovery since the beginning of the

1

Figure 1: Brent prices in Europe from 1988 to 2016[1].

crisis, it is still of the utmost importance for air-lines to optimize their operation. Marginal profitmargins imply not only small profits, but more im-portantly, the smallest variation in costs (e.g., jetfuel prices) can mean non-profitable operation andlosses of millions in this industry.

Considering these facts, fuel conservation strate-gies are extremely important in the context of PGA.Even though there are complete fuel efficiency pack-age tools available on the market (to be discussedin Section 2.1), they invariably come with a heftyprice tag. This work is a preliminary way for PGAto evaluate the potential of implementing fuel con-servation strategies on their new fleet, focusing onwhat were considered the most important and read-ily available operational procedures.

1.1.1 1% Fuel Conservation Estimations

One interesting value to calculate is the amount ofsavings, both in kg of jet fuel and in euros, thatcould be obtained for the company if it reducedits fuel consumption by 1%. To do this, a typicaloperational profile of the company is needed, andEmbraer provides an estimate for a 600 NM trip,correponding to a takeoff weight of 42 925kg, tripfuel of 3147kg and a total trip time of 90min.

Reducing trip fuel by 1% would the result in 31kgo fuel saved per flight [2], which multiplied by thenumber of flights a fleet of 20 of these aircraft per-forms a year yields the surprising value of 1180 met-ric tons of fuel saved per year [2]. However, PGAstill has a slightly smaller fleet of only 9 aircraft, andtherefore these estimates must be adjusted. Still,this returns the impressive value of 531 metric tonsof potential fuel savings a year.

Considering the current jet fuel price in Europeof 446.6 $/metric ton provided by IATA [3], thismeans a total yearly saving estimate of approxi-mately 237k$, equivalent to 212ke. Note that ap-parently small amounts of fuel economy translateto very significant financial differences, making thisa very important topic when it comes to cost reduc-tion.

1.2. ObjectivesThe main objective of this thesis is to evaluate theimpact of fuel conservation strategies on a regionalairline performance. For this work, the focus wasset on operation optimization not because it has thehighest saving potential, but because it is the most

simple and direct way for an airline to start im-plementing fuel conservation. Therefore, the wholeprocess was divided into three main objectives:

• Data collection from the aircraft recorder sys-tems;

• Data selection, processing, and estimation ofsavings;

• Statistical data analysis, with the evaluation ofeach saving measure.

1.3. StructureThis work is divided into five different chapters.The second chapter discusses the state of the artin this industry, an overview of the current avail-able solutions and the tools used during the exe-cution of this work. The third chapter enumeratesthe methodology used for such analysis, the savingsmeasures and the associated fuel economy metrics.The fourth chapter details the results of the anal-ysis on the provided sample, discussing the vari-ous routes and measures. Finally, the fifth and lastchapter presents the conclusions of this work, leav-ing suggestions for future work in the area.

2. State of the Art & Analysis Tools2.1. Market offerNowadays, due to the growing fuel economy and en-vironmental sustainability awareness, the numberof companies offering complete software solutionsto elevate airline efficiency has faced a great expan-sion. This includes Crew and Fleet optimization, aswell as Maintenance and, most important for thiswork, Fuel Efficiency. Two of the most importantsuppliers of these services are Honeywell and Ope-nAirlines.

2.1.1 Honeywell - Aviaso

Aviaso was founded in 2003, in Switzerland, andbought by Honeywell in September 2015 [4]. Aviasomethod to take on the Fuel Efficiency theme is di-vided into four main steps. First, it is important tocollect the relevant data from the many airline de-partments and IT infrastructures, from flight plan-ning, to operations, maintenance, etc. Then, thedata must be checked for quality assurance. Afterhaving all the relevant data properly checked, fuelsavings potential is calculated and the current fuelconservation program progress is monitored. Fi-nally, it is imperative to convey this data to each ofthe responsible entities, customized for their neces-sities [5].

2.1.2 OpenAirlines - SkyBreathe

Founded in 2006, OpenAirlines is a company fo-cusing on three main aspects of airline efficiency- Fuel, Crew and Fleet efficiency. By optimizingthe resources across the various airline departments,OpenAirlines promises fuel savings in the region of2% to 5% [6], corroborated by IATA estimates of3% to 5% for a systematic airline optimization.

2

Much like the competition, Skybreathe alsopresents a few steps in its fuel conservation opti-mization analysis, five in this case. Both the dataintegration and quality control are similar, in termsof concept, with the difference that the analysisis separated into two different steps here - savingscomputation and data analysis. Finally, the datacommunication step to the different entities is givena great deal of importance, with fully customizablereports and dashboards for all types of users [6].

SkyBreathe dashboard allows for a comprehen-sive understanding of the airline efficiency. It al-lows for a route analysis, flight phase analysis, bestand worst practices (whether the saving potential isbeing fulfilled), individual aircraft savings and evenpilot savings [6].

2.1.3 Embraer fuel conservation strategies

Like most aircraft manufacturers, Embraer pro-vides manuals and technical sheets that focus onresource optimization either in terms of planning,operation or maintenance ([2], [4]). As stated ear-lier, this work focused on operation optimizationand, therefore, Embraer recommended savings pro-cedures constituted the foundation of this study.These procedures are presented and analyzed fur-ther in depth in Section 3.

2.2. Data capture and extraction2.2.1 Quick Access Recorder

For fuel efficiency measures to be analyzed and im-plemented, it is necessary to have flight data regard-ing the flight planning and operations. In order toobtain the data from the various sensors and param-eters computed by the airplane, the Quick AccessRecorder (QAR) was used. It is used by airlinesto improve flight safety and operational efficiency.Since usually a QAR is not a mandatory system oncommercial flights, it is not designed to survive acrash, like the Flight Data Recorder (FDR). Thisalso means that it is a more flexible system for air-liners, as it can process data at much higher ratesthan the FDR, and frequently for longer periods oftime.

The QAR unit used on PGA Embraer 190 fleet ismanufactured by SAGEM, and its part number isED35E109-04-00. This QAR dataframe is fully con-figurable - which parameters to be recorded and thesample rates are all programmable, as is the startand stop logic for the recording. For PGA aircraft,the QAR is set to record whenever the plane is ener-gized, in order to keep a more comprehensive recordfor safety and fuel consumption analysis. The elec-trical energy for the plane systems can come fromthe engines, the APU or an external GPU.

2.2.2 Analysis Ground Station

To analyze, extract and produce reports with theQAR raw data, a special software tool is neces-

sary, and for this work, the Analysis Ground Station(AGS) software by the QAR manufacturer SAGEMwas used. Due to the complexity of the software,only a small part of its potential was used for thisstudy. Essentially, it was used to filter the flights byroutes and aircraft, then to produce reports basedon the relevant variables, and finally to export thisdata, later to be read and processed by MATLAB.

3. MethodologyIn order to evaluate the fuel saving potential forthis airline, a group of significant measures thatwere simultaneously easy to evaluate and to put topractice were chosen and characterized. To do this,many articles by several aircraft producers and in-dustry associates were analyzed and compared ([2],[4], [7]).

Then, to analyze the effect of each of the mea-sures on the airline operations, a set of routes (city-pairs) was defined. This allowed for a rough statis-tical analysis of various flight scenarios and corre-sponding fuel savings, evaluating the financial via-bility of the previously defined measures.

3.1. Fuel Savings MeasuresFor this work, five main fuel saving measures werechosen for representing relatively simple day-to-dayoperational procedures that can help reduce fuelcosts. These are: avoiding the use of the Auxil-iary Power Unit and Thrust Reverser systems, andshutting them down as soon as possible; Single En-gine Taxi In and Out; and using the Idle Descent,consisting of setting the engines to idle thrust dur-ing descent.

It should be stressed that all the savings esti-mated in this work are calculated for optimal con-ditions and thus, their full saving potential can bevery hard to achieve under real-life operation. Beit because of ATC, flight delays, meteorological as-pects, mechanical condition of the aircraft or manyother reasons, the important result is to recognizehow much fuel could be saved by simply accomplish-ing these measures, if all the unpredictable aspectsof flying went according to plan.

3.1.1 Auxiliary Power Unit

The APU, or Auxiliary Power Unit, is a devicedesigned to generate pneumatic and electrical ACpower for the various systems of the aircraft. Itconsists of a gas turbine engine, located in the air-plane tailcone, that runs on regular jet fuel and canprovide bleed air for starting the engines and forthe air conditioning packs, coupled with an electri-cal AC generator that supplies 115V 40kVA to theelectrical system.

Considering jet fuel prices, using the APU onground is much more expensive than running aGPU as an electrical power source. Similarly, ex-ternal air carts can be used to power the air condi-tioning packs, saving valuable fuel on ground.

3

Companies are invited to evaluate if continuoususe of APU at the gate instead of GPU is thebest option, considering that main APU compo-nents fail by cycle, and therefore, for really shortturnarounds, the marginal fuel saving might notjustify the extra APU maintenance costs; GPUpower, like external air carts, is most of the timeleased from ground handling companies and can beexcessively priced in some situations, or charged bythe hour, turning it into a pricier option.

For when APU continuous use on the ground isactually the most cost effective measure, it is recom-mended its usage time be minimized. This meansonly turning the APU on after landing and turningit off immediately after starting up the engines.

Conditions Unfortunately, evaluating the sav-ings potential on turnarounds would require a con-stantly up-to-date database of fuel prices, GPU andexternal air leasing prices in every airport and allthe conditions associated with the lease.

Therefore, the only condition set for the savingsanalysis of the APU, is that it should be turnedoff whenever there is at least one engine running.This means turning off the APU as soon as the firstengine is started and turning it on only immediatelybefore shutting the last engine down. A buffer timeof 60s is granted in both of these situations, for thecrew start-up/shutdown checklist time.

Savings estimates Whenever the above condi-tion is not met, and the APU is running simulta-neously with the engines, a fuel saving can be es-timated. This is done by converting the APU FuelFlow from PPH to kg/s and integrating it over thattime interval. The result is the amount of kg of fuelthat could have been saved, had the APU been usedefficiently.

3.1.2 Single Engine Taxi

In order to improve fuel savings, a single enginetaxi can be used, delaying the start of the secondengine prior to takeoff and shutting one engine justafter landing when taxiing in. A set of conditionsinfluence this maneuver, including: ramp weight;ramp gradient; engine warm up and cooling downperiod; contaminated taxiways; and the higher pilotworkload associated.

These aircraft are relatively light and thus requireless power to taxi, however, the engine warm up andcooling down periods must be met to allow for en-gine thermal stabilization - the second engine mustbe kept running for at least 2 minutes at idle beforeselecting high thrust settings or shutdown. It is alsoadvised that, when taxiing out, the second engineshould be started with the airplane static to avoidpilot heads down condition during taxi.

During single engine taxi, the fuel flow is approx-imately 5 kg/min or 300 kg/h, due to an increment

in thrust compared to normal taxi thrust per engine[2].

Conditions For SETO (Single Engine Taxi Out),the only limitation is the second engine warm uptime, and therefore, the only condition is to startthe second engine 2 minutes before takeoff [2]. Sim-ilarly, for SETI (Single Engine Taxi In), the onlycondition is to shutdown the second engine 2 min-utes after landing, in other words, 2 minutes aftertaxi out starts.

Savings estimates Whenever the above condi-tions are not met (i.e. the second engine is runningmore than 2 minutes before takeoff or after landing)a fuel saving can be estimated. This is done by cal-culating the difference between the actual aircraftfuel flows (FF1 + FF2) and the average SET fuelflow (300 kg/h) and integrating it.

3.1.3 Thrust Reversal

Reverse thrust can be used to stop the aircraft ina shorter runway length. This is essentially accom-plished by redirecting the engine exhaust forward,rather than backward, providing deceleration. Eventhough this system can help reduce wear on thebrakes, it also usually means higher fuel consump-tions and engine wear, depending on the amount ofreverse thrust selected by the pilot. With full re-verse thrust, fuel flow can reach 3200kg/h, a valuesimilar to takeoff thrust configuration.

The higher the aircraft speed, the more efficientreverse thrust is, since it uses more air mass andtherefore produces bigger brake forces. However,applying high reverse thrust at lower speeds caninduce an inlet vortex, exhaust gas and FOD in-gestion, especially in contaminated runways, andshould therefore be avoided. On snow or ice coveredrunways, it can even lead to low forward visibilitydue to a ”whiteout” (snow being propelled forwardby the engines).

Conditions One should note that, consideringfuel efficiency, the thrust reversers should only beengaged when absolutely necessary, due to run-way length or other limiting factors like inoperativebrakes. According to the GP1999 by Embraer [2],FOD ingestion can occur below 80 KIAS, so it isadvised to turn off maximum reverse thrust belowthat speed. As the airspeed reduces, so does thethrust reversers efficiency, and so it is stated thatat 60 KIAS reverse should be canceled altogetherin such way that it will be completely stowed whenreaching normal taxi speed.

Savings estimates ”Keeping full reverse thrustactuated until airplane stops completely will in-crease approach and landing fuel by 10 kg [2].”Therefore, it is expected that the savings calculatedwith this method fall below that number. When-ever reverse thrust usage does not comply with the

4

above stated conditions, a saving is calculated bysubtracting to the actual engines fuel flow, the aver-age idle fuel flow, and integrating it over that periodof time.3.1.4 Idle Descent

The descent phase usually represents a lower fuelconsumption than the climb or cruise phases, witha trip time percentage that should be around 10%for short and medium range flights, and fuel flowsmany times smaller. From a fuel consumption pointof view, the descent should be done as fast as pos-sible, using high speeds. However, this can causepassenger discomfort, due to the cabin pressure rateof change, and for really high speeds, can cause thetrip fuel to actually increase due to the extendedcruise period and high fuel flows during descent.

To maximize fuel savings, idle thrust coupledwith a constant flight path angle is recommendedby Embraer and enforced by PGA in their SOP [7],due to the lower fuel flows, effectively decreasingdescent fuel. This procedure is called Idle Descent.

Currently, many worldwide operational safetyregulations require a speed limit of 250 KIAS be-low 10000 ft. The manufacturer stresses that theEmbraer 190 was designed and flight tested for birdimpacts up to speeds of 300 KIAS, and can there-fore fly those speeds below 10000 ft safely. It isestimated that the elimination of these speed limitsboth in the descent and climb phase would resultin savings between 14 and 25 kg of fuel per flight,while contributing to reducing the flight time (andtherefore time costs).

Conditions For the savings analysis, the enginesthrust setting was taken into consideration. Assuch, whenever the Thrust Lever was not set to idlein any of the engines during the descent phase, a fuelsaving opportunity was identified.

Savings estimates Whenever at least one enginehas its thrust setting different from idle during de-scent, the potential fuel saving is calculated takinginto account an average idle fuel flow per engine.However, idle fuel flow depends linearly with the op-eration altitude [8], and therefore must be estimatedevery second. This estimation was made based ona flight in our database with a nearly perfect idledescent, that resulted in the following linear fit:

FF (kg/h) = −0.016 × h(m) + 377.013 (1)

This means calculating the extra fuel burned byadding both engines fuel flow, subtracting the av-erage estimated idle fuel flow for that altitude, andintegrating it over the period in which the condi-tions were not met.

3.2. Route choiceHaving the savings measures correctly imple-mented, it became imperative to define a samplethat could represent this regional airline operation.

Figure 2: Idle Fuel Flow variation with Altitude

In that sense, the flight data was analyzed and thevarious flown city-pairs were listed both by distanceand number of flights, as of 30th June 2016. In or-der to obtain more homogeneous results, three city-pairs were chosen, one short, one medium and onelong-haul, taking into account the number of flightsin each category. Additionally, each city-pair rep-resented and outbound and inbound flights, whichmeant the sample was comprised of 6 routes:

• LIS-OPO & OPO-LIS (flights between Lisbonand Oporto);

• LIS-BCN & BCN-LIS (flights between Lisbonand Barcelona);

• LIS-NCE & NCE-LIS (flights between Lisbonand Nice);

Then, for each of these routes, 10 flights wereselected and run through the simulation, in order tohave a significant sample to analyze statistically.

4. Results4.1. Route Analysis4.1.1 LIS-OPO & OPO-LIS

Being the shortest route analyzed, it is to be ex-pected that a flight between Lisbon and Oporto hasa fairly low efficiency, as less time is spent cruisingand more time in non-optimal phases of flight liketaxiing, taking off and climbing. However, this doesnot necessarily mean higher saving potentials, buta careful study of an average flight can help identifyand justify the best savings to be implemented.

Figure 3: LIS-OPO phase time and fuel consump-tion

In this short route, taxi times are very prominent(Figure 3). We can see Taxi Out is actually thelongest flight phase for this route, reaching above

5

25% of the total trip time and Taxi In times justunder 10%. Still, the phase with the most fuel con-sumed is clearly Climb, with about 40% of the fuelconsumed, but only under 15% of the total time,which is a very similar duration to Cruise.

For the returning flights, taxi times are very rel-evant, but here the longest flight phase is actuallyDescent, which is a consequence of the very shortdistance of the flight. Also like before, the phasewith the most fuel consumption is Climb, which isexplained by its rather high relative duration com-bined with very high fuel flows. It is also relevantnoticing Taxi Out times with a much shorter dura-tion, nearly halved from 25.5% to 12.5%, as well asTaxi In times with a fairly higher duration, reaching11.9%, compared to the previous 8.7%. Both thesedifferences come from the fact that Lisbon Inter-national Airport is much more congested and withmore complex and longer taxiways, therefore taxitimes are usually much higher than in Oporto.

From Table 1 it is possible to see that SETO is thehighest potential saving measure for the LIS-OPOroute, with nearly 2% total trip fuel that can besaved from correct implementation of this measure.In fact, the difference is pretty dramatic, with APUand SETO savings practically halved for the reverseroute (OPO-LIS), while Idle Descent reaches morethan 3.5%.

Idle Descent savings potential also relates to thehigher traffic found in Lisbon International Airport,because many times pilots are encouraged by ATCto accelerate or to keep a certain speed dependingon other departing and/or arriving aircraft. It isalways preferable to adapt the descent phase thanto be forced to enter a holding pattern, thus IdleDescent savings potential more than doubles whenlanding in Lisbon compared to Oporto.

Route Unit Thrust Rev. APU SETO SETI Idle Desc. Total

LIS-OPO

kg 0,49 10,56 26,57 1,92 19,55 59,09

Total % 0,03 0,74 1,87 0,14 1,38 4,16

Relative % 0,83 17,87 44,96 3,24 33,09 100,00

OPO-LIS

kg 0,54 5,66 12,76 2,33 54,55 75,85

Total % 0,04 0,38 0,85 0,16 3,65 5,08

Relative % 0,72 7,46 16,83 3,08 71,92 100,00

Table 1: LIS-OPO and OPO-LIS average savingsestimates

As flight distances and durations are fairly simi-lar for the inbound and outbound flights of the samecity-pair, comparing savings estimates in % betweenthem is viable, however, it is always preferable toanalyze the results in absolute values. This also al-lows for comparison across different flight distances,presented further ahead. Here, for example, it ispossible to see that the inbound flight has a savingpotential more than 15kg higher, despite 26.57kgof fuel wasted in the Taxi Out phase for the out-bound one, largely outweighed by the fact that IdleDescent has a saving potential 35kg higher when

landing in Lisbon.Finally, regarding the comparison between each

measures relative percentage in terms of the totalsaving SETO is, as expected, the main focus for fuelconservation in the LIS-OPO route (45%), followedby Idle Descent (33%). For the inbound flight, thepreponderance goes to Idle Descent measures, at72%, with SETO down to a more reasonable 17%.For both flights, Thrust Reverser operation andSETI have minimal relative percentages, the firstbecause it is not used very often, and the secondbecause the procedure is nearly optimized.

4.1.2 LIS-BCN & BCN-LIS

On a medium-haul route like the connection be-tween Lisbon and Barcelona, the efficiency is ex-pected to be greater than that seen on the previousshorter routes. Due to the longer flight distances,a bigger percentage of the time is spent cruising,with an expected percentage-wise reduction of taxitimes.

Figure 4: LIS-BCN phase time and fuel consump-tion

In fact it is possible to identify in Figure 4 thethree main flight phases - Climb, Cruise and De-scent - as the longest, which is to be expected formedium-haul flights. This is also valid for fuel con-sumption, as these three phases represent around90% of the total fuel consumed for the fight. De-spite this, Taxi Out times are still significant, withover 10% of the time spent on this phase, whichonly proves that the Lisbon Airport has a low effi-ciency when it comes to taxiing, much due to thehigh amount of traffic exceeding the airport ca-pacity. The main difference for the inbound flightin this route is the slight reduction in Taxi Outand Descent relative times, with the correspondingincrease in the percentage of time spent cruising,which usually means a more efficient flight. Eventhough Barcelona airport is much bigger, complexand carries more traffic, Taxi Out times still averagearound 50 seconds less than in Lisbon, which onlyproves the latter’s lack of capacity for the currenttraffic it deals with.

For these routes, both departing from Lisbon andBarcelona (Table 2), Idle Descent is clearly the fuel

6

conservation strategy with the highest saving po-tential. It is to be expected that savings measuresfor the three main phases of flight become more sig-nificant as the flight distance increases, consideringthe amount of time spent on these phases is higher,percentage-wise. The fact that Idle Descent rep-resents savings of 5.35% (169.97 kg) per outboundflight and only 2.94% (101.54 kg) for inbound oneshas to do with a multitude of factors, including theapproach procedure for each airport, the bigger ex-perience with the maneuver in Lisbon airport, butmore importantly, because of ATC imposed limi-tations. Barcelona airspace is very crowded, andbecause of that some of its regulations are stricter.Pilots are often encouraged to accelerate the air-craft engines in order to keep a certain speed, de-scend faster or above all, avoid holding. Of course,this means longer periods of time not complyingwith the idle thrust policy, therefore a fuel savingcan be calculated by the program for a longer pe-riod of time, resulting in higher fuel conservationestimates.

Route Unit Thrust Rev. APU SETO SETI Idle Desc. Total

LIS-BCN

kg 0,23 8,84 25,64 4,14 169,97 208,83

Total % 0,01 0,26 0,75 0,12 5,00 6,14

Relative % 0,11 4,23 12,28 1,98 81,39 100,00

BCN-LIS

kg 0,51 9,15 14,64 1,76 101,54 127,60

Total % 0,01 0,23 0,37 0,04 2,58 3,25

Relative % 0,40 7,17 11,47 1,38 79,57 100,00

Table 2: LIS-BCN and BCN-LIS average savingsestimates

It is hard to ”blame” any specific measure for thebig difference of almost double the total savings per-centage between the outbound and inbound flightlegs. In fact, almost every fuel conservation strat-egy potential is halved for flights arriving in Lisbon,which leads to think that the whole LIS-BCN routehas a lower efficiency when compared to BCN-LIS.Despite the difference in absolute savings, the sav-ings relative distribution between both flight legsis relatively similar with around 80% coming fromthe Idle Descent procedure and 12% from the SETOmaneuver.4.1.3 LIS-NCE & NCE-LIS

A flight connecting Lisbon and Nice is already con-sidered a long-haul flight for many regional airlines,even though the Embraer E190 has a range of nearlytriple that distance. This makes for a potentiallymore efficient flight, since the aircraft spends moretime in the conditions it was designed to operate- cruising - instead of maneuvering, accelerating ortaxiing on the ground.

For this route, Cruise really stands out as thelongest and most fuel consuming phase of the flightfor both legs (Figure 5), which is to be expected.Fuel consumption in the three main phases reachesabove 90% of total values, and therefore, the longerthe flight, the more important it is to optimize op-erations and planning for these flight phases. For

the inbound flight, like before, Taxi Out times arereduced by almost 200s, and Taxi In times are closeto 40s longer on average, strengthening the hypoth-esis that Lisbon airport is over encumbered, andthus ground operations suffer from it.

Figure 5: NCE-LIS phase time and fuel consump-tion

Concerning the actual savings estimations foreach measure, it is possible to identify Idle Descentas the highest potential procedure for the LIS-NCEflight (Table 3), with close to 3% of the total tripfuel that could be saved by correct implementationof this measure alone. Thrust reverser operationand SETI are fairly optimized, while APU usageand SETI present potentials of around 0.25% and0.5% respectively. As for the flight leg landing inLisbon, the distribution of the savings is rather sim-ilar, however savings values are significantly lower.The exception seems to be the APU usage, whichpresents a potential slightly above 0.2%, which isgreater than that of the outbound leg, percentage-wise.

Route Unit Thrust Rev. APU SETO SETI Idle Desc. Total

LIS-NCE

kg 0,35 8,80 25,24 3,33 157,39 195,11

Total % 0,01 0,16 0,47 0,06 2,93 3,63

Relative % 0,18 4,51 12,94 1,71 80,67 100,00

NCE-LIS

kg 0,32 12,12 14,32 2,31 56,30 85,38

Total % 0,01 0,24 0,28 0,05 1,10 1,67

Relative % 0,37 14,20 16,78 2,71 65,95 100,00

Table 3: LIS-NCE and NCE-LIS average savingsestimates

In fact, Idle Descent absolute savings drop from157.39kg when landing in Nice to 56.30kg for alanding in Lisbon, almost one third of the value,which is a very significant difference. This is dueto the previously discussed factors associated withthis maneuver, aggravated by the fact that due tonoise abatement policies aircraft cannot overfly theCap d’Antibes peninsula unless weather conditionsmake it absolutely necessary. On the other hand,the remaining maneuvers potential does not facesuch a big decline, as SETO goes down to almosthalf and Thrust Reverser and SETI savings are rel-atively constant. APU usage, however, faces almost4 kg more savings for the inbound flight, an increaseof nearly 50%. The result is the reduction of IdleDescent supremacy in terms of relative percentage,

7

with a significant increase in APU contribution, aswell as a small increment for SETO.

4.2. Total Sample AnalysisAfter studying all the routes one by one, a samplewith all the 60 flights was constructed, and the sametype of analysis was put into practice to develop abaseline fuel conservation performance analysis forthe company as a whole, based on the five strategiesapplied in this work.

Figure 6: Total average phase time and fuel con-sumption

Analyzing all the flights, it is clear that Cruiseis the longest and especially most fuel consumingphase overall, followed by Climb and Descent (Fig-ure 6). One special note regarding taxi times, whichreach a total of almost 16% of the total flight time- still a very significant value. Comparing the totaland individual routes results, the similarity betweenthe total results and the LIS-BCN route strikes im-mediately. Being the intermediate-length flight, itmakes sense for the results of the three differentroutes average to fall somewhere in the middle,which explains the similar flight phase durationsand fuel consumption.

Figure 7: Total average savings

Idle Descent, with a little above 2.5% fuel to besaved (Figure 7), is clearly the most significant sav-ing measure, followed by SETO marginally above0.5% and APU with about 0.25%. It is interestingto note that the more a given measure is dependenton ATC coordination, the higher its saving poten-tial, as illustrated by the dominance of Idle Descentand SETO. This, however, is not a coincidence, and

is actually the main reason state of the art softwareand companies (Section 2.1) are trying to attractthe ATC authorities for a transversal effort in theindustry towards optimizing fuel conservation.Route Unit Thrust Rev. APU SETO SETI Idle Desc. Total

Total

kg 0,41 9,19 19,86 2,63 93,22 125,31

Total % 0,01 0,27 0,57 0,08 2,70 3,62

Relative % 0,32 7,33 15,85 2,10 74,39 100,00

Table 4: Total average savings estimates

Regarding the fuel economy estimation throughthe different measures (Table 4), the similarityseems contained only to the BCN-LIS flight, interms of both kilograms and percentage. This is dueto the surprisingly high saving potential for the LIS-BCN flight, clearly the highest overall, percentage-wise. One flight that is even closer to the to-tal average sample value, in terms of the trip fuelto be saved in percentage, is the LIS-NCE flight,at 3.63%. Once again the dominance of the IdleDescent measure over all the others is apparent,and the only instance where this is not the case isthe LIS-OPO outbound flight leg, in which SETOclaims a higher saving potential (Table 1). Overall,this distribution follows the common trend amongall the routes, and allows for defining a priority listamong the different measures, with Idle Descentcoming first, SETO and APU usage following rela-tively close to one another, and SETI and ThrustReverser operation with practically negligible con-tribution.

Figure 8: Average saving per route in %

The fact that most of the implemented opera-tional procedures, with the exception of Idle De-scent, are related to ground operations, allows topredict that the savings potential should decrease,in percentage, with the increase of flight distance -as the relative time spent in the non-airborne stagesof flight should also decrease. Indeed, this tendencyis confirmed by Figure 8, where the shortest city-pair links present higher values of relative savingthan that of the LIS-NCE route overall. However,the LIS-BCN flight presents the highest saving po-tential, on average, and this has to do with the sub-stantial savings for the Idle Descent measure on alloutbound flights, but especially on this route. One

8

can notice the absolute savings from the remain-ing measures, in kg, are relatively constant, whilethe Idle Descent implementation tends to vary sig-nificantly from one city-pair to another, and evenbetween inbound and outbound flight legs for thesame city-pair. This difference is what explains thedominance of the LIS-BCN route, in terms of rela-tive savings, as this route has more kg of fuel to besaved in Idle Descent than any other.

Figure 9: Average saving per route in kg

In terms of absolute savings, it is expected thatthe longer the flight distance is, the bigger its sav-ing potential, however this is not quite the case,as depicted in Figure 9. While the LIS-OPO routepresents the lowest saving potential in kg, as ex-pected, flights departing to and from Barcelona out-weigh those that fly to Nice. Savings potentials foroutbound and inbound flights seem in line with eachother, with the exception of Idle Descent, whichproves this maneuver has a big margin for improve-ment when it comes to the Barcelona flights, andwhile it is fairly straightforward to explain why theOPO-LIS has a smaller saving potential for Idle De-scent than the other two inbound flights (due tolower cruising altitudes and therefore shorter de-scents), it is not quite clear why flights landing inLisbon coming from Barcelona and Nice have sodistinct values, as both flights cruise altitudes andeven approach procedures are fairly similar.

Analyzing all of the relative savings resultsin groups of outbound (blue) and inbound (yel-low) flights, it is even more apparent the under-optimization of all flights leaving Lisbon Interna-tional Airport, and this is mainly due to two rea-sons: higher SETO average savings potential for alloutbound flights, compared to inbound flights forthe same route, which proves Lisbon Airport dif-ficulty to deal with the increasing amounts of airtraffic it receives, and therefore its lower ground effi-ciency; and higher Idle Descent average savings po-tential, especially in Barcelona and Nice, much dueto higher air and ground traffic, therefore stricterATC requirements and regulations, as well as lessfuel efficient approach procedures, with the grow-ing number of paths designed for noise abatementin large cities.

4.3. Savings Estimation

Perhaps as important as calculating how many kgof fuel could be saved via the implementation ofthe measures described in this work, is estimatinghow that translates into yearly savings, in financialterms.

Total

Distance (NM) 32 845

Duration (s) 366 606

Fuel Burned (kg) 207 479

Saving Potential (kg) 7518

Average

Distance (NM) 547

Duration (s) 6110

Fuel Burned (kg) 3458

Saving Potential (kg) 125

Average Saving Potential (%) 3,62

Table 5: Final savings estimates results

Comparing Table 5 with the typical operationalprofile described by Embraer and presented in Sec-tion 1.1.1, it is possible to see that in fact the esti-mates are not too far from reality, regarding averageDistance (547 NM vs. 600 NM), Duration (6110svs. 5400s) and Fuel Burnt (3458kg vs. 3147kg).Of course these estimates would represent the com-pany operation better the more flights and routeswould be analyzed, and possibly come even closer toEmbraer estimates since the LIS-OPO route, for ex-ample, is less flown than the others and represents amuch higher fuel to distance ratio. For a 1% savingsestimation, according to the results obtained fromthe sample, there would be 0.01 × 3458 = 34.5kgof fuel saved per flight, which using Embraer es-timates for a 600 NM flight would result in ap-proximately 591 metric tons of fuel saved per yearfor PGA fleet. In turn, this represents around591 × 446.6 = 263 941$, nearly 235ke, in yearlysavings, which is fairly close to Embraer estimateof 212ke.

However, for PGA case, applying just these fivesimple operational procedures an average total sav-ing of 3.62% was obtained. This means an averagesaving of 125kg of fuel per flight, which for a typi-cal operational profile of 600 NM, according to Em-braer, results in about 2141 metric tons yearly sav-ings in fuel. Considering the average curent price ofjet fuel in Europe of 446.6 $/mt, this means savingsof 956k$, or 850ke per year.

Nearly one million euros in savings if the full po-tential of these five savings measures was attainedthrough its correct implementation is a very notableresult, and certainly very significant for any airliner.It is important to remind that this result is propor-tional to fleet size.

9

4.3.1 Additional Savings

If the full potential of these measures could beachieved under all operating conditions, then PGAwould be transporting 3.65% of unused extra fuel oneach flight, equivalent to 125kg. In turn, this wouldincrease takeoff and landing weight, leading to tripfuel increase and even premature wear of the land-ing gear, brakes and tires, increasing maintenancecosts as well. This means that if these measurescould be reflected in flight planning, reducing theamount of fuel that is uplifted to the aircraft with-out compromising safety, an additional fuel savingcould be attained.

Figure 10: Expected fuel burn increase adding200kg of extra fuel [2]

According to Embraer, for each 200kg of extrafuel uplifted to the aircraft, the trip fuel increasechanges depending on the trip distance as shownin Figure 10. Even though PGA E-190 is the leastinfluenced by extra weight when compared to therest of the E-Jets family, it still presents some im-portant potential. According to the estimations onthis work, the average distance a PGA flight coversis 547NM, and we can see according to the graphthat would translate into approximately 0.25% ofadditional fuel saved, if 200kg of extra fuel was notuplifted. However, the average fuel saving calcu-lated is 125kg per flight, so the only conclusion pos-sible is that the real saving value would be some-where below this number, as there is no informationabout its variation with the weight.

If we assume, for the sake of this estimation,that the saving increases with the weight somewhatlinearly, then we can project an additional savingof about half that presented in the graph, as theweight is reduced from 200kg to 125kg. This wouldresult in a saving estimation of about 0.125%, or4kg per flight, which translates into 68 metric tonsyearly, equivalent to an additional 27ke saved.

5. Conclusions

This project studied the impact of the implemen-tation of fuel efficient operating measures on PGAnew Embraer 190 fleet.

5.1. AchievementsThe results obtained in this work are very encourag-ing, as they show that even with simple day-to-daymeasures, big airlines can save a significant amountof fuel, and therefore, money. However, the fullsavings potential obtained for this work is not at-tainable without the cooperation and consequentoptimization of local ATC itself. These results areimportant for airlines to show the local ATC the im-pact their regulations can have, as away to sensitizethem to this issue and count on their support.

5.2. Future WorkThe next step in terms of airline optimization is ap-plying the same philosophy to both flight planningand maintenance procedures optimization. Also im-portant would be to gather the pilots’ feedback re-garding the measures described in this study, in or-der to assess its feasibility, ease of application andeventual limitations. This would also sensitize thepilots to the potential savings that can come fromtheir direct actions, and thus help with the trans-verse effort that is required within the company toachieve such goals. Unfortunately, such feedbackwas not possible to gather during this work dura-tion, much due to the constant training and flyingthe pilots were attending because of the new fleet.

References[1] EIA Europe Brent Spot Price. http://www.

eia.gov/dnav/pet/hist/LeafHandler.ashx?

n=pet&s=rbrte&f=w. Accessed: 13/09/2016.

[2] EMBRAER. EMBRAER 170/175/190/195Fuel Conservation (GP-1999), November 2004.

[3] IATA Fuel Price Analysis. http://www.

iata.org/publications/economics/fuel-

monitor/Pages/price-analysis.aspx. Ac-cessed: 28/09/2016.

[4] EMBRAER. EMBRAER 190 Airplane Opera-tions Manual (AOM-1502-164), volume 1 edi-tion, January 2016.

[5] Honeywell Aviaso Fuel Efficiency.https://www.aviaso.com/products/fuel-

efficiency/. Accessed: 28/09/2016.

[6] OpenAirlines - SkyBreathe Fuel Efficiency.http://openairlines.com/skybreathe-

fuel-efficiency/. Accessed: 28/09/2016.

[7] PGA. Sandard Operating Procedures ManualE190 (SOP-M), 00 edition, February 2016.

[8] Tasos Nikoleris, Gano Chatterji, and RichardCoppenbarger. Comparison of fuel consumptionof descent trajectories under arrival metering.In Proc. AIAA Guid., Navigat., Control Conf,2012.

10