gy erra ersit Carbon footprint Sustainable development es i g th e be (UCFs) linked to some identified emission macro-sources e i.e., land vehicles, on-ground aircraft, airport nifican rt. Acc identified. Such strategies range from technological developments to better process management, from user consumption guidelines to free-carbon policies and so on. This concept is in line with the requirements of the Kyoto Pro- tocol (United Nations, 1998), an international agreement about uman-related ac- thought to reduce duce several envi- al with high-level egree of the com- also there is a limit is taken to reduce (ICAO) and local (such as the European Aviation Safety Agency e EASA, and the Federal Aviation Administration e FAA) levels the main govern- mental associations are fixing goals, priorities and rules in order to guarantee the sustainable development of the air transport in- dustry at global level and the wellness of local communities as well. In this light, it is desirable that the overall impacts produced by an airport node e particularly CO2 emissions at local level specifically due to the transport function e are estimated and suitably managed (Graham, 2003).* Corresponding author. Tel.: þ39 0512093341. Contents lists availab Journal of Air Trans se Journal of Air Transport Management 37 (2014) 76e86 E-mail address:
[email protected] (L. Mantecchini). about 2% of the global carbon dioxide (CO2) emissions and about 12% of the CO2 emissions within the transport sector (ICAO, 2010). In the last decades, the concept of “carbon footprint” has been developed. It concerns the total amount of greenhouse gases e specifically carbon dioxide e produced to directly and indirectly support human activities. When suitable procedures have been set, the carbon footprint of individuals, nations, organizations and others can be computed so as the amount of carbon impacts pro- duced by them and the suitable strategies to reduce it can be Commission identified transport as one of the h tivities where more efficient policies should be external impacts and greenhouse effects. As for the air transport system, airports pro ronmental impacts, particularly when they de movements. In this case not only the wellness d munities living close to their areas worsens, but to the air traffic growth if any effective measure impacts. For this reason, both at international figures (IEA, 2009) transport systems account for 26.4% of the total carbon impacts due to human activities. Aviation accounts for to build a common framework among Member States to deal with climate changes (European Commission, 2011). Particularly, the EU 1. Introduction Environmental impacts are a sig human activities, included transpo http://dx.doi.org/10.1016/j.jairtraman.2014.03.001 0969-6997/� 2014 Elsevier Ltd. All rights reserved. amount of CO2. Particularly, UCFs due to transport activities have been defined according to some relevant transport variables. The computation of UCF values for a given airport allows computing both the contribution of each macro-source and also evaluating the effectiveness of transport-related actions aiming at reducing the carbon impact. The methodology has been applied to the airport of Bologna, in Northern Italy, and its UCF values for the identified macro-sources have been computed. � 2014 Elsevier Ltd. All rights reserved. t consequence of many ording to some recent emission reductions in force as of February 2005 adopted by several Countries. In the light of sustainability, and according to the Kyoto Protocol, the EU recently proposed a series of long-term policy actions so as Air transport Airport impacts handling and terminal equipment e to compute the contribution of the single macro-source to the total Keywords: this paper, a Transport Carbon Footprint methodology has been set to identify Unit Carbon Footprints A transport carbon footprint methodolo emissions Maria Nadia Postorino a, Luca Mantecchini b,* aDepartment of Civil Engineering, Energy, Environment and Materials (DICEAM), Medit bDepartment of Civil, Chemical, Environmental and Material Engineering (DICAM), Univ Risorgimento 2, 40136 Bologna, Italy a r t i c l e i n f o Article history: Available online a b s t r a c t Airports are important nod Emissions of CO2 are amon policies and guidelines hav journal homepage: www.el to assess airport carbon nea University of Reggio Calabria, Italy y of Bologna, School of Engineering and Architecture, Viale del n the air transport system, but also local sources of environmental impacts. e most relevant ones because of their potential greenhouse effects. Many en identified at national and world level to reduce such kind of impacts. In le at ScienceDirect port Management vier .com/locate / ja i r t raman characteristic as well as engine performances. However, the goal of this paper is to estimate the airport carbon footprint, due to the relevance that carbon emissions have in increasing the greenhouse effect. In the following, after an overview on airport relevant carbon sources (Section 2), the transport carbon footprint methodology (Section 3) and the airport test case (Section 4) are described. Then, the transport carbon footprint methodology is applied to the test case (Section 5) and some conclusions following the discussion are drawn (Section 6). 2. Airport operations and relevant carbon sources Airports are interchange nodes where people transfer from land to air transport modes and vice versa (Fig. 1). Particularly, a continuous flow of arriving passengers is transformed in a discrete flow of departing aircraft and vice versa. As showed in Fig. 1, air travellers start from/arrive to the “airport catchment area” generally by using the surface access network. The size of the airport catchment area depends on many factors such as airport geographical position, surface connecting transport system, socio-economic features of potential users and existence of competing airports. The air travel demand level at a given airport depends on the airport catchment area characteristics, while the percentages of users on available surface transport modes to access the airport depend on the surface transport network. The air travel demand on mode m, dim, can be estimated by direct surveys or by using stage demand models emainly within a discrete choice approach e of the form: f Air Transport Management 37 (2014) 76e86 77 By following the indications of the Kyoto Protocol, some guidelines specifically dealing with airport greenhouse gas reduc- tion have been produced (e.g., ACRP, 2009; ACI, 2009). The airport carbon footprint should be estimated and strategies thought to reduce the greenhouse emissions in order to be less than pre- scriptive environmental constraints, if these ones have been fixed by National Governments. The importance of the airport carbon footprint was recognized by ICAO since 2007, when the work “Carbon emission calculator”, successfully launched in 2008, star- ted and was then followed by other reports (e.g. ICAO, 2010, 2011). Other Organizations and Regulators promoted the development of plans and actions finalized to the estimation of airport carbon ef- ficiency. As an example, the Airport Carbon Accreditation pro- gramme launched in 2009 by ACI Europe has the aim to reduce airport carbon emissions and, ultimately, neutralize their carbon footprint. The future development of the air transport industry as a driving force for the socio-economic growth of the society is conditioned to the ability to reach e and tenacity to preserve e the sustainability of civil aviation operations. This will be possible only under coor- dinated activities by the concerned stakeholders, with special re- gard to the assessment of environmental impact of flight operations and to the development of green technologies and specific policies to ensure an optimum and harmonious equilibrium between aviation and environment. Generally, more attention has been devoted to the reduction of noise, particularly in terms of aircraft technological improvements (e.g., Aircraft Noise Certification as fixed by the Annex 16, ICAO) and/or constraints at airports (e.g., reduction of the movements during some day periods, following the EU Directive 2002/49/EC). This kind of aircraft impacts as well as social and monetary ones have been studied in the literature (as in Yu et al., 2004; Schürmann et al., 2007; Unal et al., 2005; Lu and Morrell, 2006), but always with reference to specific aspects. A comprehensive approach able to identify the “airport carbon footprint” due not only to aircraft- related operations but also to land airport activities (e.g., handling) and surface connecting transport modes has been pro- posed by Postorino (2010a). The goal of this paper is to introduce a “transport carbon foot- print” (TCF) methodology for a given transport (sub)system or transport node. Here the specific case of airport nodes is considered and the TCF methodology has been set to identify the airport car- bon footprint by following a comprehensive approach. More in details, the aim is to set a methodological approach able to help to identify the percentage contribution of carbon emission macro- sources in order to evaluate the effectiveness of transport related policies aimed at reducing the airport carbon impact. Particularly, transport quantities e such as travellers, ground access transport modes and aircraft e are relevant determinants of several carbon emissions macro-sources due to direct and indirect consumptions (e.g., energy required to move ground vehicles and aircraft, energy required for passenger terminal functions). Starting from a logical operating architecture, first relevant transport variables, to which some main carbon emission macro-sources are related, are iden- tified. Then, for each macro-source the Unit Carbon Footprint (UCF) related to its relevant transport variable is defined. The aim is to have reference unit indicators (UCF) e whose specific value de- pends on the considered airport e that measure both the contri- bution of the corresponding macro-sources to the whole airport carbon impact and then the expected reduction due to transport related actions directly operating on the transport variable that generates more relevant impacts. It is worthwhile to note that other important airport-related emissions are nitrogen and sulphur oxides (NO and SO ) and M.N. Postorino, L. Mantecchini / Journal o x x particulate matter (PM), which depend on the aviation fuel doðs; k1; k2; :::; knÞ ¼ no Y c pðkcÞ where do(.) is the travel demand with origin o, travelling for trip purpose s and characterized by the choice dimensions k1, k2,., kn; Holding apron Taxiway exit Parking/internal circulation Catchment area and surface access network Aircraft Passengers La nd si de A ir si de Airport traffic zone/Airways Runways Taxiways Apron Terminal A TC Fig. 1. Airport operations. no is the number of potential travellers at origin o; p(kc) are the choice percentages referred to the kc choice dimension such as the origin and/or destination airports, the surface connecting modes, the air services at airports (see for example Harvey, 1987; Innes and Doucet, 1990; Pels et al., 2003; Kroes et al., 2005; Suzuki, 2007; Hess, 2010). The percentages of users on available surface transport modes, say p(m), are often estimated by using random utility models such as Logit, Nested-Logit, Cross-Nested Logit models (see for example Train, 2003, for an overview), but also by direct surveys. The estimation of the car (or lorries for goods) percentages is particularly desirable because cars (lorries) contribute to the greenhouse effects due to their carbon dioxide emissions. At a first level these emissions can be computed by considering the (average) taxiway configuration and features (as route length, high-speed taxiways, last-minute start-up) then the amount of taxi/idle can vary significantly from airport to airport. Generally speaking, the most important aircraft emissions are carbon dioxide (CO2), water vapour (H2O) in terms of condensation trails, nitrogen and sulphur oxides (NOx and SOx) and particulate matter (PM). An estimate by the Air Transport Action Group verified that new aircraft are 70% more fuel-efficient than they were 40 years ago. Furthermore, carbon monoxide emissions have been reduced by 50%, while unburned hydrocarbon and smoke have been cut by 90%. Then, old aircraft use much more fuel per passenger-km than new aircraft of similar size, but in any case airlines tend to optimize the use of their aircraft in order to reduce operational costs and then fuel consumption. Many research efforts M.N. Postorino, L. Mantecchini / Journal of Air Transport Management 37 (2014) 76e8678 characteristics of circulating vehicles so as to identify the carbon contribution at airports. However, the carbon contributions can also be referred to the overall trip to/from the airport. In this case, the travelled mean distances should be estimated in order to compute the carbon contribution at aggregated level or, more in details, by using road demand assignment models (Sheffi, 1985). Still in terms of landside, the energy consumption for terminal operations are another source of emissions linked to the expected number of passengers at the airport. In fact, terminal facilities (such as check-in and security control desks; ticket offices; baggage fa- cilities, but also restaurants and retail spaces) require energy and then can be considered sources of carbon emissions according to the way such energy has been produced. On the airport airside, the main sources of air pollution are exhaust from aircraft and diesel engines, direct fuel emissions from refuelling aircraft and larger dust particles mainly from brakes, tyres, asphalt. The number of movements at a given airport depends on the expected air transport demand, and furthermore the related aircraft environmental ground effects depend on the level of technology and on the operating fleet mix. Both passengers and goods should be considered, but goods volumes are generally smaller than passenger volumes (Eurostat, 2011) although situa- tions can be very different from Country to Country. The number of aircraft (or movements) able to support the expected air travel demand at the airport depends on the aircraft average load factor. Furthermore, depending on the number of movements, the number of land vehicles operating within the airside can also be computed. Both aircraft and handling vehicles produce impacts at ground level (Fig. 2). Ground-based aircraft emissions can be estimated on the basis of the main operations performed at the airside, i.e. landing/take- off operations and taxiing. The percentage of aircraft-related emissions during landing and take-off depends mainly on the aircraft characteristics, while for taxiing it depends also on the Aircraft average (breakeven) load factor Airport user demand in time interval T Number of movements in time interval T Carbon impact contribution Fig. 2. On ground aircraft and handling eq have been made to test alternative fuels for aviation, particularly new generation of sustainable bio-fuels to contribute to a more friendly transport development. Apart from landing, take-off and taxiing, on-ground aircraft emissions include those associated with the use of auxiliary power units (APUs) at gates (IPCC, 1999). APUs are engine-driven gener- ators providing the aircraft with necessary energy (e.g. for air conditioning, lights, and so on) during the time it is at the gate. Then, aircraft continue producing emissions also when they do not move. Fuel used by APUs is only a small part of the overall aircraft fuel consumption (according to some estimates it ranges from 0.8% to 3.5% of the overall aircraft fuel) and the energy required by them can also be obtained by airport ground-based electrical equipment. Finally, handling equipment is the other source of carbon di- oxide emissions at airports for operations such as refuelling, cleaning, catering, baggage/staff/crew transfer. Handling-based emissions depend on the number of movements, the flight sched- uling, the airside size and configuration as well as on the charac- teristics of the handling vehicles themselves (age, type of fuel, fuel efficiency). Then, according to the airside and vehicle characteris- tics, handling operations can produce significant carbon emissions. 3. The transport carbon footprint methodology The logical operating architecture of the transport carbon foot- print (TCF) methodology is represented in Fig. 3. As regards methods and procedures suggested by ACI (2009) and ACRP (2009), the proposed one is different because it does not simply suggest a structural design or an inventory of emission sources, but mainly helps planners/operators to forecast the effects of actions aimed at reducing the airport carbon footprint thanks to the use of Unit Carbon Footprint (UCF) as described below. Particularly, unlike other proposed methods and procedures the starting point is the identification of some relevant transport variables to which some corresponding carbon emission macro-sources e and then the Handling equipment per aircraft Aircraft ground-based activities - On-board activities (e.g., air conditioning) - Landing/Taxiing/Take-off Carbon impact contribution uipment carbon impact contribution. - Landing/ Take-off/ - Taxiing Number of movements in time interval T Handling equipment per aircraft on: ns A Handling vehicles and equipment rt TC f Air Transport Management 37 (2014) 76e86 79 associated UCF e can be related. In fact, airport primary function is to guarantee that goods and passengers can be carried from one place to another by using aircraft and then the reference initial variables should be transport ones. The most relevant and easily available transport variables to which carbon emissions are related are the airport air demand (identified in terms of yearly passengers, pax) and the number of movements at the airport (identified in terms of yearly movements, Airport user demand in time interval T Carbon impact contribution Average load factor Ground (car) access mode Carbon impact contribution Energy consumpti - APU - Airport operatio Carbon footprint Passengers UCF Airport catchment area Fig. 3. The airpo M.N. Postorino, L. Mantecchini / Journal o mov). It should be noted that the identification of the yearly pas- sengers as transport variable allows taking into account not only the demand generated from/attracted to the catchment area, but also the transfer demand at hub airports that, in turn, relates to the number of required aircraft. On the other hand, the estimate of the air demand in the airport catchment area allows computing the carbon contribution due to the use of land vehicles for airport ground access. Air demand and movements contribute to the carbon dioxide emissions both directly (e.g., direct vehicle emissions due to pas- sengers using car for ground access, aircraft and handling vehicles) and indirectly (e.g., energy consumption/production due to handling equipment and terminal operations). Then, the airport carbon emission macro-sources linked to the identified variables are land vehicles going in/out the airport, on-ground aircraft, land vehicles for handling operations, airport handling and terminal equipment. The share of CO2 associated to each macro-source can be esti- mated so as to obtain the UCF for each of them. More in details, if MS is the identified macro-source of CO2 emissions that can be associated to the transport relevant variable RV and TAMS the total amount of CO2 due to MS, then the corresponding UCFRV-MS can be computed as: UCFRV�MS ¼ TAMS=RV (1) UCFs linked to the identified macro-sources and relevant vari- ables are defined so as to compute the contribution of the single macro-source to the total amount of CO2. Obviously, the specific values will be different from airport to airport due to the charac- teristics of RV, MS and TAMS. The macro-sources directly associated to passengers and their UCF are identified in: - ground access modes (car %) (UCFpax-car) - energy produced/consumed for airport terminal activities (UCFpax-energy). Carbon impact contribution Carbon impact contribution IRPORT Prescriptive constraints Movements UCF F methodology. The macro-sources directly associated to movements and their UCF are identified in: - on-ground aircraft during landing, take-off and taxiing e i.e. standard LTO cycle (LTO-T) e operations (UCFmov-LTO-T) - handling vehicles and airport equipment (UCFmov-handling). LTO cycles depend on the aircraft class and then they should be estimated by considering the aircraft mix at a given airport. According to ACI (2009), emissions sources can be categorized in “scopes”, labelled as 1, 2, 3 (Table 1). Here, ground access modes could be identified in terms of Scope 3, on-ground aircraft operations and handling vehicles could be still identified as Scope 3, but airports can take direct actions to reduce Table 1 “Scope” categories of sources. Scope 1 GHG emissions from sources that are owned or controlled by the airport operator Scope 2 GHG emissions from the off-site generation of electricity (and heating or cooling) purchased by the airport operator Scope 3 GHG emissions from airport-related activities from sources not owned or controlled by the airport operator. Scope 3A: emissions which an airport operator can influence, even though it does not control the sources Scope 3B: emissions which an airport operator cannot influence to any reasonable extent. Source: ACI, 2009 The handling equipment is summarized in Table 4. According to Air them (e.g., alternative fuelled handling vehicles, proper taxiway and taxiway exits design), while energy produced/consumed for airport terminal activities and airport equipment can be identified in terms of Scope 1 and 2. This classification also helps to identify operational boundaries for the emission computation according to the already cited ACRP report (2009). Passengers and movements can be considered both indepen- dent inputs, but also as dependent variables. More in details, starting from the airport demand as indepen- dent input, for example, the number of movements can be esti- mated by using an average load factor for that airport. In turn, the number of handling vehicles (equipment) operating on the airside can be estimated by using the average number of handling vehicles (equipment) per aircraft. Then, the macro-sources associated to movements and their UCF are obtained indirectly from passengers: - passengers/ (load factor) / (nr. of movements)/ UCFmov- LTO-T - passengers/ (load factor)/ (nr. of movements)/ (handling vehicles and airport equipment per aircraft)/ UCFmov-handling. The number of passengers too can be considered the result of an air demand model (Postorino, 2010b) applied to the airport catchment area (Fig. 3) and then the macro-sources associated to both passengers and movements and their UCF are obtained indi- rectly from the transport demand model: - catchment area/ (demand model)/ (passengers)/ (car %) / UCFpax-car - catchment area/ (demand model)/ (passengers)/ (energy produced/consumed)/ UCFpax-energy - catchment area / (demand model) / (passengers) / (load factor)/ (nr. of movements)/ UCFmov-LTO-T - catchment area / (demand model) / (passengers) / (load factor)/ (nr. of movements)/ (handling vehicles and airport equipment per aircraft)/ UCFmov-handling According to the short overview presented in Section 2, there could be several approaches and models/methods to quantify transport demand, emission factors and fuel consumption. The aim of this paper is not to provide specific models for each element, but an architecture to forecast the impacts of actions aimed at reducing the carbon emissions. In fact, when the UCFs have been estimated, according to the expected values of the transport relevant variables the corresponding transport carbon footprint can be estimated and the impacts evaluated. More in details, the several macro-sources could have different unit impacts, then the effectiveness of the planned actions can be measured so as to identify the most suc- cessful one. Moreover, if prescriptive environmental constraints have been set, the expected carbon impacts due to several planned actions can be compared with these ones. In other words, the procedure helps to forecast the environ- mental effects of airport developing strategies particularly in terms of carbon footprint produced, directly or indirectly, by transport variables. From the stakeholders point of view, the UCF associated to one or more relevant variables taking into account the transport function can be used to identify what are e if any e the critical aspects of a given airport growth plan and what kind of measures can be adopted to assure a sustainable development. According to the classifications in terms of scopes (ACI, 2009), not all the emission (macro-)sources are owned by airport operators directly asked to reduce the airport carbon footprint, but this remarks the need to create synergies with planners and local/national au- thorities as the final goal is the achievement of sustainability and M.N. Postorino, L. Mantecchini / Journal of80 better life quality. data recorded by the airport operator, for each aircraft operation there is on average the use of: 1 ASU, 1 GPU, 1e2 conveyor belts, 1e 2 passenger steps, 1e2 apron bus, 1 TOW; 1 toilet and water truck, 1 trailer, 1 cargo loader and 1e2 cars. During the year 2012, there were 67,257 aircraft movements at Bologna airport. The fleet mix (Table 5) is mainly composed by narrow-body aircraft (60.2%), regional jets and commuter aircrafts (35.2%), turbo-prop aircraft (2.1%). Wide body aircraft account only for 2.5%. Finally, Table 6 summarizes the consumption of primary and As final comment to conclude this section, transport variables as well as forecasted emission estimates are intrinsically unreliable due to the nature of the involved quantities and phenomena (de Neufville and Odoni, 2003) and this is a common issue in trans- port simulation and planning. The TCF methodology helps to identify possible future scenarios the boundary conditions being known, expected or supposed. As generally happens when dealing with transport issues, the shorter the forecasting period the better the predicted scenario. 4. Bologna international airport The city of Bologna, in Northern Italy, has a strategic geographic position with a good land accessibility by road and high speed trains, making the Bologna’s airport catchment area very wide (Fig. 4). The airport is located about 5.5 km from the city centre and there is an aero-bus service connecting the airport terminal, the rail station and the city centre. The catchment area (estimated by means of direct surveys) has recently modified due to the effects generated by the introduction of the high speed train link from Milan to Rome, via Bologna and Florence. The journey time be- tween Bologna and Florence is currently around 30 min and more than 100 daily trains are scheduled. The airport catchment area can be defined with reference to land extension or user demand. In the first case it represents the area generating (attracting) travellers at an airport, i.e. the area where potential air travellers for a given airport start or end their air trip. In terms of user demand, the airport catchment area can be defined as the number of air travellers using a given airport. The airport has a two-floor passenger terminal, recently restored and upgraded. The terminal total surface is 36,100 m2 e 5500 of them are dedicated to commercial activitiese and there are 24 boarding gates and 10 security checkpoints. There are also a general aviation terminal, a freight area with cargo facilities, two airport handlers with complete facilities and equipment. The airport serves 104 airports in 34 countries. During the year 2012, Bologna airport handled almost 6 million of passengers, with an increase of 1.2% compared to 2011. However, the overall aircraft movements were 67,257 (�0.4% compared to 2011). The analysis of the airport ground access modes, made by direct counts and passenger surveys, shows a predominance of cars (private, accompanying, rented or taxi) due to the efficiency and capacity of the road network and to the proximity of the airport to the city centre, although the last years have shown a trend towards a modal share balance. The results are summarized in Table 2. In terms of airside and equipment, the airport has a single runway (12e30 oriented), whose main characteristics are sum- marized in Table 3. The hourly peak mixed capacity is 24e26 movements/h; there are 11 taxiways and two apron areas (Fig. 5) named area M (92,500 m2 and 13 parking spaces) and area W (63,000 m2 and 13 parking spaces). Transport Management 37 (2014) 76e86 secondary energy sources registered in 2012. M.N. Postorino, L. Mantecchini / Journal of Air Transport Management 37 (2014) 76e86 81 5. Application of the TCF methodology to Bologna airport test case The TCF methodology has been applied to estimate the carbon footprint of Bologna airport, by considering the several contribu- tions as depicted in Fig. 3. The 2012 data base has been used. According to Fig. 3, four different macro-sources of CO2 emis- sions have been considered: - average emissions due to ground access modes; - average emissions due to energy production and consumption for airport terminal activities; - average emissions due to landing, take-off and taxiing on- ground aircraft; - average emissions due to handling vehicles and airport equip- ment (ground support equipment, APU). The simulation and approximation of CO2 unit emissions due to several sources have been widely studied in the literature. Partic- ularly, for the aim of this paper the reference unit emission factors Fig. 4. Bologna airpor Table 2 Airport ground access mode share. Airport access mode Share Private car 28% Accompanying car 26% Rented car 5% Taxi 14% Bus 27% Source: Bologna airport, 2012 due to airport sources e such as handling vehicles and airport equipment e have been taken by Greenhouse Gas Protocol Italian ISPRA, Italian greenhouse gas inventory 1990e2010, among some available others (e.g. the previously cited ACRP Report 11, 2009 and the Airport Air quality guidance manual ICAO Doc 9889, 2011). Starting from Fig. 3, the first considered emission sources are passengers using the car as airport ground access mode. Such emissions should be estimated carefully by taking into account both in and out airport car traffic flows. Most airports periodically conduct surveys and have fixed or mobile vehicle count data sys- tems at suitable places close to the airport area. Two main approaches can be considered to obtain such emis- sion estimates. The first one (A), more approximated, considers the estimated number of daily vehicles accessing the airport and the average travelled distance. The second approach (B) is a detailed and accurate one. It considers the airport access road network and evaluates the direct and marginal effect of airport-related traffic emissions. In particular, the marginal effect includes the emissions due to the increased congestion on the road network generated by airport car traffic. Both approaches need the estimate of the average daily road vehicles accessing the airport. To obtain this value, the share of air travellers using car to access the airport has been estimated starting t catchment area. Table 3 Runway characteristics. RWY designation Dimension (mt) RWY strength and surface Approach type 12 2803 � 45 PCN 71/F/B/X/T ASPH ILS cat I/IIIb 30 2803 � 45 PCN 71/F/B/X/T ASPH ILS cat I M.N. Postorino, L. Mantecchini / Journal of Air Transport Management 37 (2014) 76e8682 from the transport model available for the Bologna Municipality area. More in details, the whole road network for this area if made by 9158 links and 287 traffic analysis zones. The average daily link traffic flows are available for all the links of the network. However, only the traffic flows on the relevant ground airport access road networke extracted from thewhole available road networke have been considered (Fig. 6). According to this estimates, the number of daily road vehicles accessing the airport is about 7000 veh/day on Fig. 5. Bologna aerodrome chart. Source: ICAO average (2012 data). This value has also been confirmed by some direct traffic counts carried out on the relevant airport access roads. Assuming an average occupancy coefficient equal to 1.8 (a common value in the area, also indirectly confirmed by some oc- casional counts), the estimated number of yearly passengers (reference year 2012) using the car to access the airport is 4,536,000. This value also corresponds to about 70% of the yearly number of passengers handled by the airport during the year 2012 and is coherent with the estimated mode share reported in Table 2 (73% by considering private cars, accompanying cars, rented cars, taxis). Table 4 Handling equipment (reference year 2012). Type of equipment Total number Airstart unit (ASU) 2 Aircraft cooling unit (ACU) 1 Ground power unit (GPU) 11 Conveyor belts 11 Passenger steps 19 Apron bus 8 TOW-TUGS 3 TOW-BARLESS 5 Toilet service truck 2 Potable water truck 1 Trailers 24 Cargo loaders 6 Transporter loaders 2 Cars 20 Minibus 3 According to the approach A, ground access vehicles emissions can be evaluated by considering the estimated daily number of vehicles accessing the airport and the average travelled distance. This approximated approach is more used in the common practice because it simply needs the use of the two previous data. The estimated number of daily vehicles is 7000 veh/day, as described before. The average travelled distance estimated by airport specific surveys is 25 Km. To compute the emissions due to the number of vehicles for the estimated travelled distance, the average fuel emission factors have been taken by official data (ISPRA, 2012) according to the average national share between gasoline and diesel fuel (53% gasoline and 47% diesel fuel, ISTAT, National Institute for Statistics, Annual Report, 2012). The yearly emission value is then 20,823,250 kg/year. According to the approach B, the CO2 emissions caused by air travellers using cars to access the airport and including marginal effects have been estimated starting from two initial known data from previous airport surveys: 1) the total amount of km covered by airport vehicular traffic during a peak reference hour e available from the relevant ground airport access road network introduced before; 2) the average link travel time during the same peak reference hour referred to a unit length (1 km) link, on the same relevant ground airport access road network. The choice of the peak hour is a common rule so as to identify the most critical situation. The effect of congestion can be measured by comparing the reference time needed to travel along a unit length link at the reference speed and the actual time derived from the previous Table 5 Bologna airport fleet mix (reference year 2012). Aircraft type % Narrow body (A319, A320, A321, B737, MD80) 60.2% Regional jets and commuter aircraft (CRJ2, CRJ7, CRJ9, Embraer190, DH, F70, RJ85) 35.2% Wide body aircraft (B747, A330, B767, B757, Antonov124) 2.5% Turbo-prop aircraft (Atr72, F50, SB20) 2.1% surveys. It is well known that the actual travel time is greater than the reference one due to the effects of congestion (Sheffi, 1985; Highway Capacity Manual, 2010) and the emissions linked to the travelled unit length link are then greater than those computed without considering congestion. This effect can be conveniently measured by referring to an increased distance at the reference speed e virtual travelled distance in the following e so as to obtain the same travel time as during congestion. The estimated average virtual travelled distance coherent with the actual travel time is 45.5 km, i.e. about 82%more than the actual one. The main input data are shown in Table 7 and the estimated amount of yearly CO2 emission is 37,858,060 kg/year. In the following, the value obtained by using the more precise approach has been considered, so as to take into account the increased congestion on the road network due to the airport presence. As for the second identified relevant variable (see Fig. 3), the the airporte starting from the cruising altitude and then along the glide path e and it is followed by the landing and taxiing to the gate. Then, the cycle continues when the aircraft taxis back out to the runway for take-off and climbs out to the airport traffic zone. For Bologna airport, the LTO CO2 contribution (Table 8) have been evaluated by using EDMS, one of the most common and popular aircraft emission models, developed by FAA since 1998. EDMS uses the EUROCONTROL Base of Aircraft Data (BADA, http://www. eurocontrol.int/services/bada) for aircraft performance modelling and emissions calculation. The database, constantly updated, con- tains the emission factor of the 90% of the aircraft types. The used data set refers to year 2012 and concerns aircraft fleet mix, operation time, measured taxi-in and taxi-out time. Concerning ground support equipment, the emissions have been computed by considering the fleet of ground support vehicles (as reported in Table 4), the corresponding fuel consumption and the emission unit reference factors for gasoline and fuel oil (Table 9). Finally, the emissions from airport equipment stationary sour- ces have been grouped into two categories: electrical energy usage and in-situ combustion activities. According to the most recent estimates at national level (ISPRA, 2012), the share of CO2 emissions from electrical energy con- sumption in airport is 396 gCO2/kWh. This emission factor results from the 2010 Italian electricity mix (38.5% crude oil; 36.3% natural gas; 13.1% primary electricity; 8% solid fuel; 4.4% renewable). The Table 6 Airport energy consumption (reference year 2012). Source of energy Reference year: 2012 Methane 588,420 (m3) Heating oil 196,419 (kg) Electrical energy 14,193,321 (kWh) Photovoltaic energy 76,916 (kWh) M.N. Postorino, L. Mantecchini / Journal of Air Transport Management 37 (2014) 76e86 83 number of movements correspond to landing and take-off aircraft operations. In 2011 total greenhouse gas emissions in Italy from this source category were about 2.0% of the national total emissions from transport. According to the IPCC Guidelines and Good Practice Guidance (IPCC, 2006) and the EMEP/CORINAIR Guidebook (EMEP/ CORINAIR, 2007), the emissions are estimated starting from the number of aircraft movements broken down by aircraft and engine type (derived from ICAO database if not specified). Aircraft CO2 emissions and fuel consumptions during the Landing-and-Take-Off (LTO) cycle are due to two single aircraft operations e landing and take-off. The standard LTO cycle starts whit the aircraft approach to Fig. 6. Airport ground access road network. Source: Bologna airport. electrical energy consumed during the year 2012 is reported in Table 6 (14,193,321 kWh) and then the CO2 due to these sources has been quantified in 5,620,555 kg. These emissions are ascribed to passengers because most part of the energy has been consumed for terminal-related passengers activities. As regards the in-situ combustion activities, their contribution to the CO2 emission has been estimated in a similar way, by using reference unit emission factors provided by ISPRA (2012) and data in (Table 6). The results are shown in Table 10. The methane con- sumption has been ascribed to passengers because it mainly refers to terminal heating operations, while the diesel consumption is Table 7 Airport ground access network relevant data. Input data Road airport sub-network length 210 (km) Average virtual travelled distance 45.5 (km) Table 9 CO2 Emission for ground support equipment vehicles. Source Emission factor Consumption (2012) CO2 emission (kg) Gasoline 2.2536 (kg/l) 17,966 (l) 40,488 Fuel oil 2.6604 (kg/l) 434,272 (l) 1,155,337 Source for emission factors: ISPRA, 2012 M.N. Postorino, L. Mantecchini / Journal of Air Transport Management 37 (2014) 76e8684 ascribed to handling operations because it mainly refers to engine generators. The total amount of CO2 emitted during the reference year 2012 from the identified airport sources related to the transport function has been computed as the sum of the several contributions (Table 11). The estimated airport carbon footprint due to the transport function can be ascribed to passengers andmovements (see Section 3). Table 12 reports the UCF referred to the several macro-sources computed by starting from the values of CO2 in Table 11. To summarize the results obtained in Table 12, the first two macro-sources are associated to passengers, while the other two are associated to aircraft movements. According to Eq. (1), the UCF for each macro-source has been obtained by dividing the total amount of CO2 ascribed to the given macro-source by the corre- sponding relevant variable. By assuming that no mitigation actions are taken and the boundary conditions are the same (in particular for ground access modal share), the expected carbon footprint linked to the relevant transport variables can be obtained by referring to forecasted passengers and movements for the short period. Some estimates predict an average increase of the travel demand at the airport by 2013 of 3.8% (Bologna airport traffic statistics, update Sept. 2013) and the expected number of passengers is so 6,184,814. At the same time, a reduction in the number of movements is forecasted at a rate of �3.4% and the expected number of movements is 65,232. According to Table 12, the expected carbon footprint is summarized in Table 13. Very briefly, the increase of the load factor, all the other things being the same, as a consequence of the movement number decrease produce a slight reduction of the CO2 emission thus confirming an improvement of the environmental performances. 6. Summary, discussion and future developments The TCF methodology presented in this paper proposes a gen- eral approach to identify Unit Carbon Footprints (UCFs) linked to some relevant transport variables so as to compute the unit contribution of the identified emission macro-sources to the total amount of CO2. Transport variables have been defined in terms of passengers and movements, while macro-sources have been identified coher- ently with the guidelines proposed by ACI (2009) and ACRP (2009). In fact, carbon impacts produced at an airport are due not only to aircraft but also to ground vehicles (such as handling vehicles; cars used by travellers to go in/out the airport). Following the TCF Average travel speed 37.2 (km/h) Average fuel emission factor 163 (g/km) methodology, average emissions due to ground access modes and emissions due to energy production and consumption for airport terminal activities have been ascribed to the airport passengers, while emissions due to landing, take-off and taxiing on-ground Table 8 CO2 emissions due to aircraft operations e Bologna airport, reference year 2012. Jet fuel consumption (kg) 14,169,247 CO2 emissions (kg) 45,089,125 aircraft; emissions due to handling vehicles and airport equip- ment (ground support equipment, APU) have been mainly referred to the number of aircraft movements. Concerning the relevant variables, they not only synthesize relevant transport characteristics, but also the characteristics of the catchment area, the average aircraft load factor, the handling equipment per aircraft. The application of the TCFmethodology and the estimates of the UCFs to the test case of Bologna airporte one of themost important Italian national airport e suggests some interesting comments. First of all, the general results are in line with the existing literature. Particularly, according to some recent studies (BAA London Gatwick Airport, 2009; Budd et al., 2011; Miyoshi and Mason, 2013; Ryley et al., 2013), and as obtained in Tables 12 and 13, emissions due to passenger to and from airports are one of the main causes of airport-related environmental impacts. The other still more important carbon emission contribution is due to aircraft, particularly during their LTO cycle (Dodson et al., 2009; Song and Shon, 2012). Apart from the specific results that are in line with the above literature, the main emissions at the airport belong to Scope 3 (CO2 emissions from ground access vehicles and LTO cycle) and then are outside the airport control thus suggesting the need to adopt a broader point of view when identifying sustainable developing strategies. Also due to airport ownership types e particularly gov- ernment or mixed public/private owned e synergic transport ac- tions with Local Authorities can be adopted to reduce impacts due to the airport related activities such as car access. As an example, improved public transit systems may help to reduce the passenger surface access contribution to the airport carbon footprint. One important issue, already addressed in several contexts, is that there is no single control of the problem and then future legislation needs to regulate emissions at airports through reduc- tion of CO2 by involving several actors directly and indirectly responsible for them. Inventories of emission sources and classifi- cation of Scopes, suggested by ACI (2009) and ACRP (2009), are some of the tools to achieve such goal. The identification of re- sponsibilities may have significant implications in the effectiveness of the carbon mitigation (Daley and Preston, 2009). To give an example, in some cases airport operators accept responsibility for aviation emissions produced during the whole LTO cycle, while in some others they limit their duties to aircraft emissions generated at the gate or during manoeuvres on the apron and taxiways. From a wider point of view, where many actors could be involved, both directly and indirectly, the carbon footprint reduc- tion can be obtained by adopting specific measures referred to the involved transport relevant variable starting from the estimated Table 10 CO2 Emission for airport energy production. Source of energy Reference unit emission factor Consumption (2012) CO2 emission (kg) Methane 1.958 (g/m3) 588,420 (m3) 1,152,126 Diesel 3.112 (g/kg) 196,419 (kg) 611,256 Source for reference unit emission factors ISPRA, 2012 Regarding passengers and the related macro-sources, useful to the increase of car traffic flows in its neighbouring, some specific measures targeted at reducing such airport contributions concern Table 11 Summary of estimated airport CO2 emissions due to the transport function. Source Emitted CO2 (kg) Induced vehicular traffic 37,858,060 Electrical energy 5,620,555 Energy produced in airport from methane 1,152,126 Energy produced in airport from diesel 611,256 Aircraft LTO 45,089,125 Handling ground equipment 1,195,825 M.N. Postorino, L. Mantecchini / Journal of Air UCFs e which represent the unit emission due to the several macro-sources. Regarding movements and the related macro-sources, useful measures mainly concerns fuel efficiency and improvement of the airport energy efficiency. Fuel efficiency in aircraft engines e particularly during the LTO cycle e means less fuel burned and lower emissions, but good re- sults in terms of reduced emissions can also be obtained by the use of alternative fuels, particularly biofuel (Rosillo-Calle et al., 2012). EC-funded R&D projects (www.biofuelstp.eu/air.html#studies) aimed at checking the use of biojetfuels in commercial aviation tried to give answers to both regulators and commercial aviation operators so as to reduce the climate negative effects, cut costs and reduce volatility of fuel supply. From the airport point of view, encouraging airline carriers to improve their fleet is an important issue to reduce the airport carbon footprint, as well as the use of larger aircraft or the increase of load factors so as to decrease the number of movements e the number of passengers being constant. Furthermore, airports can also act directly by improving the airport layout, mainly the taxiway system. In fact, improving taxiways, air-side terminal and runway configurations to reduce taxiing distance and terminal area congestion e as well as efficient gate allocation techniques to minimize taxiing time after landinge helps to reduce the emissions due to taxi-in/out. Some regulatory measures can also be thought to reduce impacts, such as emission-based incentives and landing fees, but this kind of measures is more controversial. As for the ground support equipment, here some improvements can be obtained by converting gasoline and diesel powered ground support equipment to electric ones. According to some studies, electrifying ground support equipment as well as desulphurizing jet fuel, avoiding the use of APUs and use of single engine taxiing can reduce significantly the airport carbon impact (Yim et al., 2013). However, from a wider point of view the conversion to electric power could simply shift the carbon impact from the airport to another unit. As an example, coal-burning power plants providing electric power produce carbon emissions that cannot be directly ascribed to the airport, but this source of power does not contribute to minimize the environmental effects indirectly due to the airport. Finally, the ground handling equipment characteristics and effi- Total 91,526,947 ciency can be improved by converting the current fleet to hybrid or electric vehicles. At the same time, a direct reduction of in-situ emissions of CO2 might be obtained through the production of Table 12 CO2 emissions due to the identified macro-sources and related UCFs (year 2012). CO2 emission macro-source Value (kg) UCF Car (UCFpax-car) 37,858,060 6.3 kg/pax Energy production/consumption (UCFpax-energy) 6,772,681 1.1 kg/pax Aircraft LTO (UCFLTO-T) 45,089,125 670.4 kg/mov Energy for ground support equipment, APU (UCFmov-handling) 1,807,081 26.9 kg/mov improved transit systems, implementation of policies and actions to reduce the use of private car e such as airport parking facilities for emission friendly vehicles e alternatively fuelled taxis. Discounted fares could be provided to encourage the use of transit systems, while priority parking for emission friendly vehi- cles could encourage passengers to drive this kind of vehicles. Similarly, specific contracts could require the use of environmental friendly taxis as well. On the other hand, market-based policy ap- proaches as environmental taxes or emission charges at local level could be thought as a way to regulate the airport access by cars. To summarize, the TCF methodology help planners and airport operators to compute the carbon contribution of the different relevant sources and then estimate their share so as to promote a sustainable airport development by adopting transport related strategies that minimize the larger contributions. The focus here was on transport related actions because of the primary airport role as transport node allowing passengers to move from one origin to a destination by using aeroplanes. Further improvements can be obtained by best exploring the modelling of some steps, as the relationship among handling ve- hicles, scheduled aircraft and airside configuration as well as the introduction of infrastructure and ground access transport systems characteristics among the relevant variables. Acknowledgements The authors acknowledge the Airport of Bologna for kindly making available the data used in this study and an anonymous reviewer for his constructive and valuable suggestions. References measures concerns again the use of alternatively fuelled vehicles and the improvement of the airport energy efficiency. The first option is completely in Scope 3 and then depends on wider policies aiming at reducing the impact of car traffic flows. However, as the airport represents a local source of emissions due energy from renewable sources (photovoltaic, biomasses power plants, bio-fuels.). Table 13 2013 CO2 forecasted emissions. CO2 emission macro-source Value 2013 (kg) Car (UCFpax-car) 38,964,328 Energy production/consumption (UCFpax-energy) 6,803,295 Aircraft LTO (UCFLTO-T) 43,731,533 Energy for ground support equipment, APU (UCFmov-handling) 1,754,741 Total 91,253,897 Transport Management 37 (2014) 76e86 85 Air Transport Action Group. www.atag.org. Airport Cooperative Research Program (ACRP) Report 11, 2009. Guidebook on Preparing Greenhouse Gas Emissions Inventories. Transportation Research Board, Washington DC. Airports Council International (ACI), 2009. Guidance Manual: Airport Greenhouse Gas Emissions Management. BAA London Gatwick Airport, 2009. Gatwick Airport Climate Change Report. London. Budd, T., Ison, S., Ryley, T., 2011. Airport surface access in the UK: a management perspective. Res. Transp. Bus. Manag. 1 (1), 109e117. Daley, B., Preston, H., 2009. Aviation and climate change: assessment of policy options. In: Gössling, Upham (Eds.), Climate Change and Aviation. Issues, Challenges and Solutions, pp. 347e372. de Neufville, R., Odoni, A., 2003. Directive 2002/49/EC of the European Parliament and the Council of 25 June 2002 relating to the assessment and management of environmental noise. Airport Systems. Planning, Design and Management. McGraw-Hill. Dodson, R.E., Houseman, E.A., Morin, B., Levy, J.I., 2009. An analysis of continuous black carbon concentrations in proximity to an airport and major roadways. Atmos. Environ. 43, 3764e3773. European Commission, 2011. A Roadmap for Moving to a Competitive Low Carbon Economy in 2050. COM(2011) 112 final. Brussels, 8.3.2011. Eurostat, 2011. Energy, Transport and Environment Indicators, 2011 edition. Publi- cations Office of the European Union, Luxembourg. Graham, A., 2003. Managing the Airports: an International Prospective, second ed. Elsevier. Harvey, G., 1987. Airport choice in a multiple airport region. Transp. Res. A 21, 439e 449. Hess, S., 2010. Modelling air travel choice behaviour. In: Postorino, M.N. (Ed.), De- velopments of Regional Airports: Theoretical Analyses and Case Studies. WIT Press, U.K. ICAO, 2010. Environmental Report. http://www.icao.int/environmental-protection/ Pages/EnvReport10.aspx. ICAO, 2011. Airport Air Quality Guidance Manual. Doc 9889. http://www.icao.int/ environmental-protection/Documents/Publications/FINAL.Doc9889.1stEdition. alltext.en.pdf. IEA, 2009. In: Fulton, L. (Ed.), Transport, Energy and CO2. IEA, Paris. Innes, J.D., Doucet, D.H., 1990. Effects of access distance and level of service on airport choice. J. Transp. Eng. 116 (4), 507e516. Intergovernmental Panel of Climate Change (IPCC), 1999. Fourth Assessment Report. http://www.ipcc.ch/ipccreports/sres/aviation. Intergovernmental Panel of Climate Change (IPCC), 2006. Guidelines for National Greenhouse Gas Inventories, IPCC, 2006. http://www.ipcc-nggip.iges.or.jp/ public/2006gl/. ISPRA, 2012. Italian Greenhouse Gas Inventory 1990e2010, National Inventory Report. http://www.isprambiente.gov.it. ISTAT, National Institute for Statistics (2012) Annual Report 2012. Kroes, E., Lierens, A., Kouwenhoven, M., 2005. The airport network and catchment area competition model: a comprehensive airport demand forecasting system using a partially observed database. RAND Europe, Conference Reference: ERSA, 2005. Lu, C., Morrell, P., 2006. Determination and applications of environmental costs at different sized airports e aircraft noise and engine emissions. Transportation 33, 45e61. Miyoshi, C., Mason, K.J., 2013. The damage cost of carbon dioxide emissions pro- duced by passengers on airport surface access: the case of Manchester Airport. J. Transp. Geogr. 28, 137e143. Pels, E., Nijkamp, P., Rietveld, P., 2003. Access to and competition between airports: a case study for the San Francisco Bay Area. Transp. Res. A 37, 71e83. Postorino, M.N., 2010a. Environmental effects of airport nodes: a methodological approach. Int. J. Sustain. Dev. Plan. 5 (2), 192e204. Postorino, M.N., 2010b. Air demand modelling: overview and application to a developing regional airport (Chapter 4). In: Postorino, M.N. (Ed.), Developments of Regional Airports: Theoretical Analyses and Case Studies. WIT Press, U.K. Rosillo-Calle, F., Teelucksingh, S., Thrän, D., Seiffert, M., 2012. The Potential Role of Biofuels in Commercial Air Transport e Biojetfuel. Commissioned by IEA Bio- energy Task 40 Sustainable International Bioenergy Trade. www. bioenergytrade.org. Ryley, T., Jaafar Elmirghani, J., Budd, T., Miyoshi, C., Keith Mason, K., Moxon, R., Ahmed, I., Qazi, B., Zanni, A., 2013. Sustainable development and airport surface access: the role of technological innovation and behavioral change. Sustain- ability 5, 1617e1631. Schürmann, G., Schäfer, K., Jahn, C., Hoffmann, H., Bauerfeind, M., Fleuti, E., Rappenglück, B., 2007. The impact of NOx, CO and VOC emissions on the air quality of Zurich airport. Atmos. Environ. 41, 103e118. Sheffi, Y., 1985. Urban Transportation Networks: Equilibrium Analysis with Math- ematical Programming Methods. Prentice Hall, Englewood Cliffs. Song, S.-K., Shon, Z.-H., 2012. Emissions of greenhouse gases and air pollutants from commercial aircraft at international airports in Korea. Atmos. Environ. 61,148e158. Suzuki, Y., 2007. Modelling and testing the ‘‘two-step’’ decision process of travellers in airport and airline choices. Transp. Res. E 43, 1e20. Train, K., 2003. Discrete Choice Methods with Simulation. Cambridge University Press, Cambridge, MA. Unal, A., Hu, Y., Chang, M.E., Odman, M.T., Russell, A.G., 2005. Airport related emissions and impacts on air quality: application to the Atlanta International Airport. Atmos. Environ. 39, 5787e5798. United Nations, 1998. Kyoto Protocol to the United Nations Framework Convention on Climate Change. Yim, S.H.L., Stettler, M.E.J., Barrett, S.R.H., 2013. Air quality and public health impacts of UK airports. Part II: impacts and policy assessment. Atmos. Environ. 67, 184e192. Yu, K.N., Cheung, Y.P., Cheung, T., Henry, R.C., 2004. Identifying the impact of large urban airports on local air quality by nonparametric regression. Atmos. Environ. 38, 4501e4507. M.N. Postorino, L. Mantecchini / Journal of Air Transport Management 37 (2014) 76e8686 A transport carbon footprint methodology to assess airport carbon emissions 1 Introduction 2 Airport operations and relevant carbon sources 3 The transport carbon footprint methodology 4 Bologna international airport 5 Application of the TCF methodology to Bologna airport test case 6 Summary, discussion and future developments Acknowledgements References