Case Studies on Transport Policy 9 (2021) 1097–1105

Contents lists available at ScienceDirect

Case Studies on Transport Policy

journal homepage: www.elsevier.com/locate/cstp

Resident air transport subsidy impact on airport ground operations: Gran

Canaria airport case study
Roberto Rendeiro Martín-Cejas
Departamento de Análisis Económico Aplicado, Universidad de Las Palmas de Gran Canaria, Campus Universitario de Tafira, Edificio Departamental de CC.EE. y EE.,
Módulo D., 35017 Las Palmas de Gran Canaria, Spain

A R T I C L E I N F O A B S T R A C T

Keywords: This paper seeks to analyze the impact of regional aviation subsidies for Gran Canary airport; specifically, the
Airline effect on airport capacity, air carrier economics and the environment. It was found that aircraft taxiing opera­
Airport tions time increased producing negative effects on airport capacity management, air carrier economies and the
Subsidies
environment in terms of decreasing airport capacity available and the increased fuel and emissions costs.
Emissions
Taxiing time
However, those losses have to be balanced against the social benefit of increased resident mobility.
Queueing theory

1. Introduction effects of PSOs and subsidies on the efficiency of operators, others have
studied their design in several countries in EU. Those studies found that
Air transport is key to improving the mobility of people, in order to PSOs reduce the level of competition on protected routes and increase
guarantee economic and social cohesion across the European Union the operation cost of European carriers; they also identified weaknesses
(EU). However, some air transport routes in the EU are far from prof­ in the regulation of PSOs in some countries. Fageda et al. (2012)
itable, mainly those located in island regions. In light of this, the Eu­ compared subsidised routes (domestic flights from Gran Canaria) with
ropean Commission has established a public service obligation (PSO) unsubsidised international flights from Gran Canaria and found that
mechanism and subsidy for those routes. The number of PSOs routes has non-residents pay more than international passengers. Valido et al.
expanded in several European countries. France and Norway are the (2014) showed that non-resident passengers could be driven out of the
countries with the greatest number of PSO routes in Europe (Williams market if the flow of resident passengers was high. Santana (2009) found
and Pagliari, 2004). In Spain, PSO routes have been imposed on inter­ that European airlines subject to the PSO mechanism have higher costs.
island transport services since 1998. Together with the PSO, Canary Calzada and Fageda (2014) analysed the effects of PSOs on the level of
Islands residents enjoy subsidies on fares to improve connectivity and competition and flight frequency offered by airlines in the European
accessibility to services on the main islands and mainland Spain (San­ aviation market and found that routes protected by PSOs offer a high
tana, 2009). flight frequency in Spain with respect to unprotected routes with similar
Economic interventions, such as subsidies, aim to lower market characteristics. The relevant finding of this study is that PSOs increase
prices and raise demand (output). Specifically, subsidies in the aviation market concentration.
market can lead to expansion in the aviation system. One of the objec­ Much less attention has been paid to the effects of PSO mechanisms
tives of market intervention can be to improve populations’ air mobility and subsidies on airport infrastructure. Aviation subsidies can lead to
within and from peripheral regions. Benefits from improved mobility the creation of additional airport capacity and encourage regional
can include better job opportunities, easier access to health services and development in peripheral areas (Gössling et al., 2017). However, these
increased leisure travel and tourism for the region, among others. The policies can also produce perverse effects on airport services manage­
EU’s long-term policy is to enhance economic and social cohesion across ment. For instance, delays caused by congestion in airports can cause
the EU. However, the results depend on the way in which the PSO economic losses for airlines and environmental impacts. If an airport is
mechanism and subsidies are adopted with respect to air transport ser­ close to its ‘saturation level’ or has a highly seasonal traffic flow, main
vices (Williams and Pagliari, 2004). operational issues, such as aircraft taxiing for landing and take off, can
According to Fageda et al. (2016), some studies have examined the become more congested.

E-mail address: roberto.rendeiro@ulpgc.es.

https://doi.org/10.1016/j.cstp.2021.05.010
Received 11 April 2020; Received in revised form 7 May 2021; Accepted 20 May 2021
Available online 25 May 2021
2213-624X/© 2021 World Conference on Transport Research Society. Published by Elsevier Ltd. This is an open access article under the CC BY-NC-ND license
(http://creativecommons.org/licenses/by-nc-nd/4.0/).
R. Rendeiro Martín-Cejas Case Studies on Transport Policy 9 (2021) 1097–1105

This paper seeks to close this gap by analysing the impact of aviation distribution of aircraft is a constant distribution. For one server, the
subsidies on Gran Canaria Airport (airport code LPA). Air transport parameters of the Kendall sign would be G/D/1 (Kariya et al. 2011). This
growth from subsidies can exacerbate the need to invest in additional model is shown in Fig. 1.
airport capacity and can produce collateral effects on air carriers’ For this case, the average waiting time for take off Wq in the sta­
economies in terms of increasing their operational costs. Therefore, as tionary state1 is derived by the following equation:
air transport incrementally increases at local levels, its environmental
ρ Cλ2 + Cμ2
impact and economic effects have to be readdressed. In this context, this Wq = × (1)
study contributes to the literature on the impact of subsidies on airport μ(1 − ρ) 2
operations, specifically taxiing operations, at LPA. Delays in taxiing af­ In Eq. (1), Cλ and Cμ are the coefficients of variations of the arrival
fects the economics of air carriers, as well as producing environmental distribution on the runway and the processing distribution, respectively.
impacts in terms of increased greenhouse gas emissions. Queue-based Those coefficients of variation are derived by the division of the stan­
modelling is generally used to analyse aircraft processes at an airport dard deviation and the average value of distribution. According to the
and specifically ground operations congestion (Kariya et al., 2011; Itho diagram processing shown in Fig. 1, the coefficient of the variation of the
and Mitici, 2019). arrival distribution is Cλ = 0 due to the arrival rate at the runway being
The structure of this paper is as follows: Section 2 describes the constantly distributed. Moreover, the coefficient of processing distri­
analytical approach applied in this case study. Section 3 defines the bution (Cμ ) needs to be estimated due to the absence of direct obser­
background of the case study and discusses the results, specifically vation or simulation. This later parameter has been estimated using a
analysing the impact of subsidies on LPA’s taxiing operations by esti­ landing-intervals model for a sample of several approach speeds, for
mating taxiing delay due to congestion in ground operations. Likewise, both aircraft family. The estimated value of Cμ is equal to the ratio
this section calculates the environmental and economic impact of delays among standard deviation and the average value of the sample (Kariya
in ground operations. Finally, the conclusions highlight the main find­ et al., 2011).
ings of the study.
2.2. Landing-intervals model
2. Analytical approach
A simple model of landing intervals was developed to estimate the
Different aircraft types interact at airports and compete for use of average rate of processing aircraft on runways (Harris, 1974). This
airport capacity. Thus, the mix of aircraft is crucial in optimising ca­ methodology gives an approximation of the average processing rate for
pacity and adequately controlling air services (Yu and Lau, 2013). take off using the ‘ultimate capacity concept’ for a mix of aircraft landing
Subsidised regional traffic can exacerbate congestion and may produce on a single airport runway. The landing-intervals model assumes error-
delays at airports. Therefore, to quantify how much the increased traffic free approaches and that pilots are able to precisely maintain the
from subsidies makes airport ground operations delay flights, the con­ required separations and speeds using instrument flight rules (IFR). Two
flicts between aircraft types must be considered. situations were considered: the ‘overtaking case’ (Fig. 2a), in which the
The analytical approach is as follows: first, congestion that causes trailing aircraft has a speed equal to or greater than that of the lead
bottlenecks (in terms of average waiting time) before take off will be aircraft, and the ‘opening case’ (Fig. 2b), in which the speed of the lead
estimated using queuing theory (Kariya et al., 2011; Itho and Mitici, aircraft exceeds that of the trailing aircraft. The following minimum
2019). On the other hand, the conflicts between aircraft will be evalu­ separation function can be applied. In this function aircraft are grouped
ated employing a simple model of landing intervals (Harris, 1974) into n discrete speed classes and a matrix of minimum intervals, so that
considering two aircraft types (i.e., ATRs and the Boeing 737 and Airbus the minimum time separation for each combination of approach speeds
320 (B737/A320) aircraft families). Using these procedures, taxiing can be estimated by the following equations (Ashford and Wright,
time will be estimated. 1992):
( ) δ( )
2.1. Airport ground operations delay model using queuing theory m vj , vi = vj ≥ vi (2)
vj

In this paper, delays in airport ground operations are analysed ( ) δ

(
1 1 (
)
)
through aircraft congestion during taxiing departure operations. m vj , vi = + γ − vj < vi (3)
vj vj vi
Congestion occurs due to the concentration of departures near to the
runway. This congestion phenomenon can be studied using queuing where vi is the speed of aircraft i, γ is the length of common approach
theory to specifically analyse take off waiting time at LPA. To apply the ( )
path, δ is the minimum safety separation between aircraft and m vj , vi is
theory to aircraft taxiing, various parameters for the operational the error-free minimum time separation over the threshold for aircraft j
modelling of aircraft have to be established. The number of aircraft that following aircraft i. The matrix of minimum intervals can be formed for
arrive on the runway at the unit time (arrival rate) is called λ. The aircraft with speed class i following aircraft with speed class j:
number of aircraft that take off at each unit time (processing rate) is [ ]
called μ. The take off density is assumed to be ρ = λ/μ (0 < ρ < 1). The [ ( )] mi,i mi,j
M = m vi , vj = (4)
queue process is described using the Kendall sign, as shown in Table 1. mj,i mj,j
Kendall’s notation in the form of A/S/c means that A describes the
aircraft distribution between each arrival to the runway, S the distri­
bution of processing time and c the number of the servers. In this case, (Arrival) (Processing)
General Constant
the arrival at the runway is a general distribution, and the processing distribution distribution

Table 1 Fig. 1. Kendall sign for take off process: G/D/1.

Parameters of Kendall sign. Source: Kendal (1951).
Parameter Arrival distribution Processing distribution

M Poisson Exponential
D Constant Constant 1
Stationary state is an operational concept in which the processes merely
G General General
reproduce themselves with no changes.

1098
R. Rendeiro Martín-Cejas Case Studies on Transport Policy 9 (2021) 1097–1105

C=1
− (6)
m
This model assumes that runway occupancy time during landing is
less than the time separations during approach and has no effect on
capacity.

2.3. Taxiing time in route

Next, considering the aircraft mix at LPA, the taxiing time in route
has to be estimated for two kind of aircraft: ATR aircraft and the other
narrow-body aircraft, specifically from the B737/A320 families. The
expression used to estimate taxiing time in route for both kinds of
aircraft was as follows:
Distance
Taxiingtime = + (powerback × %powerbackperformance) (7)
taxispeed
Next, this analytical approach (formulas 1 to 7) was used to analyse
the impact of subsidies on congestion at LPA. The environmental and
economic impact of the subsidies will be estimated for LPA for the peak
month of December 2018. First, the analytical approach determines the
delay in taxiing operations; second, the environmental and economic
effects resulting from that delay will be derived in terms of the annual
increase in tons of carbon dioxide emissions and fuel costs, respectively.

3. Case study

3.1. Background

For Canary Islanders, it could be said that their proximity to the

mainland is essential for a better quality of life. The connectivity of the
Canary Islands to large cities in Europe and Africa tempers the
remoteness of the islands and, at the same time, means that they can be
competitive with other European regions. Tourism depends on people’s
mobility and has a strong role to play in the economic development of
the islands, and air transport is essential to support this. Thirty-five per
cent of the Canary Islands’ GDP (gross domestic product) is generated by
national and international tourist flows, which support 40.4% of jobs in
the region (Impactur Canarias, 2018). Therefore, there is a strong
argument for subsidising aviation in the Canary Islands.
In this sense, residents’ mobility, whether for leisure, work or health,
makes the transport sector crucial to guaranteeing proximity, not only to
the mainland of Spain but also between islands. For this reason, air and
sea transport from the islands to the mainland, and between islands, has
been subsidised since 1982 under a compensation scheme to reduce the
extra costs incurred by freight and passenger traffic as a result of the
islands’ remoteness from the Spanish mainland and EU territories.2
Recently, this allowance has been increased from 50% to 75% of the
travel price. This new subsidy began in early June 2018, and within the
following six months, interisland air traffic volume increased by 19.8%,
compared to the same period in 2017. In contrast, traffic from the Ca­
nary Islands to mainland Spain increased in the same period by only
Fig. 2. Landing interval model ‘overtaking case’ (a) and opening case’ (b).
around 2.5%. In percentage terms, subsidies for air traffic stimulated
more interisland air traffic than that from the Canary Islands to the
This matrix associates each one of the n speed aircraft classes with a
Iberian Peninsula (Table 1 - Aeropuertos Españoles y Navegación Aérea-
probability of occurrence [P1 , ⋯, Pn ]. These probabilities are the per­
AENA, 2019). This increase in interisland traffic has been mainly for
centages of the various speed classes in the aircraft mix divided by 100.
Avions de Transport Régional (ATR) turboprop aircraft flow, and because
Thus, the expected minimum landing interval or weighted mean service
of the special operational characteristics of these kinds of aircraft, it
time can be approximated by the following formula:
could affect operational issues for any airport on the islands
∑
−
m= Pi mij Pj (5) One of the main characteristics of air traffic in the Canary Islands is
ij the importance of regional or interisland air traffic. The preponderance
of turboprop airplanes (ATR families) operating at Canary Islands air­
Parameter mis approximate to the processing rate of aircraft before
−
ports may have implications for airport capacity management. Table 2
take off. Finally, the hourly saturation capacity is the inverse of the
weighted mean service time:
2
Real Decreto 1316/2001

1099
R. Rendeiro Martín-Cejas Case Studies on Transport Policy 9 (2021) 1097–1105

Table 2 Table 3
Traffic evolution (ATMs) for LPA airport. Source: Compiled by authors with data Main PSO route characteristics (December 2018). Source: Compiled by author
from AENA. with data from the website of Binter and AENA.
Aircraft type/ 2016 2017 2018 2019 PSO routes FD1 ATM2 Pax3 LF4 SE5
year
Gran Canaria (LPA) -Tenerife Norte 20 1152 79,319 68.8 €40
Turboprop1 34,898 37,068 48,175 50,746 (TFN)
(9.9%)* (6.2%) (30%) (5.3%) Gran Canaria (LPA) -Tenerife Sur (TFS) 4 224 11,630 52 €42
Narrow-body2 77,102 (14%) 81,483 82,852 75,706 Gran Canaria (LPA) – Lanzarote (ACE) 14 776 64,236 82.8 €46
(5.7%) (1.7%) (-8.6%) Gran Canaria (LPA) – Fuerteventura 12 680 51,314 75.4 €50
Total traffic 112,000 118,551 131,027 126,452 (FUE)
1
*In brackets percentage increase of the ATM (Air transport movement) by Average daily frequency.
2
aircraft type. Monthly air traffic movements of ATR aircraft.
1 3
ATRs 72 and 42. Monthly resident passenger number.
2 4
Boeing 737 and Airbus 320 families. Average load factor.
5
Average fare subsidies: (ticket price for non-resident – ticket price for
shows the traffic evolution for LPA over the last four years for different resident).
aircraft types, mainly ATRs 42 and 70, and narrow-body airplanes such
as the B737/A320 families. It is apparent from Table 2 that ATR traffic found also that the benefit of the price discount is transferred to the
has steadily increased. The subsidy increase became available in June airlines to the detriment of both island residents and other passengers on
2018, and in the following six months, the percentage of ATR aircraft these routes. Hence, control of the flight price becomes necessary to
traffic increased by 30% in contrast to the same period in the previous guarantee PSO service at the ‘right price’.
year. This means that regional aviation absorbed most of the subsidy Binter Canarias also connects the Canary Islands with the Iberian
increase. As is apparent from Table 2, 2018 is the best moment to esti­ Peninsula (Murcia, Palma de Mallorca, Pamplona, Vigo, Zaragoza,
mate the impact of the subsidy increase in air transport movement at Santander and Victoria) and a few years ago started its international
LPA. expansion to Africa, flying to destinations such as Agadir, Casablanca, El
The Canary Islands have eight airports, including two on Tenerife. Aaiún, Marrakech, Banjul, Punta Delgada, Dakar, Nouakchott and
Those airports can be classified into two groups: those with heavy in­ Dakhla. The airline also runs flights to Lisbon and Madeira. Similarly, air
ternational traffic (Gran Canaria, Tenerife Sur, Fuerteventura and Lan­ traffic from the Iberian Peninsula to the Canary Islands has steadily
zarote) and the other four (Tenerife Norte, La Gomera, El Hierro and La increased in recent years, as it has also been influenced by an increase in
Palma) servicing mainly national and local traffic. The Spanish Civil fare subsidies. However, this air market has strong competition
Aviation authority ranks the Canary Islands as the third Spanish region regarding price from low-cost airlines such as Air Europa, Ryanair,
in terms of passenger traffic in 2018: its airports accommodated 45.3 Vueling and Norwegian. In fact, these four airlines account for 93.5% of
million passengers, which represents 17% of total passenger traffic in the market (Gundelfinger-Casar and Coto-Millán, 2018).
the Spanish airport network. (The top two regions are Madrid (22%) and The subsidies that the Spanish government gives to regional air
Catalonia (20%).) The AENA top 10 ranking of greatest passenger vol­ transport carriers flying between islands, and from the Canary Islands to
ume in 2018 included the airports of Lanzarote (7.3 million), Tenerife the Iberian Peninsula, are considered cash funding. This is in essence a
Sur (11 million) and Gran Canaria (13.6 million) (AENA, 2019). transfer of money from the government to the transport sector, which
In 2018 Canary Islands airports received more than 18 million non- uses it as revenue. This market intervention, as is apparent from Table 2,
interisland passengers (foreigners and from the rest of Spain). Nearly 14 incentivises air traffic and therefore impacts airport capacity
millions of these inflows were from Europe. However, the interisland management.
passenger volume for the same year was about 4.4 million. This repre­ LPA is open during the whole year and 24 h per day, which repre­
sents an increase of at least 40.6% from the previous year (AENA, 2019). sents 8,760 h per year of available capacity. The saturation level is about
The entrance into the interisland air market of two new air operators, 53 ATMs/peak-hour (Plan Director del Aeropuerto de Gran Canaria,
Canaryfly, at the end of 2012, and Air Europa, at the end of 2017—the 2005). However, it is a seasonal airport and has peak periods during
latter attracted by the subsidy increase—is one of the factors that ex­ which demand is very close to the saturation level. Currently, LPA is far
plains this continued growth in interisland traffic. However, Canaryfly from being a congested infrastructure. However, this does not mean that
has been absorbed by Binter (although it maintained its brand image) taxiing operations are not congested at peak periods, resulting in air
and Air Europa has stopped operating in the Canary Islands’ interisland carriers wasting time, increasing fuel consumption and greenhouse gas
air market. Hence, Binter exerts real monopoly power in the Canary (GHG) emissions.
Islands’ interisland air traffic market. Calzada and Fageda (2014) found LPA has two parallel runways (designated 03L/21R and 03R/21L);
empirical evidence of market concentration in routes protected by PSOs however it does not allow simultaneous take off and landings. The
in European countries. All Interisland routes in the Canary Island have airport uses only one runway (03L/21R), since it is adjacent to the ter­
been operated under public service obligations (PSOs) since this type of minal area and is available the whole time. LPA has a competitive use
subsidy started in 1998 (Santana, 2009). All PSO routes are operated by between short-range aircraft and regional turboprops. This distribution
one operator, Binter Canarias. Table 3 shows some characteristics of is understandable because of the environment in which the airport is
PSOs routes. placed, that is, on an archipelago where aircraft are of great importance
As is apparent from Table 3, the main PSO routes have an average in the transportation of people and cargo. This is one of the main
load factor of about 70%. The route with most traffic density has an operational characteristics of LPA that must be considered. In 2018
average daily frequency of 20 flights and is the route that generates most there were 48,175 ATM (turboprop aircraft) flights, mainly from the
subsidies. The average subsidy per passenger for those PSO routes was ATR-72 and ATR-42 families. In contrast, short-range aircraft, mainly
€44. Gössling et al. (2017) point out that due to the lack of a competitive B737s and A320s, accounted for 82,852 ATMs (AENA, 2018). In terms of
tender, the subsidy might not be set at the most efficient level. In percentage, 36.8% of all aircraft movements at LPA were ATR aircraft.
addition, Santana (2009) found evidence that airlines employing the This study focuses on the operations of B737s, A320s and ATRs
PSO programme have higher costs. Consequently, airlines operating the because together they represent 89.4% of the aircraft types operating at
PSO service have incentives to establish higher airfares than those LPA (Lorenzo-Aparicio and Rendeiro Martín-Cejas, 2017). It is apparent
specified by the administering authority. Calzada and Fageda (2012) from Table 1 that ATR traffic has increased steadily. Turboprop aircraft

1100
R. Rendeiro Martín-Cejas Case Studies on Transport Policy 9 (2021) 1097–1105

have some characteristics, such as slow-speed landing, that might affect (Doc 9830-Advanced Surface Guidance and Control System, 2004). The
operations of other aircraft families at LPA and cause congestion. Delays estimation of taxiing time was performed by taking as a reference the
mainly occur during the taxi-out phase due to departure congestion and distances to the farthest parking stand from the runway for the two types
due to interference among aircraft flying the final approach path and of aircraft. The ATR parking stands are distant from the terminal
landing (Ignaccolo, 2003). building; the farthest stand (P00) (see Appendix B) is about 1,200 m
from the runway (see the green line in Fig. 2). For the B737/A320 family
of aircraft, the farthest finger (T01) is 2,375 m from the runway (see the
3.2. Estimation of the subsidy impact on LPA ground operations
blue line in Fig. 3). In these calculations, the farthest parking stand was
selected considering the worst scenario for compensating an aircraft
An analytical approach was implemented for a peak month
common route without conflicts effects which were not considered.
(December), which was established by inspecting the flow data at LPA
The configuration of taxiing at LPA implies the existence of a com­
for the last three years. Thus, the model was implemented for the peak
mon route where a conflicts between aircraft might occur. However, if a
period of 2017 and 2018, that is, before and after the subsidy increase.
natural segregate subsystem to separate the taxiing of ATRs from B737/
To estimate Equation 1 of the analytical approach, the parameters (Cμ ),
A320 families is applied (see Fig. 3), potential interference between
λ, μ and ρ have to be calculated. The coefficient of variation of arrival
taxiing aircraft is removed. Hence, to estimate taxiing time in route for
distribution is Cλ = 0, due to the arrival rate at the runway being
both aircraft families, the ‘no conflict scenario is considered. In this case,
constantly distributed. To do that, the landing interval model was
the taxiing time in route only depends on the distance from the parking
applied next.
and the aircraft’s taxiing speed. Therefore, the taxiing time in route
According to the wake vortex categorization, the aircrafts’ landing
without aircraft interference and for the dates already established, for
and take offs at LPA airport belongs to category C (lower heavy) in
both years of the study, would be as follows:
RECAT-EU wake turbulence separation based on ICAO Doc 4444 PANS-
Taxi time in route ATR = (Distance / taxi Speed) + (Power-back Time
ATM. In this sense, the minimum separation to guarantee safety sepa­
× 0.3) = 69.57 seg.
ration (δ) would be of 3 nautical miles (Rooseleer et al, 2016). The
Taxi time in route 737s/A320s = (Distance / taxi Speed) + (Power-back
matrix of minimum intervals for the airports was estimated using 6
Time × 0.5) = 221.66 seg.
nautical miles as the length of common approach path (γ). The approach
The cumulative effect of taxiing time in route, the time spent waiting
speed for ATRs was in the interval of 85–130 knots (approach speed
for take off, the increase in fuel costs for air carriers and CO2 emissions
units) and 115–160 knots for B737/A320 families. The probabilities of
are shown in Tables 5 and 6. As shown in Table 6, an increase in time
occurrence for ATRs and B737/A320 families were 0.368 (36.8%) and
spent waiting for take off represents significant losses for both air car­
0.632 (63.2%), respectively. Using the above dates, the complete matrix
riers and the airport. These are primarily associated with the increase in
M (minimum time separation between aircraft for each combination of
fuel consumption and unwanted CO2 emissions. According to data
approach speeds) and the ultimate capacity for LPA are as follows:
published by Avions de Transport Régional (2000), the fuel consump­
[ ]
127 193 tion for taxiing, for ATRs aircraft families, was approximately 6 kg/min.
M=
28 94 For the B737/A320 aircraft families, the fuel burned in taxiing was, on
average, 13.6 Kg/min (Lorenzo-Aparicio and Rendeiro Martín-Cejas,
2017). The relationship of 3.15 kg of CO2 per kg of fuel burnt allows us
to estimate the increase in CO2 emissions. Additionally, the evolution of
1 1
CLPA = − = =18.5arrivals/h the fuel price published in IATA (2019), for December of 2017 and 2018,
m (127×0.368+28×0.632+193×0.368+94×0.632)
was 1.86 €/kg and 1.81 €/kg,3 respectively. Using the values estimated
The results of the landing interval and queue models for year 2018 previously (see table 4), it is therefore possible to compare the values for
are summarised in Table 4. waiting time, taxiing time, fuel cost and CO2 emissions at LPA for the
Before starting the estimation of taxiing time in route, certain aspects periods before and after the subsidy increase. These results are shown in
must be clarified. ATR aircraft very often have a ‘power-back unas­ Table 5. Table 6 shows the annual losses due to the subsidy increase.
sisted’, in contrast to the B737/A320 aircraft families. It will be assumed This is an overestimation because it considers a peak month (December)
that ATRs take 31.89 s on each power-back and that on 30% of occasions as a representative month for both years (2017 and 2018). The annual
ATR aircraft perform power-backs before taxiing, while B737/A320 estimation uses the ATM flow from Table 1. However, this could be
aircraft families take 205.83 s and perform power-back on 50% of oc­ compensated, because the conflicts in taxiing route operations between
casions (Lorenzo-Aparicio and Rendeiro Martín-Cejas, 2017). With aircraft types was not considered, and thus the waiting time before
respect to other ground operations on taxiing, both ATRs and B737s/ taking off may have been greater.
A320s are restricted to a maximum speed of 20 m/s. This is a reasonable
restrictions according to ICAO standards that suggest a speed of 25.8 m/ 3.3. Discussion
s (50 knots) for aircrafts on straight taxiways such as those at LPA airport
Subsidising PSO air routes in the Canary Islands has shown signifi­
Table 4 cant perverse effects in terms of increased fuel consumption and emis­
Results of landing interval and queue models for year 2018. sions. As is apparent from Tables 5 and 6, the potential economic and
environmental loss for Canary Islander due to the implementation of the
Parameters Waiting time
(s.) subsidy increase is clear. It could have also a negative effect on the use of
the main runway (03L/21R) at LPA (due to space constraints, though
λ = 16 ATM/hours = 0.267 ATM/min. Wq = 219 s.
μ = (ultimate capacity/60 min.) = 18.5/60 = 0.308 arrival/
this is not quantified in this study). LPA airport has two parallel runways
min. (03L/21R, 03R/21L) however only one is currently in use 95.7% of the
ρ = λ/μ = 0.267/0.308 = 0.8668 (take off density) time, due to annual weather conditions. Further, these two runways do
Cλ = 0Cμ = 0.5876* not allow for mixed and simultaneous operations of take offs and
*Coefficient of variation of the arrival distribution estimated for different
combination of approach speeds (low, medium and high) as approximation to
3
According to the average fuel price for December 2017 and 2018, using a
the rate of aircraft before takeoff: 1− . (while aircraft is landing takeoff is pre­
m conversion factor of $1 = 0.87€ for 2017 and $1 = 0.84€ for 2018 and a
vented).
kerosene density of 817 kg/m3.

1101
R. Rendeiro Martín-Cejas Case Studies on Transport Policy 9 (2021) 1097–1105

Fig. 3. Segregate subsystem for taxiing.

dependence on aircraft types (traffic homogeneity). The most significant

Table 5
part of the total delay occurs during the taxiing phase, and the service
Subsidy impact per ATM and aircraft type at LPA airport.
time is strongly correlated with the degree of the homogeneity of the
Year Aircraft Waiting Taxiing Fuel Fuel cost CO2 traffic mix (Ignaccolo, 2003).
type time3 (s.) time4 (s.) (kg) (€) (kg)
In addition, from Table 3 and for the PSO route between LPA and
2017 ATRs 168 69.57 23.76 44.2 74.84 Tenerife Norte Airport (TFN), the subsidy for December 2018 was about
B737s/ 168 221.66 88.32 164.27 278.21 €3,172,760. This revenue from the subsidy could create an incentive to
A320s
2018 ATRs 219 69.57 28.85 52.23 90.89
develop new routes from the islands to mainland Spain or to increase the
B737s/ 219 221.66 99.88 180.78 314.62 frequency of existing ones. The interisland routes are operated by Binter
A320s Canarias S.L. under the Public Service Obligations (PSO) regime (San­
3
For each year ‘waiting time before takeoff’ is considered the same for both
tana, 2009). Currently, there are two air operators on the Canary Islands,
aircraft types. Binter and Canaryfly; however, the latter has been a Binter subsidiary
4
Taxiing time does not vary from one year to the next because of the segre­ since 2017. In 2019, three new routes to the Iberian Peninsula were
gated subsystem for taxiing (see Fig. 2). created, to Murcia, Pamplona and Zaragoza. In the current year Binter
was planning to introduce more peninsula destinations. However, the
emergence of COVID-19 could halt Binter’s expansion strategy, which
Table 6 for year 2020 perhaps has to be postponed depending on the evolution of
Annual losses of subsidy increase per aircraft type for taxiing operations. the pandemic and the response to it.
Aircraft type Annual fuel cost increase (€) Annual increase in CO2 (tons) In summary, two main effects from the subsidies can be pointed out:
ATRs* 877,775 1,604.9
a more intensive use of LPA’s capacity and an incentive to develop new
B737s/A320s 1,592,772 3,397.51 routes from the Canary Islands to the Iberian Peninsula or/and to in­
Total 2,470,547 5,002.4 crease the frequency of existing interisland routes. Two main conse­
*Annual increase in CO2 for ATRs: (48,175 × 90.89) – (37,068 × 74.84) = quences should be considered: first, the need to invest in additional
1,604.9 tons. CO2. airport capacity or to improve it to accommodate an increase in traffic,
and, second, the improvement of Canary Islanders’ mobility. For the
landings because the separation is insufficient (210 m) to guarantee first, a smart solution using LPA’s design features could be to separate
being out of wake vortex influence (Lorenzo-Aparicio and Rendeiro the operational flow according to aircraft type, that is, creating a
Martín-Cejas, 2017). Regarding runway capacity and considering the turboprop-regional subsystem (airport-within-airport), minimising
two aircraft types, it is apparent from Table 5 that the runway occupancy conflicts between aircraft types in the taxiing phase (Lorenzo-Aparicio
time (waiting plus taxiing time) for 2017–2018 has increased by about and Rendeiro Martín-Cejas, 2017). For the second, cost-benefit analyses
16.2% of its available time. would be required if the social and economic value of the subsidy has to
There are many other factors that can potentially affect service time be estimated.
in the taxiing phase, such us runway configuration, limited capacity due
to weather conditions or air traffic control (ATC). The study has 4. Conclusion
demonstrated that those effects are linked to the competitive use be­
tween short-range aircraft and regional turboprops at LPA. As pointed The subsidy applied to travel between the islands, and from the
out above, this is one of the main operational singularities of LPA. As is Canary Islands to mainland Spain, now represents about 75% of the
apparent from Table 2, the percentage increase of the ATM by turboprop market price. As mentioned above, the subsidy increased from 50% to
aircraft (ATRs) in 2018 was higher than the narrow-body families, 75% in June 2018. As a consequence, traffic flow at LPA increased and
because of the subsidy effect. ground operations became more busy due the conflicts between
The optimisation of airport operations is one of the main ways to turboprop-regional aircraft and narrow-body aircraft. Thus, congestion
reduce congestion, increase capacity and decrease fuel consumption and may appear to be a perverse effect, at LPA. An analytical approach was
emissions. This problem can be approached in specific ways, by developed to estimate the subsidy’s impact on LPA ground operations.
considering the characteristics of airport operations. In this sense, one of The study showed that the implementation of a subsidy increase for
the main features of an airport using queuing theory is its strong Canary Islands residents has produced substantial economic and envi­
ronmental losses. From the perspective of the airlines, this subsidy

1102
R. Rendeiro Martín-Cejas Case Studies on Transport Policy 9 (2021) 1097–1105

causes an annual fuel cost increase of about 2.47 million euros and an increase produces for Canary Islands inhabitants. Also, further work
increase in CO2 emissions of about 5,002 tons. In addition, it has to be needs to quantify the reduction in LPA’s available runway capacity at
pointed out that this subsidy produces an increase in ground operations peak periods. This would allow us to determine how this subsidy in­
time (waiting plus taxiing time) and, therefore, an increase in 2018 of crease affects the quality of airport service. Finally, a more technical
the runway occupancy time as it is apparent from table 5. These losses, analysis might explore the implementation of similar procedures while
however, have to be balanced with the social and economic benefits that considering conflicts between aircraft in taxiing route operations.
each regional inhabitant derives from the subsidy in terms of improve­
ment in mobility for any purpose. Nevertheless, as I have pointed out Acknowledgements
above, these losses are underestimated, because the conflicts in taxiing
route operations between aircraft were not considered. This research did not receive any specific grant from funding
A natural extension of this work would be to implement a cost- agencies in the public, commercial or not-for-profit sectors.
benefit analysis of the mobility improvement that this subsidy

Appendix. Gran Canaria Airport Plan

1103
R. Rendeiro Martín-Cejas Case Studies on Transport Policy 9 (2021) 1097–1105

References Harris, R., 1974. Models for runway capacity analysis. Report N◦ FAA-EM-73-5. McLean,
VA.: Mitre Corporation.
ICAO, 2004. Doc 9830-Advanced Surface Guidance and Control System.
AENA Air traffic statistics http://www.aena.es/csee/Satelite?pagename=Estadisticas/
Ignaccolo, M., 2003. A simulation model for airport capacity and delay analysis.
Home 2018 Accessed 2/4/20.
Transportation Planning and Technology 26 (2), 135–170.
AENA Air Traffic Statistics http://www.aena.es/csee/Satelite?pagename=Estadisticas/
Impactur, Estudio del impacto económico del turismo sobre la economía y el empleo de
Home 2019 Accessed 2/4/20.
las Islas Canaria 2018 Exceltur.
Ashford, N., Wright, P., 1992. Airport Engineering. A Wiley-Interscience Publication.
Itho, E., Mitici, M., 2019. Queue-based modelling of the aircraft arrival process at a
John Wiley & Sons INC.
single airport. Aeroespace 6, 103. https://doi.org/10.3390/aerospace6100103.
Avions de Transport Régional, 2000. ATR 72-500 Performance Data.
Y. Kariya T. Mase S. Yoshihara J. Ota Analysis of congestion of taxiing aircraft at a large
Calzada, J., Fageda, X., 2014. Competition and public service obligations in European
airport 2011 Phuket, Thailand.
aviation markets. Transportation Research Part A 70, 104–116. https://doi.org/
F. Lorenzo-Aparicio R. Rendeiro Martín-Cejas The economic and environmental benefits
10.1016/j.tra.2014.10.009.
of runway capacity optimization: The case of Gran Canaria airport Journal of Airport
Calzada, J., Fageda, X., 2012. Discounts and Public Service Obligations in the Airline
Management 11, 4 2017 420–435 (16).
Market: Lessons from Spain. Review of Industrial Organization 40 (4), 291–312.
Plan Director del Aeropuerto de Gran Canaria, 2005. Estudio de la situación actual del
https://doi.org/10.1007/s11151-011-9331-7.
Aeropuerto de Gran Canaria. Ministerio de Fomento, Dirección General de Aviación
Fageda, X., Jiménez, J.L., Díaz, C., 2012. Fare differences between domestic and
Civil.
international air market on routes from Gran Canaria. Journal of Air Transport
Rooseleer, F., Treve, V., De Visscher, I., Graham, R., 2016. Wake Turbulence Re-
Management 25, 8–10. https://doi.org/10.1016/j.jairtraman.2008.12.001.
Categorisation on approach and Departure for Safe and More Efficient Air Traffic
Fageda, X., Jiménez, J.L., Valido, J., 2016. Does an increase in subsidies lead to changes
Management, European organization for the Safety of Air Navigation (Eurocontrol).
in air fares? Empirical evidence from Spain. Transportation Research Part A 94,
30th Congress of The International Council of the Aeronautical Sciences.
235–242. https://doi.org/10.1016/j.tra.2014.10.009.
Santana, Isabel, 2009. Do Public Service Obligations hamper the cost competitiveness of
Gössling, S., Fichert, F., Forsyth, P., 2017. Subsidies in Aviation. Sustainability 9 (8),
regional airlines? Journal of Air Transport Management 15 (6), 344–349. https://
1295. https://doi.org/10.3390/su9081295.
doi.org/10.1016/j.jairtraman.2008.12.001.
Gundelfinger-Casar, J., Coto-Millán, P., 2018. Measuring the main determinants of
tourism flows to the Canary Island from mainland Spain. Journal of Air Transport
Management 70, 83–90. https://doi.org/10.1016/j.jairtraman.2008.12.001.

1104
R. Rendeiro Martín-Cejas Case Studies on Transport Policy 9 (2021) 1097–1105

Valido, J., Socorro, M.P., Hernández, A., Betancor, O., 2014. Air transport subsidies for Glossary
resident passenger when carriers have market power. Transportation Research part E
70, 388–399.
Yu, C., Lau, H., 2013. Integrated Optimization of Airport Taxiway and Runway AENA: Aeropuertos Españoles y Navegación Aérea
Scheduling. J. Autom. Control Eng. 2(4), 338-342. ATM: air transport movement
Williams, George, Pagliari, Romano, 2004. A comparative analysis of the application and ATR: Avions de Transport Régional
use of public service obligations in air transport within the EU. Transport Policy 11 GDP: gross domestic product
(1), 55–66. https://doi.org/10.1016/S0967-070X(03)00040-4. GHG: greenhouse gas
Knots: approach speed units
LPA: Las Palmas Airport (Gran Canaria Airport)
PSO: public service obligation

1105