Optimization of Airline Support Facility Space at Non-Major Airports of India Using Goal Programming

Per the commercial land policy and manuals of the Airports Authority of India [20], a portion of airport land and building space is allotted on lease to about nineteen and twenty-nine activities, respectively (see Table 1).

Out of the nineteen possible options for land and twenty nine for floor area, four essential airline supporting facilities and four passenger supporting facilities were considered respectively for the present study.

2.1 Land

The assessment of an airline's decision-making process regarding the continuation and initiation of new operations at an airport is contingent upon the evaluation of both traffic demand and the responsiveness of airport operators in terms of expanding or constructing new ground facilities. It is imperative to scrutinize the allocation of land for various purposes, with leasing options particularly beneficial for four key facilities: fuel stations, ground servicing equipment (GSE), flying schools or hangars catering to commercial general aviation aircraft/non-scheduled operators (NSOP)/MRO, and porta-cabins. The significance of these allocations is rooted in the following reasons.

2.1.1 Fuel stations

Airlines may opt to fuel their aircraft at specific airport locations based on factors such as convenience, logistics, and economics. Encouraging fuel companies to establish fuel stations at every operational airport becomes crucial to ensure a consistent and reliable fuel supply, especially considering uncertainties such as seasonal changes and emergencies. Additionally, the potential financial benefits for fuel companies can extend beyond aviation fuel by serving regular vehicles, given the strategic locations of airports with good highway access. This diversification can contribute to the financial health of fuel companies, ensuring timely rentals for airport operators. Operational airports universally feature at least one fuel station, with allotment areas across non-metro Indian airports ranging from approximately 400 sq. meters to 7400 sq. meters. Figure 1 below shows a typical fuel station photo.

11.png

Figure 1. Fuel station

2.1.2 Ground servicing equipment (GSE)

Various types of GSE, essential for loading and unloading passengers and cargo, towing, auxiliary mobile power, and servicing stationary aircraft, necessitate parking space at airports. While these GSEs are typically parked on the apron's edge, they occupy valuable paved apron space, potentially causing congestion and hindering free movement. Allocating a separate land area, distinct from the main apron, for GSE parking offers a more organized solution with quick access during ground operations. Each GSE unit may require a minimum land area patch of 30 sq. meters, and with the anticipated increase in the number of aircraft and operations, the demand for parking spaces for GSEs is expected to rise significantly. Figure 2 shows a typical ground servicing equipment (GSE) photo.

2.png

Figure 2. Ground service equipment (GSE)

2.1.3 Hangars / flying schools for scheduled airlines, non-scheduled operators (NSOP), and cargo

Scheduled airlines and non-scheduled operators (NSOP), including general aviation/charters, establish their bases at metro or major airports, where they station fleet aircraft for daily operations. These aircraft are often parked at remote bays exposed to the sky. Periodic checks, as per stipulated frequency and minor servicing, necessitate hangar sheds. Some State Government Aviation and private agencies run flying schools/academies from hangars at select airports. However, challenges such as high fees, limited student enrollment, restrictions on flying hours at congested airports, and delays in obtaining approvals have hindered these schools from adhering to their pre-decided training schedules. With Indian airline companies placing record orders for additional aircraft to meet forecasted aviation growth, the need for establishing flying training schools at more airports has become apparent. Reports suggest increased Revenue Market forecasts for Ground Handling Agencies (GHA) - INR 39 billion by 2017, General Aviation - INR 16 billion by 2017, and Maintenance, Repair, and Overhaul (MRO) - INR 70 billion by 2020. However, unforeseen reasons, financial constraints, and airline closures have impeded the realization of these projected potentials. Figure 3 below shows a typical aircraft hangar photo.

2.1.4 Porta-cabins

Scheduled and non-scheduled airlines, along with their supporting ground handling agencies, fuel stations, and other aviation-related entities, often require covered office space not available within airport buildings, airside, or cityside. Even if some space can be made available by the airport, it may not meet their specific location and urgency requirements due to higher rental rates. Factory-built or custom-made Porta-Cabins offer an attractive alternative due to their modular design, speed, and ease of placement or relocation as needed by the agencies or the airport.

It is noteworthy that Juhu Airport in Mumbai, handling about 100 small aircraft/NSOPs, including ONGC's offshore helicopter operations, features hangars and Porta-Cabins for approximately 22 NSOP/helicopter operators. Despite no regular operations of commercial airlines, Juhu Airport remains profitable, relying on lease and rentals from land, hangars, Porta-Cabins, and other sources. Therefore, it is recommended to publicize and allocate land for Porta-Cabins at every non-major airport. Figure 4 below shows a typical Porta-Cabin photo.

3.png

Figure 3. Aircraft hangers/flying schools

4.png

Figure 4. Porta-cabin

2.2 Sources of revenue from space rentals inside the terminal building

The comprehensive breakdown in the right column of Table 1, located under Paragraph 1 in the Introduction section, outlines twenty-nine potential categories of space or floor area allottees/concessionaires within the terminal buildings of Indian airports. Each commercial outlet or concessionaire within this space caters to the diverse needs of passengers during their time in the airport terminal, providing essential products and services. The total floor area of a terminal building is meticulously calculated based on international and domestic norms, prescribed service levels, and the volume of passengers during peak hours of aircraft operations.

Distinct differences emerge between airports handling international operations and those exclusively managing domestic flights, influencing the types, numbers, and locations of commercial outlets. Areas within an airport terminal generally encompass public concourses, departure check-in, secure holds, and arrival sections, tailored to accommodate the flow of departing, arriving, and transiting passengers between the city and air sides.

Table 1. Airport-sources of rental revenue from land lease and covered space (non-aeronautical revenue)

Expression of interest is invited from interested agencies, and open competitive bidding by the airport operator determines the allocation of floor areas to various commercial outlets. These areas are awarded either to the highest individual bidder or a Master Concessionaire. In instances where non-major airports experience minimal and fluctuating aircraft movements, clustering these airports can stimulate interest and responses from potential bidders.

5.png

Figure 5. Commercial outlet

The diverse commercial outlets serving passengers within the terminal building, with floor area rentals at notified rates, can be categorized into four main sub-heads: Retail, Food and Beverages (F&B), Offices, and Cab, Bank, and others. Additionally, advertisements predominantly on vertical surfaces inside and outside the terminal and city-side car parking serve as significant revenue-generating outlets. Contracts for these spaces may extend beyond a year, with monthly revenue comprising rent for floor area or a license fee or royalty—a percentage offered by the allottee to the airport operator based on sales.

To illustrate the allocation process and revenue generation from the earmarked floor area for commercial outlets, an Excel-based input utilizing the Goal Programming concept is presented in the subsequent paragraphs. Figure 5 shows a typical commercial outlet photo.

4.1 Use of goal programming for optimization

Caballero, R., Gómez, T., and Ruiz, F., in their paper titled "Goal programming: realistic targets for the near future," assert that Goal Programming (GP) is one of the most widely used multi-criteria decision-making techniques, specifically designed for real applications [19]. To underscore the advantages of Goal Programming over other optimization techniques, a paper titled "Comparison between goal programming and other linear programming methods" was consulted [21]. Unlike general optimization using linear programming, where the objective function can be fully maximized or minimized under certain constraints to achieve the desired outcome, Goal Programming deals with multi-objective optimization. It represents an extension of linear programming tailored to handle conflicting objectives. The optimization process involves minimizing positive and negative deviations from overall set targets when utilizing airport land and space resources under various constraints, including limited area, facilities with varying revenue returns, and a predefined revenue budget. The formulation of objective and constraint functions for Goal Programming was carried out, and the solution was obtained using the Simplex Program method with Excel Solver software. The selection of the number of non-major airports (sample), inputs, outputs, and a typical sample testing of data for one airport were based on the guidelines and challenges mentioned above. While the detailed test tables and outcomes for the other 22 airports are not included in the main manuscript to conserve space, they are available upon request from the author. The summary outcomes for all twenty-three airports are presented in table format in the subsequent paragraphs.

4.1.1 Model used for land lease mix

Knowing the average minimum areas required for each of the four types of facilities, i.e., fuel station, ground service equipment (GSE), hanger/flying school and port cabin, and the land area that can be spared, the airport operator may want to know the mix of numbers of land areas for allotting to the airline operators, fuel agencies or other related agencies.

a) The Assumed variables and equations for the weighted goal programming model are as under:

1. Fuel Station

$\mathrm{X} 1+\mathrm{d} 1^{-}-\mathrm{d} 1^{+}$        (1)

2. Ground Servicing Equipment (GSE)

$\mathrm{X} 2+\mathrm{d} 2^{-}-\mathrm{d} 2^{+}$           (2)

3. Hanger /Flying School

$\mathrm{X} 3+\mathrm{d}3^{-}-\mathrm{d3}^{+}$           (3)

4. Porta Cabin

$\mathrm{X} 4+\mathrm{d} 4^{-}-\mathrm{d} 4^{+}$           (4)

where, X1, X2, X3, and X4 represent the number of areas of land lease for fuel station, ground service equipment (GSE), hanger, and port cabin, respectively. d1, d2, d3, and d4 represent deviations in the above areas from the target value for the fuel Station, GSE, hanger, and port cabin, respectively. d1^- and d1⁺ represent negative and positive deviations from the target values of the number of fuel station areas. A similar interpretation holds good for the other three variables corresponding to the other three types of land lease areas.

b) Defining the goal constraints:

The total number of land lease areas for four types of facilities:

$\mathrm{X} 1 \mathrm{~A} 1 \mathrm{i}+\mathrm{X} 2 \mathrm{~A} 2 \mathrm{i}+\mathrm{X} 3 \mathrm{~A} 3 \mathrm{i}+\mathrm{X} 4 \mathrm{~A} 4 \mathrm{i}+\mathrm{d} 5 \mathrm{i}^{-}-\mathrm{d} 5 \mathrm{i}^{+}=$Total Available Area at i airport  (5)

A1i represents considered an average area for type 1 (for example, fuel station) facility at the airport i. A similar interpretation holds good for the other three variables corresponding to the three different types of land lease areas. D5i^- and d5i⁺ represent the negative and positive deviation from the assumed land area limit at the airport i.

Total land lease revenue for four types of facilities:

$X 1 L 1 i+X 2 L 2 i+X 3 L 3 i+X 4 L 4 i+d 6 i^{-}-d 6 i^{+}=$Total Lease Revenue at i airport        (6)

L1i represents lease revenue for type 1 (Fuel Station) facility at the airport i. A similar interpretation holds good for the other three variables corresponding to the other.

Three types of land lease areas. D6i^- and d6i⁺ represent the negative and positive deviation from the assumed land area limit at the airport i.

$\mathrm{X} 1$ to $\mathrm{X} 4, \mathrm{~A} 1$ to $\mathrm{A} 4, \mathrm{~L} 1$ to $\mathrm{L} 4, \mathrm{i}$ and $\mathrm{d} 1$ to $\mathrm{d} 6 \geq 0$        (7)

c) The objective function for this weighted goal programming model:

Minimize sum of $\%$ deviations or $\operatorname{Min} \sum 1 / \mathrm{tg}\left[\left(\mathrm{di}^{-}+\mathrm{di}^{+}\right]\right.$        (8) 

where, t g is the target value of goal g.

4.1.2 Model used for commercial outlet numbers mix

Knowing the average areas required for each of the four types of commercial outlets, i.e., retail, food and beverages, offices and cab, etc., and the floor area earmarked. An airport operator may want to know the mix of numbers of outlets for allotting for setting up those four types of outlets by the interested agencies or concessionaires.

a) The assumed variables and equations for the weighted goal programming model are given below:

1. Retail

$\mathrm{X} 1+\mathrm{d} 1^{-}-\mathrm{d} 1^{+}$              (9)

2. Food & Beverages(F&B)

$\mathrm{X} 2+\mathrm{d} 2^{-}-\mathrm{d} 2^{+}$              (10)

3. Offices

$\mathrm{X} 3+\mathrm{d} 3^{-}-\mathrm{d} 3^{+}$              (11)

4. Cab etc.

$\mathrm{X} 4+\mathrm{d} 4^{-}-\mathrm{d} 4^{+}$              (12)

where, X1, X2, X3, and X4 represent the number of areas for retail, F&B, offices, cab, etc., respectively. d1, d2, d3, and d4 represent Deviations in the above areas from Target value retail, F&B, offices, cab, etc., respectively. d1^- and d1⁺ represent negative and positive deviations from the target values of the number of retail areas. A similar interpretation holds good for the other three variables corresponding to the three different types of floor area allotments.

b) Defining the goal constraints:

Total Number of Floor Areas for four types of outlets:

$\mathrm{X} 1 \mathrm{~A} 1 \mathrm{i}+\mathrm{X} 2 \mathrm{~A} 2 \mathrm{i}+\mathrm{X} 3 \mathrm{~A} 3 \mathrm{i}+\mathrm{X} 4 \mathrm{~A} 4 \mathrm{i}+\mathrm{d} 5 \mathrm{i}^{-}-\mathrm{d} 5 \mathrm{i}^{+}=$Total Earmarked Floor Area at i airport     (13)

A1i represents considered an average area for type 1 (for example, Retail) outlet at the airport i. A similar interpretation holds good for the other three variables corresponding to the three different types of floor areas. D5i^- and d5i⁺ represent negative and positive deviations from the assumed and earmarked floor area limit for commercial outlets at the airport i.

Total Floor Rental Revenue for four types of outlets:

$\mathrm{X} 1 \mathrm{~F} 1 \mathrm{i}+\mathrm{X} 2 \mathrm{~F} 2 \mathrm{i}+\mathrm{X} 3 \mathrm{~F} 3 \mathrm{i}+\mathrm{X} 4 \mathrm{~F} 4 \mathrm{i}+\mathrm{d} 6 \mathrm{i}^{-}-\mathrm{d} 6 \mathrm{i}^{+}=$Total Rental Revenue at i airport   (14)

F1i represents floor rental revenue for type 1 (for example, Retail) outlet at the airport i. A similar interpretation holds good for the other three variables corresponding to the other.

Three types of floor areas. D6i^- and d6i⁺ represent the negative and positive deviation from the assumed and earmarked floor area limit for commercial outlets at the airport i.

$\mathrm{X} 1$ to $\mathrm{X} 4, \mathrm{~A} 1$ to $\mathrm{A} 4, \mathrm{~F} 1$ to $\mathrm{F} 4, \mathrm{i}$ and $\mathrm{d} 1$ to $\mathrm{d} 6 \geq 0$     (15)

c) Objective function for this weighted goal programming model:

Minimize sum of $\%$ deviations or $\operatorname{Min} \sum 1 / \mathrm{tg}\left[\mathrm{di}^{-}+\mathrm{di}^{+}\right]$     (16)

where, t g is the target value of goal g.

4.2 Basis, source, and sample for determining the number of non-major airports for study and data collection

The Airports Economic Regulatory Authority (AERA), established in 2009 under the Ministry of Civil Aviation, governs tariffs, fees, and service quality in the aviation sector. Currently, AERA oversees twenty-three major airports, including joint ventures with annual passenger traffic exceeding 1.5 million. Non-major airports, excluding those under joint ventures and entirely operated by the Airports Authority of India (AAI), comprise eighty-three airports.

According to a consultation paper, non-major airports are categorized by passenger traffic as follows:

•   Cluster 1 (C1): More than 1 million passengers per annum (MPPA) - 10 airports

•   Cluster 2 (C2): Between 0.1 and 1 MPPA - 21 airports

•   Cluster 3 (C3): Less than 0.1 MPPA - 52 airports (Total: 10 + 21 + 52 = 83) - Category A

Deductions include:

•   Civil enclaves or defense and joint civil operations - 20

•   Airports with no traffic operations - 18

•   Airports under the regional connectivity scheme - 20 (Total deductions: 20 + 18 + 20 = 58) - Category B

The remaining number of non-major airports after deductions (A - B = 83 - 58) is 25. Out of these, twenty-three airports were selected for the present study, as depicted in Figure 6 below.

6.png

Figure 6. Twenty-three non-major airports of India selected for study (shown in Green)

4.3 Breakup of aeronautical & non-aeronautical revenue of airports authority of India (AAI)

7.png

Figure 7. Break up of non-aero revenue of airports Authority of India during 2017-18 (1 US $=70 INR in 2018)

Typically, over 70% of the land at an airport, encompassing the runway, related pavements, and basic strip area, constitutes the operational area. Structures outside this operational area adhere to permissible height criteria outlined by the obstruction surfaces of ICAO and FAA Manuals. As per the Airports Authority of India [22], excluding joint venture companies, the revenue breakdown (in million INR, with 1 US $ = 70 INR in 2018) is as follows:

•   Airport Services: 425.77

•   Aero-Navigational Services: 349.53

•   Sub-total of Aeronautical Revenue (AR): 775.30

•   Non-Aeronautical Revenue (NAI): 145.10

•   Total Revenue (TI): 920.40

The contribution of aeronautical revenue (Airport Services + Aero-Navigational Services) and non-aeronautical revenue to AAI's total revenue was 84% and 16%, respectively. Further insight into AAI's non-aero revenue is illustrated in Figure 7. Rent and services, along with trading and concessionaires, collectively contribute 44% to the total non-aeronautical revenue (NAR) and 7% to the overall revenue. Refer to Figure 7 for a detailed breakdown of AAI's non-aero revenue components.

4.4 Model inputs

The analysis included compiling details of aircraft and passenger movements, area of land allotment along with its lease revenue and total revenue for twenty-three non-major airports during 2017-2018 and working out a desirable land lease allotment mix at all those airports for setting up the above four types of facilities by the airline & other supporting operators.

Assuming a hypothetical airport area features for working out a mix of four types of commercial outlet floor areas, i.e., retail, food and beverages, offices and cab, bank, and others, along with their rental revenue within the earmarked floor area of airport terminal building and the corresponding possible maximum rental revenue.

Inputting of above data in the form of an Excel spreadsheet using the Weighted Goal Programming Model and obtaining the optimal mix number of four facilities for each of those twenty-three airports. The increased revenue expected from the new land lease mix allotment model is compared with the existing allotted area to know the enhancement in total revenue.

The consolidated findings from the analysis of land lease combinations for four types of facilities (fuel stations, ground service equipment - GSE, hangars, and porta-cabins) using the Goal Programming model [15, 20] and Excel Solver are presented in Tables 5 and 6 as discussed in the preceding section.

1. For each of the twenty-three airports, the land allotment areas ranging from 1830 sq. meters (0.452 acres) to 19136 sq. meters (4.73 acres) would accommodate all four facilities (refer to Table 5) [12].

2. The lease revenue for each airport location, based on notified annual rates per sq. meter, varies from 0.83 million INR to 28.99 million INR (refer to Table 5, with an exchange rate of 1 US $ = 70 INR in 2018).

3. The expected total lease revenue for the land lease areas of all twenty-three airports sums up to 267.88 million INR, covering 216,285 sq. meters (53.45 acres) (refer to Table 5, with an exchange rate of 1 US $ = 70 INR in 2018).

4. The cumulative increase in total land area and revenue, considering the proposed land lease mix for the four facility types, is 95% for land area and 77% for land lease revenue, compared to the existing land allotment revenue at notified rates (refer to Table 6, with an exchange rate of 1 US $ = 70 INR in 2018) [10].

5. Comparing the ratios of land lease revenue to total revenue between the existing allotment and the proposed model, there is approximately a 95% increase in the latter (refer to Table 6, with an exchange rate of 1 US $ = 70 INR in 2018).

6. The analysis incorporates average values for area requirements and average floor rent per sq. meter per month from six airports. Assumptions include a 20% floor area limitation for commercial outlets and corresponding income, forming the basis for the weighted Goal Programming Model.

7. In Table 4 under Para 5, the Solver Model's output of "0" in the Objective Function Cell signifies the successful minimization of deviations based on the selected weightage for each of the six deviations. The resulting total area is 3,840 sq. meters within the earmarked limit of 4,000 sq. meters, generating total rental revenue of 58.56 lakh Indian Rupees or 5.85 million. Further scenarios can be explored by modifying input target numbers and prioritizing types of allotments, even with minor deviations, without complete depreciation (refer to Table 4, with an exchange rate of 1 US $ = 70 INR in 2018).

8. By adopting the presented methodology and results, the optimal mix of land and space allotments for enhancing revenue from both airport land and space resources can be determined within the given limitations [2, 7, 10].

In a chapter titled "Statistical Methods as Optimization Problems," George Mason University's publication highlights various approaches to accommodate multiple objectives and constraints in optimization problems. The simplest method involves forming a weighted sum, where constraints are incorporated as a weighted component of the objective function, allowing control over the extent to which the constraints are met. In optimization, there often requires interaction between decision-makers and the optimization procedure.

While formulating an optimization problem, careful consideration is essential to ensure it accurately captures the objective of the real problem. The impact of assumptions about the real problem can be magnified in optimization, necessitating caution in both formulating and analyzing the underlying problem. Even when the problem is correctly formulated, difficulties may arise in applying the optimization problem in a statistical method. In contrast to regular regression models, which can be validated using statistical parameters/tests such as sample size, R square, p-values, and assessments of multicollinearity, the validation of multi-objective models like Goal Programming involves addressing practical resource allocation problems under constraints while minimizing deviations from the objective function. The author acknowledges the absence of direct references to specific statistical tests for validating Goal Programming models. Furthermore, emphasizing that interpreting and validating a model solely from a statistical viewpoint may not always be feasible or justified. The above process, including the basis for data (pertaining to twenty-three airports), assumptions, etc., was thoroughly explained in the preceding subsections covering Methods and Data Analysis. The outcomes and validation of the model, presented as an optimal mix of land allotments for each of the twenty-three airports, were summarized in Tables 3 and 6 and discussed above.

This paper focuses on sorting, compiling, and modeling data from twenty-three non-major airports out of a total of eighty-three classified by the Ministry of Civil Aviation. Including the remaining airports would broaden the study's scope, allowing for a comprehensive assessment of potential revenue enhancement through land leasing and space renting. The model, utilizing the Weighted Goal Programming approach, calculates the number of units for leasing land to four types of facilities, encompassing both land and building space. For assessing additional revenue from spare land and building space, other frameworks or models, such as dynamic programming, non-linear frameworks, or simulations, could be explored. Future work may involve comparing the results with other multi-objective optimization methods, conducting sensitivity/risk analyses, and considering dynamic traffic and financial factors. Exploring alternate methods will contribute to a more comprehensive understanding of the potential revenue generation at non-major airport.