Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.
Forecasts are informed predictions of future aviation activity that are supported by careful assessment and analysis of historical trends in traffic demand, projected economic growth, and any other relevant factors that may affect growth in the local aviation market. The more robust the analysis, the more reliable the predictions will turn out to be, particularly for the short term. For the medium to longer term, forecasts provide an important guide to airport planners as to when additional airport infrastructure may be required, but they must have regular review and adjustments, as necessary, to reflect any unexpected changes in market conditions. Forecasts of future levels of aviation activity serve as the basis for effective decisions about the terminal plan. Planners use forecasts to determine the need for new or expanded facilities. In general, forecasts should be based on the latest available data and provide justification for the proposed development. The level of effort required to produce a planning forecast will vary significantly from airport to airport. Considerable effort, including the use of elaborate forecasting tools and techniques, may be warranted in the case of more complex projects. A more cursory update of an existing forecast, on the other hand, may be all that is required for simpler projects. Planners should determine the appropriate level of forecasting effort in the course of pre-planning and scoping the study. A number of forecasts are readily available for use in developing and evaluating the master plan forecast. These include the FAA Terminal Area Forecast (TAF), state aviation system plans, and other planning efforts. Most of these forecasts lack the detail about peak period activity required for terminal planning, and thus, at a minimum, the terminal planner must extend the forecast to include this needed detail. âPlanners preparing forecasts of demand or updating existing forecasts should consider socio- economic data, demographics, disposable income, geographic attributes, and external factors such as fuel costs and local attitudes toward aviationâ (13). FAA AC 150/5070-6B provides a good general overview of the forecasting process and the information that a planner should consider in preparing a forecast: ⢠Economic characteristics: The economic characteristics of a community will affect the demand for air traffic. In addition to national and regional economic activity, these characteristics include specific, identifiable, local activities that distinguish the geographic area served by the airport. The type of industry in an airportâs service area also will affect aviation demand, with manufacturing and service industries tending to generate more aviation activity than resource industries such as mining. ⢠Demographic characteristics: The demographic characteristics of an areaâs population also affect the demand for aviation services. Demographic characteristics influence the level, 80 C H A P T E R I V Forecasts
composition, and growth of both local and long-distance traffic. Factors such as average disposable income, leisure time, and recreational pursuits are also important in estimating activity, but can be difficult to measure. These factors are a good indicator of the propensity to travel and may also be significant for demand of general/business aviation aircraft purchases and use. ⢠Geographic attributes: The geographic distances between populations and centers of commerce within the airportâs service area may have a direct bearing on the type and level of transportation demand. The existence of populations and centers of commerce beyond an airportâs service area may indicate the need for additional airports that serve transportation demand. The physical characteristics of the area and the local climate may also be important, because they may stimulate holiday traffic and tourism. The role of the airport within the airport system and its relationship to other airports may also have an effect on the services that are demanded at the airport as well as the availability of alternative modes, particularly high-speed rail. ⢠Aviation-related factors: Business activity, changes in the aviation industry, and local aviation actions can markedly affect the demand for airport services. Business developments in the airline industry, such as consolidations, mergers, and new marketing agreements, can affect airline operations at a particular airport. Wider industry trends, such as the introduction of low- fare service, new classes of aircraft, and the growth or curtailment of an airline hub may also alter the level and pattern of demand. Actions taken by local airport authorities, such as changes in user charges, ground access policies, or support services, can also stimulate or hinder the demand for airport services. Investment decisions made as a result of the planning process itself can also produce change by removing physical constraints to airport growth, which should be reflected in the forecasts. ⢠Other factors: External factors that may also influence the demand for airport services include fluctuations in fuel price and the availability of aviation fuels, currency restrictions, and changes in the level and type of aviation taxes. Political developments, including rising international tensions, changes in the regulatory environment, and shifting attitudes toward the environ- mental impacts of aviation, may also affect future demand and should be considered in devel- oping or updating airport forecasts. IV.1 Methodology Long-term forecasts of annual traffic demand are usually created using one of three techniques: by computing an airportâs share of a larger system forecast, by extrapolating past activity trends into the future, or by relating the forecast of future activity to other forecasts of socioeconomic factors through regression analysis. A brief description of these techniques is described in the following sections. The technique of choice will depend on the resources available to the forecaster and the complexity of analysis required, but, either way, the forecaster will be expected to support his predictions on a firm foundation of facts, historical trends, and socioeconomic analysis. Deployment of more than one methodology and/or multiple analyses will strengthen confidence in the recommended outcome of the forecast study. IV.1.1 Share Analysis An airportâs traffic can usually be assessed as being a share of a larger traffic volume, such as that of a statewide or regional aviation system. A relatively simple forecasting technique is to examine the percentage relationship of the airport in question to the larger system. One can determine whether this relationship grows or shrinks over time, and then extend this trend into the future. This technique is useful when an independently prepared forecast for the larger area is available. Forecasts 81
Share-based forecasts can be prepared with relatively minimal effort and can be a useful tool to benchmark forecasts prepared by other methods. They work best in stable aviation markets when any changes in airline service strategies and economic trends are expected to be few or rel- atively uniform across the larger forecast area. They perform less well in more dynamic aviation markets or at times of market or economic uncertainty. Accordingly, the forecaster needs to review carefully the overall characteristics of the aviation market concerned in order to determine whether share analysis is an appropriate tool. IV.1.2 Trend Analysis One method of creating a forecast is to examine historical data to determine the trend in activity and then extend this trend into the future. Generally, at least 20 years or more of data should be examined so that periods of both economic growth and recession are captured. The forecaster can then determine whether recent performance is above or below the longer term trend line and begin to analyze the reasons why. Trend analysis is adequate when no changes to the status quo of airline service and economic activity are expected. Trends, however, tend to be upset when airlines make a radical change to their service models, declare bankruptcy, or enter/leave an airport market. Trends also tend to be upset by unexpected changes to the economic health of a region. Trend analyses tend to work best for short-range forecasts because the risks of these changes occurring are less. IV.1.3 Regression Analysis and Econometric Modeling Regression analysis derives forecasts of passenger and/or cargo activity at an airport from independently prepared forecasts of factors such as population size and profile, disposable income, breakdown between business and leisure travel, and cost of air fares. This sort of analysis compares historical data on passenger and/or cargo activity with comparable data relating to one or more independent factors, to create an equation that relates the passenger or cargo activity to the independent factors. The forecaster then uses this equation to refine forecasts of future passenger and/or cargo activity in the light of the modifying affect of the independent factors. Various statistical tests are used to evaluate the quality of the correlation of historical passenger or cargo volumes to the independent variables. Forecasters typically try to correlate the historical passenger data to multiple variables in an effort to improve the accuracy of the final result. Normally, forecasters use personal income, population, and air fares as the independent factors, because personal income and fares are factors in passenger decisions on whether to travel and whether to use air transportation. Population can be a predictor of the total volume of a market, once the propensity to travel is determined by personal income and fares. Regression analysis improves on trend analysis by relating passenger volumes to predictive factors that have independently prepared forecasts. These independent forecasts will likely capture demographic characteristics, migration trends, and employment profiles. More sophisticated independent forecasts may reflect local infrastructure improvements, industry changes, popu- lation densities, and land use controls. The use of regression analysis enables the assumptions that have informed the independent forecasts to be incorporated into the dependent forecasts of aviation activity. Thus, there is always uncertainty about the levels of types of future aviation activity at an airport. The terminal planner needs to understand the reasons behind uncertainty and then develop a plan that provides the airport flexibility to adapt to new circumstances. Airport terminal facilities are sized to accommodate the peak hour passenger volumes of a selected design day. Annual enplanements are an indicator of overall airport size; however, peak hour volumes more accurately determine the demand for airport facilities based on the specific 82 Airport Passenger Terminal Planning and Design
user patterns of a given airport. Peak hour passengers are typically defined as the number of passengers in the peak hour of an average day in the peak month (PHADPM) and are also often referred to as âdesign hour passengers.â The design hour measures the number of enplaned and deplaned passengers departing or arriving, on aircraft in an elapsed hour of a typically busy (design) day. The design hour typically does not correspond exactly to a âclock hourâ such as 7:00â7:59 but usually overlaps two âclock hours,â for example, 7:20â8:19 reflecting airline scheduling patterns. Each airport or terminal also has its own distinct peaking characteristics due to differences in airline schedules, business or leisure travel, long- or short-haul flights, the mix of mainline jets and regional aircraft, originating/terminating passenger activity or transfer passenger activity, and the balance of international and domestic passenger services. These peaking characteristics determine the size and type of terminal facilities. Thus, two airports with similar numbers of annual passengers may have different terminal requirements, even if the design hour passenger volumes are similar. The following sections will discuss how the user can proceed from annual forecasts (typically from an Airport Master Plan or the FAA TAFs) to the necessary design hour forecasts for terminal facility planning. These sections will also cover the similar process to convert annual aircraft operations to data needed to develop gate forecasts. IV.2 Data Sources Regardless of the forecasting method used, the forecasting process examines past aviation activity patterns, as well as other information, and attempts to predict what future activity will be based on the patterns of past activity. Thus, all forecasting efforts begin with gathering data about past aviation activity at the airport, as well as other information about the airportâs market that the forecaster may find useful in predicting future aviation activity. The following data sources are often tapped to prepare forecast information: ⢠Airport records ⢠Socioeconomic data ⢠Flight activity data ⢠Airline surveys ⢠Passenger surveys ⢠Other sources These data sources are described in the following sections. IV.2.1 Airport Records Virtually all commercial service U.S. airports keep monthly records of aircraft operations, passenger numbers, mail, and cargo by each airline. This information is used to invoice airlines for landing fees; therefore, these records are kept by the airport finance department. Historic information on monthly and annual traffic counts of passengers and aircraft operations should be obtained for as long a period as available, preferably long enough to cover a couple of boom economic times and a couple of economic recessions (airport activity tends to vary with popu- lation and economic activity). Traffic activity by specific airlines should be examined to identify when different carriers started and ended service, as this type of information can explain changes (either up or down) in total activity. Airport-generated traffic counts should be compared with FAA air traffic counts as a control measure. FAA records will capture total activity, of which traffic activity reported to the airport Forecasts 83
by the individual carriers will make up a significant portion. The airport usually reports the differ- ence between its records and FAA records as âmiscellaneousâ or âunknownâ activity. Usually this activity is by general aviation aircraft that do not use the passenger terminal. However, the planner should validate this with locally knowledgeable personnel at the airport. Some of this activity can include non-scheduled air charters that use the passenger terminal. In addition, this activity can qualify an airport for AIP assistance. Local airports report their information either on a calendar year or on a local fiscal year, while the FAA tends to report its annual information on a federal fiscal year basis. The terminal planner needs to be cognizant of the annual periods being used to report information by each agency and prepared to reconcile between data sources using monthly information from each agency. IV.2.2 Socioeconomic Data Aviation activity tends to vary with an airportâs service area (catchment area) population and per capita personal income. In addition, the number of flight destinations available and the price of air fares may increase the size of an airportâs service area, because customers tend to drive from further away to use low-fare or nonstop air service. The local census bureau or equivalent regional agency will likely have historic socioeconomic data. The forecaster should obtain a history of local socioeconomic data that at least covers the period of time that local airport traffic data has been available. Because historic aviation activity usually correlates fairly closely to past socioeconomic information, the forecaster should also obtain county-level forecasts of future socioeconomic trends for the airportâs service area. These forecasts are often available from a regional agency or from private data services that provide data nationally by county. IV.2.3 Flight Activity Data In addition to airport data, flight activity data is available from a wide range of sources, depending on the budget available and the level of detail desired. The FAA provides a substantial volume of aircraft activity data through its various web portals. Some of the more common data sources are described below. IV.2.3.1 Federal Aviation Administration The FAA provides data on aviation system operational performance, including airport utilization statistics, via its Aviation System Performance Metrics website (aspm.faa.gov). Some sections of this site require a user name and password (available from the FAA) for access. The site also provides access to the FAA TAFs. Not all types of data are available for all airports, because the FAA data focuses on airports that have FAA facilities such as air traffic control towers or terminal radar approach control facilities. IV.2.3.2 Bureau of Transportation Statistics The U.S. Department of Transportation (U.S.DOT) has centralized the processing and dis- seminating of various statistics on transportation usage within the Bureau of Transportation Statistics (BTS) (www.bts.gov). The BTS is home to several useful databases that describe airline and airport activity including the following: ⢠Form 41 databases, which provide airline-reported financial statements, aircraft operations, and passenger statistics. ⢠OD1A databases, which provide a 10% sample of passenger itineraries based on ticket infor- mation, passenger profiles, fares, city-pair travel, and revenues. ⢠T-100 databases, which provide information on load factors (passengers carried versus seats offered) by city-pair market. 84 Airport Passenger Terminal Planning and Design
The BTS website (www.bts.gov/programs/airline_information/sources/) provides a list of various vendors who preprocess and analyze the various BTS databases. These services may make analysis of these large data sets more efficient. IV.2.3.3 Official Airline Guide The Official Airline Guide (OAG) (29) provides flight-by-flight listings of scheduled airline activity (arriving and departing aircraft) with various data about each flight including aircraft type, number of seats, arrival and departure times, flight numbers, fares, and full flight itineraries (cities served by each flight). This data is compiled and sold by Official Airline Guide, Inc. as a manual, as data files available on CD, and as data files available for download. IV.2.4 Airline Surveys Sometimes demand for a terminal project is prompted by a change in a particular airlineâs business plan at the airport. In this case, it is very helpful to get information directly from the airline about future plans. Even when there is no specific announced plan, interviews of airline station management or property management personnel will give insights about future airline plans at the airport. However, the forecaster needs to treat forward-looking information (such as future flight schedules) from the airline company with some skepticism, because business plans do not always occur as expected and are subject to change. Ultimately, forecasters must apply their judgment about the viability of future airline changes. For example, two airlines may enter a single market simultaneously, each expecting to win the competition. The forecaster must judge whether the market is big enough to support both competitors beyond the short- term future. IV.2.5 Passenger Surveys Some forecasting information about air service is available only by asking the passenger: ⢠Trip purpose (business/government or leisure/discretionary/tourism) ⢠Visitor versus local passenger ⢠Whether the passenger considered using another airport when booking the trip Answers to these types of questions help the terminal planner understand the level of uncertainty and risk in the forecast. Surveys are usually sponsored by local airport operators to understand passenger attitudes about customer service, how long before flight departures do passengers arrive, ground transportation modes that passengers use, passenger demographics, spending at the airport (concession usage), or spending in the local economy (economic impact analyses). The information about trip purpose, type of passenger, and whether passengers considered using another airport may have been asked in past surveys, but these answers may not have been published. However, the survey sponsor will likely have a database with these survey results. Surveys reflect the conditions when questions were asked and may not reflect future con- ditions when underlying factors affect the mix of passenger types within the sample. A survey taken in winter will miss passenger profiles on summer vacations. Surveys taken before 2001 do not reflect security procedures put into effect after 2001. The analyst needs to understand the temporal context for the survey and how more recent changes in activity affect the relevance of the survey. For additional insights on correcting statistical information on passenger-related processing rates from passenger surveys, please reference ACRP Report 23: Airport Passenger-Related Processing Rates Guidebook (30). Forecasts 85
IV.2.6 Other Sources Every local airport service or catchment area has its own unique characteristics that will define its travel market. The forecaster should consider the use of any data source that will accomplish one of two goals in the forecasting process: ⢠Describe the future trends in air passenger demand ⢠Reduce the uncertainty in the forecast The forecast accomplishes these two goals by improving information quality and/or providing new/additional information. IV.2.6.1 Forecasting Basics and Sources of Forecast Information Most initial forecasting efforts provide only annual estimates of future enplaning or total pas- sengers and aircraft operations. These forecasts come from one or more of the following sources: ⢠FAA TAFs (www.aspm.faa.gov) ⢠State and/or regional aviation system plans ⢠Airport Master Plan forecasts ⢠Airline forecasts ⢠Other forecasts (such as a regional agency airport analysis or an environmental study) It is up to the terminal planner to extend the analyses provided in these forecasts to estimate the peak hour flows of passengers and aircraft that will be required to develop the terminal plan. As described in the overview, forecasting future demand has inherent uncertainties based on external factors that tend to upset the continuation of past economic trends. The terminal planner needs to understand the reasons behind uncertainty and then develop a plan that provides the airport flexibility to adapt to new circumstances. Some critical factors that create uncertainty are discussed in the following list: ⢠Airline network planning: Many U.S. airlines structure their route networks around one or more strategically located âhubsâ that have passenger flows from less heavily trafficked feeder routes consolidated before transfer to and from longer haul sectors with higher load factors. While the advent of more low-fare carriers and regional jets is increasing the proportion of point-to-point services, business at many airports is still heavily dominated by the main local hub carrier. Thus, airline decisions to open, expand, contract, or close down a hub loca- tion can radically affect the volume of traffic passing through a particular airport. Airline strategic business decisions of this sort can occur for reasons that are largely independent of local factors. ⢠Local economic factors: While local economic forecasts are generally more reliable than the forecasts of future airline plans, the more dependent a community is on a single industry or corporate activity, the more susceptible are local airport traffic levels to factors affecting that industry or activity. For obvious reasons, smaller airport traffic levels are more vulnerable in this respect than airports serving a larger, more diverse economy. ⢠Airportsâ competition for passengers: Airports have overlapping service areas. While the great majority of passengers originate within a 1 hour travel time of the airport, some passengers come from up to 3 hours away. The presence of a low-fare carrier will tend to attract passengers from longer distances if no competing service is available from a closer airport. Changes in airline service at competing airports can affect local traffic. The best response to uncertainty is to build alternative forecasts that have scenarios that are sensitive to one or more of the factors described above. Most airports now keep statistics on historic traffic activity. The planner can assess how this traffic rose and fell in response to external 86 Airport Passenger Terminal Planning and Design
factors. Study of this history in the context of airline actions, local economic events, and activity at other airports enables the development of low and high case scenarios, in addition to the main expected case, including scenarios that have different mixes of connecting versus origin/destination traffic. The purpose of having a range of forecast scenarios is to have an understanding of the magnitude of development risk associated with investing in expanded airport facilities. This helps the planner and stakeholders evaluate the various terminal expansion options and select the most prudent and cost-effective planning solution. IV.3 Typically Forecasted Information and Forecast Validation Issues While forecasts usually focus on future passenger activity, terminal plans also need estimates of belly cargo or mail and aircraft arrivals and departures (operations). Once forecasts are prepared, then there are several activities that the forecaster can conduct to gain stakeholder acceptance of forecasting results. IV.3.1 Passengers Usually, forecasters spend the most time and effort on forecasting passenger volumes, using the techniques described above, because passenger volumes form the basis for many of the analyses that the planner will use in evaluating the functional requirements of the terminal. While passenger growth can be forecast in totality, it is often more beneficial to forecast specific categories of passengers independently. Numbers of domestic origin and destination passengers tend to vary with local economic factors and local air fares. International passenger traffic will be driven more by global economic and business considerations and the relative importance of business and tourism links with the overseas regions concerned. The numbers of transfer passengers will vary depending on the size and role of the hub airport and the airline route development and network strategy. The forecaster may need to get infor- mation about connecting passenger volumes from the hub airline if there is substantial connecting activity at the airport. IV.3.2 Belly Cargo and Mail Larger aircraft, especially widebody aircraft on international flights, will also carry cargo and mail in addition to baggage in the lower compartments of the aircraft. Similar to passengers, the volume of this activity will also vary with local economic conditions and the price of air cargo. Small changes in price may divert domestic air cargo to truck, as long as the truck will get the cargo to the destination in sufficient time. The time required and the expense of cargo security may also determine whether air cargo and mail goes in the belly of a passenger aircraft, by a dedicated cargo aircraft, or by truck (if domestic). The planner should be sure to evaluate the qualities of the air cargo data they receive. Some âair cargoâ does not go by air at all, but is still recorded as air cargo. Some data is based on estimates and not actual tonnage reports. If possible, the planner should cross check various data sources to determine reasonableness. IV.3.3 Aircraft Operations Aircraft operations by commercial airlines are normally derived by determining the average number of passengers on board each aircraft, determining the expected percentage of seats filled (load factor), and then dividing the volume of boarding passengers by these two numbers to Forecasts 87
determine the total number of aircraft operations. Forecasting growth in aircraft operations also involves evaluating whether the airlines will introduce new or different aircraft types in order to tailor their services more closely to the demand profile. This element of the forecast also involves assessing the airlinesâ fleets and determining what types of aircraft are available for substitution. The aircraft manufacturers publish the aircraft they have on order from each airline. The forecaster also needs to be aware that growth in aircraft operations may be limited by airfield constraints that cannot be resolved through additional construction. In this case, average aircraft size may increase to absorb demand growth unless the relevant airlines have no plans to order larger aircraft. Without growth in the size of aircraft, airfield delays will likely grow to an unacceptable level, at which point the FAA may step in and limit the number of hourly aircraft operations when delays increase beyond an average of 20 to 25 minutes per aircraft. IV.3.4 Benchmarking and Stakeholder Buy-in Because many assumptions about future terminal size and configuration depend on long-term demand forecasts, obtaining stakeholder buy-in is an essential step in the planning process. Four steps help improve the acceptability of forecasts to stakeholders and increase the likelihood of obtaining a broader consensus. IV.3.4.1 Consider and Use Stakeholder Data in the Forecast Whenever Possible Forecasters need to consider a wide range of data sources in order to identify as many cause and effect relationships between basic information and passenger volumes. Airlines have some of the most detailed information about passenger profiles available. In addition, they may be willing to share their opinions about future market conditions at an airport, or within a regional airport system, providing they do not perceive such disclosure as a breach of commercial confidentiality. While these opinions might not be included in the final documentation, the simple statement, that stakeholder data and opinions were considered in the creation of the forecasts, improves the chance for obtaining a stakeholder consensus that the forecasts are reasonable. IV.3.4.2 Examine the Sensitivity of the Forecast to Variations in the Independent Variables or Changes in Major Assumptions The selection of independent variables and the forecast assumptions underpinning them can be attacked as being somewhat subjective. The forecaster should present a discussion on how changing independent variables and assumptions will affect the forecasts. This could include a complete forecast using alternative assumptions, or it could include a simpler discussion on whether a changed assumption will depress or stimulate traffic levels. Presenting alternatives, along with a narrative that describes the rationale for choosing a particularity of assumptions should improve the chances for acceptance of the final forecast among stakeholders. Discussing this analysis with stakeholders before completing and accepting the forecasts should further increase stakeholder acceptance of the final forecast. IV.3.4.3 Evaluate Alternative Methodologies and Compare Results While forecasters may have their favorite methodologies, not all methods work in all places. A technique that produced an acceptable outcome at one airport may be ill suited at another. Further, a methodology that worked in a previous study may no longer work in a new or repeat study. Finding the methodology that best represents the current conditions involves testing alter- natives and finding a method that provides the best explanation and support for the prediction of future conditions. Explaining the choice of methodology to stakeholders should improve their acceptance of the final forecasts. 88 Airport Passenger Terminal Planning and Design
IV.3.4.4 Benchmark (Compare) the Local Forecast to Other Forecasts, as well as Regional and National Forecasts Comparing forecasts to other relevant forecasts demonstrates that other parties share a similar opinion about the future growth of aviation activity at the airport. If the results are not broadly similar, the comparison should include an analysis of the factors which cause the forecast to differ. The benchmark forecasts could include previous forecasts at the same airport, system plan forecasts prepared by a state or regional agency, and the FAA TAF. If the forecasts will be used to support FAA Airport Improvement Program funding decisions, then the FAA expects the local forecast to be within 10% of the TAF after 5 years and within 15% after 10 years. If the forecasts do not agree with the TAF, the FAA will require compelling documentation that describes why the local forecast should be used in lieu of the TAF for funding decisions. If the planner elects to use the TAF and expects to use the forecast for a larger master planning effort, the FAA will expect the planner to prepare an analysis that validates the TAF. IV.3.5 Planning Activity Levels Because forecasts typically project demand over a particular time frame, such as 20 years, removing the time frame from the analysis and focusing on an activity level may increase the shelf-life of a future plan, because the plan still provides the appropriate response to a future level of demand even if the forecast of when future demand occurs is incorrect. Use of a planning activity level instead of a time-based forecast will focus planning decisions on the size and con- figuration of a terminal instead of whether it is advisable to proceed with any development at all. Use of planning activity levels is especially useful for long-range planning when major design and construction will not occur for some period of time after the terminal planning occurs. Use of planning activity levels also allows the planner to round activity levels to representative levels rather than focus on serving a specific demand level. IV.4 Peak Hour Demand Analysis IV.4.1 Defining the Design Hour As already discussed, the terminal planner must have, as a basic planning tool, forecasts of peak hourly passenger loadings that the terminal and its various systems may have to cope with. These loadings are also referred to as âdesign hour passengers,â a term which the Guidebook will use for consistency. The design hour is not the absolute peak level of activity, nor is it equal to the number of persons occupying the terminal at a given time. It is, however, a level of enplaning and deplaning activity that the industry has traditionally used to size many terminal facilities. The number of persons in the terminal during peak periods, including visitors and employees, is also typically related to design hour passengers. There are a number of methods for determining design hour passengers. One approach is to define the design hour as the 90th (or 95th) percentile busiest hour of the year. This approach requires keeping track of all enplaning and deplaning passengers for every flight during the year, and then ranking these by hour (usually a clock hour) to find the level of activity that accounts for 90% of the annual traffic. While used by some non-U.S. airports, it is a very data-intensive approach for which data is not available for the vast majority of U.S. airports. In the United States, the design hour is typically defined as the peak hour of an average day of the peak month. The design hour measures the number of enplaned or deplaned passengers departing or arriving on aircraft in an elapsed hour of a typically busy (design) day. The design hour typically does not correspond exactly to a âclock hourâ such as 7:00â7:59, but usually overlaps two âclock Forecasts 89
hours,â that is 7:20â8:19 reflecting airline scheduling patterns. More information about selecting the design hour is presented in the following section. IV.4.2 Estimating Design Hour Passenger Activity Estimating design hour passengers is typically a three-step process: ⢠Determine the peak month ⢠Determine the design day to be used ⢠Estimate the amount of daily activity that occurs in the design hour. This process is applied to both existing (base year) conditions as well as activity in future years. IV.4.2.1 Peak Month The peak month is based on historic patterns of passenger activity. It is recommended that 3 to 5 years of data should be collected to determine if the peak month is a constant or changing percentage of annual activity. The peak month may be different for enplaned and deplaned pas- sengers, domestic and international, and so forth. Table IV-1 is an example of monthly enplanement data used in the Spreadsheet Models to determine the peak month percentage. For this airport, the peak month for enplanements was consistently August, but the percentage varied over the 5-year period. The peak month percentage may be modified for the future based on local trends and/or anticipated changes in service patterns. IV.4.2.2 Design Day For most airports, an average day of the peak month is used for the design day; the peak month is simply divided by the number of days in that month. An alternative to using the average day of the month is to use an average weekday. This alterna- tive is often used at airports that have domestic service as the predominant activity and weekend activity is less than weekday activity. Airport records on monthly and daily passenger volumes (as recorded by the airlines) are the best source for determining whether an average day or an average weekday is the appropriate (design) day for the design hour. 90 Airport Passenger Terminal Planning and Design 2004 339,212 335,431 380,372 383,986 384,009 412,229 433,519 438,881 359,801 392,988 389,683 390,748 2005 351,751 343,331 410,799 410,089 417,314 431,319 448,310 453,798 381,840 396,737 390,193 386,018 2006 346,250 345,682 406,676 412,639 410,434 430,066 437,895 446,311 373,111 401,655 395,973 407,416 2007 371,721 365,513 432,975 433,370 438,341 452,244 456,592 478,329 388,735 414,229 390,115 366,854 2008 350,450 350,533 408,656 392,136 385,109 398,749 411,909 419,764 342,455 362,867 325,972 344,026 2004 4,640,859 386,738 438,881 August 9.5% 2005 4,821,499 401,792 453,798 August 9.4% 2006 4,814,108 401,176 446,311 August 9.3% 2007 4,989,018 415,752 478,329 August 9.6% 2008 4,492,626 374,386 419,764 August 9.3% Average Peak Month AUGUST 9.4% Year Jan. Feb. March April May June July Aug. Sept. Oct. Nov. Dec. Year Total Monthly Maximum Peak PM Average Value Month % of Year Average Peak Month Percentage of Annual Source: Landrum & Brown Table IV-1. Example of peak month percentage of annual enplanements.
At a small number of airports with highly variable activity from day to day, a specific day of the week may be chosen as the design day. For example, a vacation destination airport may have significantly more activity with larger aircraft on Saturdays than during the week. The concen- tration of larger aircraft would affect both the daily and peak hour activity. IV.4.2.3 Design Hour Passengers The design hour can be computed by one of two approaches: using a design day flight schedule (DDFS) or applying percentage factors against a design day passenger volume. The most frequently used source of data for assembling a DDFS is the OAG. Data from OAG needs adjusting to account for non-scheduled activity such as charters. Choosing the correct day for a flight schedule should take into account monthly activity statistics from the airport operator and daily flight activity from either the FAA or OAG. The chosen day should be representative of activity for the design day as discussed above. Because the primary use of the design hour is to estimate passenger activity, the day chosen for the OAG schedule should reflect a day with a representative number of scheduled seats. DDFSs should have, at a minimum, the following information: ⢠Airline name (usually listed as a two-letter code) ⢠Aircraft type (usually listed as a three-letter code) ⢠Number of seats on the aircraft (usually derived from the airline name and aircraft type information) ⢠Time of flight arrival or departure ⢠Destination or origin of the flight (used to determine whether a flight is using international or domestic terminal facilities) ⢠Flight number In addition to these six fields, flight schedules can contain several other pieces of information about each flight, such as operating airline (versus the marketing carrier), the time the aircraft left the originating airport or arrived at the destination airport, and the days of week the flight operates. Days of operation are especially important for airports with flights that may not operate daily. If an airfield simulation study is being done, a usable DDFS and fleet mix could be avail- able directly from that study. However, the planner should validate that a DDFS chosen for the airfield analysis also has representative information for a design day for passenger terminal planning. Regardless of the source of the DDFS, the terminal analyst should consider that a DDFS presents a very detailed description of a single day. The analyst should evaluate whether the flight activity schedules are truly representative (through examination of other historical schedules for the same airport) or whether it represents a unique case. See Section IV.4.4 for a discussion of the factors involved in developing a DDFS. The rolling peak hour seats can be determined from a flight schedule using a spreadsheet analysis. As shown in Figure IV-1, the peak hour (design hour) usually does not occur in a clock hour, but across two hours. This type of analysis can be done using scheduled seats (as in Figure IV-1) or flight-by-flight actual passenger loads. The advantage of using a DDFS is the great level of detail it provides. The disadvantage is that small variations in the schedule can vary results in ways that may be difficult to explain. For example, the shift in a flight by a few minutes can determine whether or not it falls within the busiest hour and thus affect the percentage of daily activity in that peak hour. Some analysis that compares percentage of activity in the design day and design hour should be used to ensure that the schedule provides representative information. Forecasts 91
92 Airport Passenger Terminal Planning and Design If flight-by-flight data is not available, the design hour is estimated as a percentage of daily activity. These percentages (enplaned and deplaned) should be based on actual passenger activity data collected from the airlines for a typical week. Actual passengers per flight data is preferable to scheduled seats because it is more accurate. For example, to position aircraft for the next morningâs departures, an airport may have a late night arrivals peak in terms of seats. However, the actual deplaning peak may be in late afternoon or early evening when load factors are higher. IV.4.2.4 Sample Design Hour Calculations Table IV-2 illustrates a typical approach to estimating future design hour activity as contained in the Spreadsheet Models. In this example, the peak month factor has been held constant based on an average of the last 5 years, although it could be varied. An average day has been used as the design day, so the future peak month is divided by 31 days. The percentage of daily activity represented by the design hour is currently estimated to be 15.4% for enplaned and 12.5% for deplaned. This estimate is based on actual flight-by-flight activity, scheduled seats, or a combination of assumptions including scheduled seats and peak hour and average daily load factors. Most planners will assume some variation in the design hour percentage of daily activity for the future based on local trends and/or anticipated changes in service patterns. If there are many âvalleysâ in the current schedule, adding off-peak flights or assuming increasing loads in off-peak ta snigeB ruoH kaeP:ruoH gnilloR kaeP 6:40 Peak Seats = 5,324 = latoT yliaDSTAES GNITRAPED 56,443 % of Daily = 9.4% 0 1000 2000 3000 4000 5000 6000 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 S ea ts Note: This chart was created in the Design Hour Determination model, which is part of the Spreadsheet Models provided in Volume 2 and was developed to work with the material in this volume. Figure IV-1. Design hour for departing seats. For additional insight and practical help in performing the determinations and methods described in this section, go to the Design Hour Determination model provided in Volume 2: Spreadsheet Models and Userâs Guide. This model takes the user through the steps to determine the peak hour planning factors that can be used when a DDFS is not available.
times may reduce the design hour as a percentage of daily activity over time. This is the type of assumption typically used and is reflected in the example. In contrast, there may be an increase in the peak hour percentage (especially in the short term) if forecasts indicate that new service would be provided in the current peaks without a concurrent increase in off-peak activity. Regardless of the analysis method used to derive the design hour volumes of passenger and aircraft activity, the method used to create the analysis must be calibrated against data or obser- vations that reflect actual operations at the airport. IV.4.2.5 Other Design Hour Considerations O&D vs. Connecting Passengers. After determining the total number of design hour passengers, it is usually appropriate to divide this volume into O&D passengers and connecting passengers. Connecting passengers usually stay on the airside area of the terminal while the O&D passengers make use of both the airside and the landside areas of the terminal. The only time connecting passengers make use of the landside facilities is when they change airlines Forecasts 93 Calendar Total Year Enplanements Base 2008 4,492,626 Forecast 2010 4,168,100 2015 4,732,800 2020 5,381,300 2025 6,104,700 2030 6,925,300 Base 9.4% Peak Month Factor 2008 419,764 ~from Peak Month Tab Forecast Peak Month Factor 2010 9.4% 391,800 2015 9.4% 444,900 2020 9.4% 505,800 2025 9.4% 573,800 2030 9.4% 651,000 Base 31 # of Days in Peak Month 2008 13,540 Forecast 2010 12,640 2015 14,350 2020 16,320 2025 18,510 2030 21,000 Base % of Average Day Enplaned % of Average Day Deplaned 2008 15.4% 2,080 12.6% 1,700 Forecast 2010 15.4% 1,950 12.6% 1,590 2015 15.0% 2,150 12.2% 1,750 2020 14.7% 2,400 12.0% 1,960 2025 14.0% 2,590 11.7% 2,170 2030 13.5% 2,840 11.5% 2,420 ANNUAL PEAK MONTH DESIGN HOUR PEAK MONTH AVERAGE DAY Source: Landrum & Brown. Table IV-2. Design hour calculation example.
and the onward airlineâs gates are located in a different concourse or terminal with no airside connector. Connecting passenger volumes tend to vary considerably from airline to airline. In general, the larger the volume of flight activity that an airline has at an airport, the greater is the likeli- hood that connecting passengers will be a part of that airlineâs total passenger volume. Informa- tion on connecting passenger volumes should be collected from the airlines if possible. Analysis of U.S.DOT databases on O&D travel (10% ticket survey) and connecting travel can also be used to derive factors that calculate connecting versus O&D traffic flows. Visitor Ratios. In addition to passengers, other populations are present in the terminal building and use its various facilities. Frequently, the largest volumes of non-passengers (besides airport, concession, and airline employees) are visitors who usually are present to send off or greet passengers. The best method for determining the number of visitors to the terminal is through a passenger survey. The survey information is then used to determine the number of visitors per arriving or departing passenger. Care should be taken to identify those visitors who actually enter the building as opposed to those who just drive the passenger to or from the airport and never actually leave their cars. IV.4.3 Determining Design Hour Aircraft Operations Three types of analyses are appropriate for aircraft operations: determining the number of arrivals; the number of departures; and the number of aircraft on the ground, either parked at contact gates or remote stands. The first two numbers are useful to evaluate the capacity of taxiways and taxilanes, while the third is useful for determining the number of contact gates and remote stands required for each aircraft type. If a DDFS is not used for future design levels, then the forecaster should evaluate the number of passengers and seats per flight operation, airline aircraft orders, and other forecast information to judge how the mix of aircraft in the design hour will change in future years. It is possible and sometimes necessary to match arriving flights to departing flights to prepare detailed analyses of gate usage. Flight numbers and aircraft operator codes can provide useful information in the matching process. However, the best information on which arriving flights become the next departures will be obtained from actual records of gate usage. More detailed information on gate analysis is presented in Chapter V. The forecaster should also examine the fleet mix across several representative hours of the day to identify other hours of the day that have critical aircraft that do not arrive or depart in the design hour. For example, an international flight may use a widebody aircraft, but this flight may not occur in the design hour. In addition, the fleet of aircraft that Remain Overnight (RON) should also be examined to determine whether overnight parking needs are different from the design hour. IV.4.4 Developing Design Day Flight Schedules DDFSs present a level of detail that can enhance the planning effort and sometimes aid in gaining stakeholder acceptance of planned improvements. The schedules contain detailed flight- by-flight information that can be time consuming to prepare, evaluate, and manipulate, depending on the reliability of the sources of information available and the software tools available to manipulate, summarize, and present the information they contain. The most usual starting point for creating an existing design day schedule is using data available from the OAG. The planner still must go through the process of estimating the design day and design hour demand levels, and then select an OAG for a day that most nearly matches the selected 94 Airport Passenger Terminal Planning and Design
design day and design hours. In addition to the OAG, the planner should also use airport and FAA records to identify non-scheduled activity that the OAG does not contain. The flight schedule initially will contain two parts: a listing of arrivals and a listing of departures. The planner may choose to link arriving flights to subsequent departing flights. Linking flights allows the planner to create Gantt charts of gate usage. More information on creating gate charts and the analysis of gate usage is described in Chapter V. DDFSs for forecast years are usually created by adding flights to the existing schedule and by editing existing flights to show new aircraft types, or changed times of arrival or departure. The following points should be considered when creating forecast schedules: ⢠Forecast design day and design hour values should be estimated prior to creating forecast schedules. Because a DDFS usually has a whole dayâs activity, hourly target values should be set. ⢠Design day aircraft fleet and airline mixes should be estimated prior to creating forecast schedules. ⢠Flight schedules should be edited one airline and one market at a time. ⢠Adding an additional flight by an airline to a market usually requires retiming the remaining flights in that market to provide a realistic frequency of flights to that market. ⢠The future ability of an airline to increase the size of an aircraft on a flight versus adding frequency of service should be considered. The planner should consult published information about an airlineâs fleet and aircraft orders or ask the airline about future fleet plans. ⢠Shorter haul markets tend to have more frequent service versus longer haul markets, although many exceptions occur. ⢠The inbound and outbound fleet mix to each market is usually the same. ⢠Airline orders for aircraft and network routing strategies should be considered when adding or changing flights for a future demand level. ⢠Long-haul and international markets have times of day when operations are not viable because of time zone constraints, especially eastbound flights. Generally, both ends of a domestic flight should arrive or depart between 6:00 a.m. and midnight. For example, a transcontinental flight that leaves Los Angeles at 6:00 a.m. (Pacific Time) will not arrive in New York until 2:30 p.m. (Eastern Time). Thus, a future DDFS for a New York airport should not show transcontinental flights arriving before 2:30 p.m. These time zone constraints often define peak hours, especially in international terminals. DDFSs developed from an OAG reflect the constraints of existing terminal facilities, runway lengths and widths, and specific airline leaseholds. If the goals of future planning include removing these constraints, the development of future flight schedules should reflect a less constrained or unconstrained future facility environment. A DDFS provides estimates of seats and aircraft operations available at the terminal. However, the planner needs to create reasonable estimates of passenger load factors to convert the seat counts into passenger volumes. These load factors will likely vary from airline to airline and by time of day. Design hour load factors usually range between 80% and 95% of seats filled. A DDFS contains a very specific sequence of flights, which will generate a very specific volume of aircraft activity at the gates, and flows of passengers and bags through the terminal building. The planner should spend some time understanding whether an analysis based on a DDFS is truly representative of a range of future activity or reflects a unique condition resulting from the specific sequence of flights in the schedule. Forecasts 95
