chapter 6 airport capacity alternatives

Chapter 6

AIRPORT CAPACITYALTERNATIVES

Photo credit: Federal Aviation Administration

Dunes International I

Contents

PageIntroduction . . . . . . . . . . . . . . . . . . . . . . . . . . 101Airside Components. . . . . . . . . . . . . . . . . . . . 102Limitations on Airside Capacity. . . . . . . . . . 103

Aircraft Performance Characteristics. . . . 103Wake Vortex... . . . . . . . . . . . . . . . . . . . . . 103Weather . . . . . . . . . . . . . . . . . . . . . . . . . . . . 103Airfield/Airspace Configuration. . . . . . . 105Aircraft Noise . . . . . . . . . . . . . . . . . . . . . . . 106ATC Equipment and Procedures.. . . . . . 107Demand Considerations. . . . . . . . . . . . . . 107

Delay and Delay Reduction. . . . . . . . . . . . . . 107Demand-Related Alternatives . . . . . . . . . . . . 109

Peak-Hour Pricing. ......... . . . . . . . . . 109Quotas. . . . . . . . . . . . . . . . . . . . . . . . . . . . . 110Balanced Use of Metropolitan Area

Airports . . . . . . . . . . . . . . . . . . . . . . . . 110Restructuring Airline Service Patterns. . . 111Reliever Airports. . . . . . . . . . . . . . . . . . . . . 112

Airport Development Alternatives. . . . . . . . 113Expanding Existing Airports. ....... . . 113Development of Secondary Runway

Operations. . . . . . . . . . . . . . . . . . . . . . 113Building New Airports. . . . . . . . . . . . . . . . 114

ATC Improvement Alternatives. . . . . . . . . . 116Airfield/Airspace Configuration

Management . . . . . . . . . . . . . . . . . . . . 116Wake Vortex Prediction. . . . . . . . . . . . . . . 116Microwave Landing System . . . . . . . . . . . 117Reducing Separation or Spacing

Minimums . . . . . . . . . . . . . . . . . . . . . . 117

Page

Automated Metering and Spacing . . . . . . 118Cockpit Engineering. . . . . . . . . . . . . . . . . . 118

Summary of Alternatives. . . . . . . . . . . . . . . . 119Future Research Needs. . . . . . . . . . . . . . . . . . 121

Wake Vortex Avoidance. . . . . . . . . . . . . . 121Wake Vortex Alleviation. . . . . ........ 121Noise . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 121Airport Design . . . . . . . . . . . . . . . . . . . . . . 121Ground Access. . . . . . . . . . . . . . . . . . . . . . 122

LIST OF TABLES

Table No. Page8. ’’Top’’ U.S. Airports, by Enplaned

Passengers, by Air Carrier Operations,and by Reported Delays.. . . . . . . . . . . . 101

9. Arrival and Departure Separations. . . . 10410. Operational Characteristics of Airports

With Potential Benefits From aSeparate General Aviation Runway. . . 115

11. Summary of Alternatives.. . . . . . . . . . . 119

LIST OF FIGURES

Figure No. Page26. Airport Hourly Capacity Varies

Strongly With Weather. . . . . . . . . . . . . . 10427. Runway Configuration. . . . . . . . . . . . . . 10528. Typical Distributions of Delay. . . . . . . . 108

Chapter 6

AIRPORT CAPACITY ALTERNATIVES

INTRODUCTION

The ability of airports to accommodate trafficcan be expressed in terms of “airside” or “land-side” capacity. “Airside” capacity is defined hereas the number of air operations—landings andtakeoffs—that the airport and the supporting airtraffic control (ATC) system can accommodatein a unit of time, such as an hour. The capacityof an airport is not a single number, but willvary with the number of runways in use, the vis-ual or electronic landing aids available, the typesof aircraft being accommodated, the distancebetween aircraft in the approach pattern, andthe noise abatement procedures in effect. Thetime each aircraft occupies the runway and thefacilities for handling aircraft on the ground, ontaxiways, or at gates also affect airside capacity.All of these factors will vary depending on theweather.

“Landside” considerations, such as the sizeand number of lounges or the adequacy of bag-gage-handling equipment, affect the number ofpassengers an airport terminal can accommo-date. Ground access, including the adequacy oftransit connections, roadways, and parkingareas for passengers’ cars, is an important part of

an airport’s landside capacity, and in some caseshas become a limiting factor on an airport’s abil-ity to handle passengers. Recent discussionabout putting a quota on operations at Los An-geles International Airport, for example, is re-lated to growing ground access problems, notlack of airside capacity.

This chapter discusses alternatives to increaseairport airside capacity. Landside problems willonly be treated here as they affect airside capa-city.

When the traffic demand for an airport ap-proaches or exceeds its capability, the result isdelay. Delay has been a major problem at theNation’s busiest airports, resulting in millions ofdollars of increased operating costs for air carri-ers and wasted time for travelers. Although sev-eral different methods of measuring delay exist(as will be discussed later) it is generally agreedthat the six airports most affected by delay in1980 were: O’Hare (Chicago), Stapleton (Den-ver), La Guardia and JFK (New York), Harts-field (Atlanta), and Logan (Boston). As shownin table 8, most of the airports which report

Table 8.— “Top” U.S. Airports, by Enplaned Passengers, by Air Carrier Operations,and by Reported Delays

Passenger Air carrier Delays overenplanements operations 30 minutes

1.2.3.4.5.6.7.8.9.

10.11.12.13.14.15.

Chicago O’HareAtlanta HartsfieldLos Angeles InternationalNew York J.F. KennedySan Francisco InternationalDallas-Ft. WorthDenver StapletonNew York La GuardiaMiami InternationalBoston LoganHonolulu InternationalWashington NationalDetroit MetroHouston IntercontinentalSt. Louis Lambert

Chicago O’HareAtlanta HartsfieldLos Angeles InternationalDallas-Ft. WorthDenver StapletonMiami InternationalSan Francisco InternationalNew York La GuardiaNew York J.F. KennedyBoston LoganWashington NationalSt. Louis LambertDetroit MetroHouston IntercontinentalHonolulu

Chicago O’HareDenver StapletonNew York La GuardiaNew York KennedyAtlanta HartsfieldBoston LoganLos Angeles InternationalSt. Louis LambertSan Francisco InternationalDallas-Ft. WorthPhiladelphia InternationalNewarkWashington NationalMiami International

SOURCE: Federal Av/at/on Adm/n/strat/on, Term/na/ Area Forecasts, Fisca/ Years 1981-92, Washington, D.G. 1981 p 13; in-terview, FAA, Air Traff/c and A/rways Fac//dies, Aug. 20, 1981.

101

102 ● Airport and Air Traffic Control System

serious delay problems rank among the top 15airports in terms of both enplaned passengersand air carrier operations.

This chapter first describes the airside compo-nents in the operation of a typical airport. Itthen reviews those major factors which influenceor limit airside capacity. Next the chapter dis-cusses the problem of delay—how it comesabout and the methods for measuring it and esti-

mating its costs. The next sections outline somealternative methods for reducing delay or in-creasing the airside capacity. These includechanging the pattern of traffic demand, expand-ing the runway system, or modifying the termi-nal area air traffic control procedures and equip-ment. Finally, some suggestions for future re-search are made.

AIRSIDE COMPONENTS

The airside capacity of an airport is governedby factors related to its runway system and theairspace above and around the airport, as wellas the terminal area ATC and navigation equip-ment and procedures.

The number of runways, their layout, length,and strength will in large measure determine thekinds of aircraft that can use the airport andhow many aircraft can be accommodated in anygiven time period. The layout depends on anumber of factors including the local terrain andpredominant direction of the wind. Federal Avi-ation Administration (FAA) safety regulationsdictate how close the runways may be to one an-other and to buildings, trees, or other obstruc-tions.

In order to land on a runway, aircraft ap-proach the runway in single file, with a safe dis-tance between them. Air traffic may enter theairspace around the airport (“terminal area”)from many directions at a number of differentpoints (“entry fixes”), and in many metropolitanareas the aircraft may be destined for one of sev-eral different airports. Thus, the task of deliver-ing aircraft one by one to a particular runway ata particular airport must begin many miles fromthe airport itself, and controllers must orches-trate the orderly merging and diverging of manydifferent traffic streams until each aircraftreaches the final approach to its destination run-way. By the same token, departing aircraft mustbe safely routed from the airport to the “depar-ture fix” where they leave the terminal area andjoin the en route ATC system.

Controllers use bothseparation to maintain

vertical and horizontalsafe distances between

aircraft, a task that is complicated by their dif-ferent performance characteristics. Jets flying ata very slow (for a jet) 160 knots will neverthelessovertake and pass slower aircraft. The controllermay assign different altitudes so that this cantake place safely, or he may vector the faster air-craft along a longer path so that it will safelyovertake and pass around the one ahead.

In good visibility conditions, tower control-lers may clear aircraft, once they are in sight ofthe airport, to make a visual landing undertower control. The pilot assumes responsibilityfor separating himself from other aircraft, withthe controller standing by to warn pilots to “goaround” in case of a potential conflict. Duringtimes of poor visibility the ATC team retains re-sponsibility for separating the aircraft on finalapproach. In this case the Instrument Flight Rule(IFR) radar minimum separation is observed, so

Photo credit Neal Callahan

The variety of airspace system users .

Ch. 6—Airport Capacity Alternatives ● 103

that distances between aircraft are greater thanin good weather. Under IFR conditions, pilotsare much more dependent on landing aids suchas the Instrument Landing System (ILS) to guidethem to the runway.

An aircraft is considered to be on the runwayfrom the moment it flies over the runway thresh-old until it turns off onto a taxiway. Angled“high-speed” turnoffs can allow aircraft to leavethe runway at higher speeds than perpendicular

convenient to most of the aircraft using a run-way is important for getting maximum capacityfrom the runway system.

Departures from the airport may take placeon a separate runway or may be “interleaved”between arrivals on the same runway. Aircraftpreparing to depart can wait beside the runwayon holding aprons until the runway is clear; thenthey can then taxi onto the runway and take offfairly quickly—the time spent on the runway for

ones. Placing the turnoffs where they will be departure is on the order

LIMITATIONS ON AIRSIDE CAPACITY

Among the major factors influencing airportcapacity are: aircraft performance characteris-tics, wake vortex turbulence, weather, airfieldand airspace configuration, aircraft noise, ATCequipment and procedures, and demand con-siderations.

Aircraft Performance Characteristics

Characteristics of the aircraft—their size, aer-odynamics, propulsion and braking perform-ance, and avionics—will affect the capacity ofthe runways they use. Pilot training, experience,and skill will also influence performance, andthe capacity of a runway can vary greatly withthe types of aircraft using it. Runway capacity isusually highest if the “traffic mix” is uniformlysmall, slow, propeller-driven aircraft. The nexthighest capacity would come with a uniform mixof large jets. Where the traffic mix is highly di-verse—with jet and propeller aircraft of widelyvarying sizes and speeds—it is usually difficultto maintain optimum spacing and optimum run-way usage, and runway capacity is reduced. Thedirection of traffic also affects runway systemcapacity. When arrivals predominate, capacityis lower then when departures predominate.

Wake Vortex

Related to aircraft performance characteristicsis the problem of wake vortexes. Aircraft pass-ing through the air generate coherent energeticair movements in their wakes, and under quies-

cent weather conditions

of 30 seconds.

the wake vortex canpersist for 2 minutes or even longer after an air-craft has passed. The strength of the vortex in-creases with the weight of the aircraft generatingit. As the use of wide-bodied jets (e.g., B-747and DC-10) became more common in the early1970’s, it became apparent that wake vortexesbehind these heavy aircraft were strong enoughto endanger the following aircraft, especially if itwas smaller. Until the potential danger of wakevortex to transport sized aircraft was demon-strated (e.g., the 1972 crash of a DC-9 landing inthe wake of a DC-10) standard separations of 3nautical miles (nmi) were required under IFRconditions. In order to prevent accidents causedby wake vortexes, FAA increased the separa-tions for smaller aircraft behind larger ones dur-ing weather conditions when persistent vortexesmay be a danger. These minimums are shown onthe right side of table 9.

Weather

Heavy fog, snow, strong winds, or icy run-way surfaces reduce an airport’s ability toaccommodate aircraft and may even close anairport completely. For a given set of weatherconditions, several of the different runway con-figurations available at an airport may be suit-able but only one will have the maximum value.Using these maximum values, and plotting themwith the percentage of the year during which dif-ferent weather conditions are likely to prevail, a

104 • Airport and Air Traffic Control System

Table 9.—Arrival and Departure Separations

Minimum ArrivalVisual Flight Rules*

Leads L H

s 1.9 1.9 1.9

L 2.7 1.9 1.9

H 4.5 3.6 2.7

Separations— Nautical MilesInstrument Flight Rules

Leads L H

s 3 3 3

L 4 3 3

H 6 5 4

Minimum DepartureVisual Flight Rules*

Leads L H

s 35 45 50

L 50 60 60

H 120 120 90

Separations— SecondsInstrument Flight Rules

Leads L H

s 60 60 60L 60 60 60

H 120 120 90

“VFR separations are not operational minima but rather reflect what field data show under saturated condition. Adapted fromParameters of future ATC Systems Re/atirrgr to A/rport Capacify/De/ay (Washington, D. C.: Federal Aviation Administration,June 1978), PP. 3.3, 3.5.

“capacity coverage curve” for any given airportcan be constructed.

An example of a capacity coverage curve isshown in figure 26. The highest hourly capacityof Boston Logan Airport is 126 operations perhour in Visual Flight Rule (VFR) weather. Thiscombination of highest capacity runway use andgood weather is available 40 percent of the year.Strong winds create crosswind componentswhich close some of the runways of that con-figuration, and hourly capacities continue todecrease as marginal weather and finally badweather cause restrictions in safely operating therunway system. There is a small percentage (2percent) of the year when poor visibility, ceil-ings, and snow completely close the airport.Notice that there is a wide variation in the hour-ly capacity from 126 operations per hour downto 55 operations per hour before the airportcloses. This is typical of many major airportswhere several runway combinations exist. Thiswide variation in hourly capacity prevents theestablishment of a single capacity value for theairport; instead, it will be variable depending onweather conditions.

It is difficult to foresee any capital investmentin runways or technological improvements toATC facilities which can completely eliminate

Figure 26.—Airport Hourly Capacity Varies StronglyWith Weather

(There is a 3 to 1 or 2 to 1 ratio between good weather/badweather capacities)

Capacity Coverage Curve—Boston Logan Airport

20t

0 10 20 30 40 50 60 70 80 90 100

Average percent of time

SOURCE: Robert W. Simpson, “Airside Capacity and Delay at Major U.S. Air-ports,” draft report prepared for the Office of Technology Assess-ment, U.S. Congress, Washington, D. C., October 1980.

this degradation of capacity with weather condi-tions. New runways can raise the overall level ofthe capacity coverage curve, but they do not

Ch. 6—Airport Capacity Alternatives . 105


Snow control at a terminal

prevent its degradation with weather. Some ofthe ATC improvements discussed later in thischapter attempt to improve overall capacity byreducing the gap between IFR and VFR perform-ances.

Airfield/Airspace Configuration

The capacity of an airport depends to a largeextent on the number of runways available andtheir interactions. For example, for a given traf-fic mix a particular runway can handle 65 opera-tions an hour in VFR conditions and 55 in IFRweather. The VFR capacity of two parallel run-ways, 2,500 ft apart, might then be 125 opera-tions per hour— twice the capacity of a singlerunway. Yet the IFR capacity of this two-run-way system would be more like 65 operations

per hour, because under IFR conditions runwaysless than 4,300 ft apart are considered “depend-ent” for purposes of landings—that is, an opera-tion on one prevents a simultaneous operationon the other. Similar safety restrictions applywhere runways converge or intersect with oneanother. Thus, not only is the capacity of eachrunway reduced during bad weather, but the ca-pacity of the airport is further reduced becausenot all runways may be fully used.

In the illustration in figure 27, the three run-ways could be used in several different ways,four of which are shown. Each of these combina-tions may have a different operating capacity,and each might be suitable for a different set ofwind, visibility, and traffic conditions. A largeairport like O’Hare might have 40 or 50 possiblecombinations of runway uses. The limitationimposed by the available runway system variesamong the top air carrier airports. ChicagoO’Hare has seven runways, Kennedy has five,and La Guardia has only two (La Guardia’sadditional short 2,000-ft runway can be usedonly for departures during good weather condi-tions). Yet the capacity relationship is not linear:La Guardia manages to handle 40 percent ofO’Hare’s total aircraft movements with less than30 percent of its runways. An adequate taxi-way/gate configuration is also needed in orderto support optimum runway usage. For in-stance, the La Guardia Airport capacity task

Figure 27.—Runway Configuration

SOURCE: Federal Awatlon Admlnlstration, Techniques for Determinmg Awport Alrside Capacity and Lle/ay, FAA-RD-74-124, June 1976,


force found that additional taxiways in one areawere critical to minimizing delays. This is be-cause space at gates was limited, and the addi-tional taxiways could be used to hold and se-quence departing aircraft during periods of con-gestion.

Aircraft Noise

Aircraft noise, especially the noise of jet air-craft, has made airports unpopular with theirneighbors. The greatest noise impact is usuallyin the areas just beyond the ends of the runways,where arriving and departing aircraft fly at lowaltitudes. If a high-noise area is occupied by afactory or a highway cloverleaf there maybe lit-tle difficulty, but such land uses as residences,hospitals, and schools are not compatible withthe amount of noise generated by an airport. Insome areas, ineffective or nonexistent zoningand land use controls over the years have al-lowed these incompatible land uses to occupyhigh noise impact areas near many airports. Thecourts have generally found that the airport op-erator is responsible for injury due to reducedproperty value, and owners of nearby prop-erty have been able to collect damages in somecases. In Los Angeles, the courts have recentlyawarded nuisance damages as well. In someareas, including Atlanta, St. Louis, and Los An-geles, airport operators have been required topurchase noise-impacted property and either useit as a buffer zone or resell it for a more compat-ible use.

One method for reducing noise is to introducequieter aircraft or, as many air carriers havebegun doing, to re-equip old aircraft withquieter engines. FAA has set standards for newaircraft that are much quieter than in the past,but noisy aircraft will remain in the fleet formany years. The increasing sensitivity of thepublic to noise may have offset much of the re-cent improvement.

FAA, at the request of individual airport oper-ators, has also developed operational proce-dures that reduce noise impact. For example, useof certain runways may be preferred, or pilotsmay be required to make approaches over lesssensitive areas, weather permitting. However,


Air use and land use

FAA has established very few mandatory noise-abatement procedures. Over the past few yearssome operators have conducted airport noisecompatibility and land use studies for use as abasis for their own noise planning. The newFederal Aviation Regulation, Part 150, requiredunder the Aviation Safety and Noise AbatementAct of 1979 (Public Law 96-193), providesoperators with guidelines for voluntary noise-abatement standards and establishes a standard-ized method for measuring noise exposure.

Many of these noise-control procedures havea negative effect on capacity, and airports withboth capacity and noise problems have foundthat the available solutions to one problem oftenaggravate the other. The highest capacity run-way configuration, for instance, may be onewhich requires an unacceptable number offlights over a residential area. Enforcing noise-abatement procedures may also cause an unac-ceptable level of delay at peak hours. Thus, air-ports must balance tradeoffs between usable ca-pacity and environmental concerns.

The FAA Administrator recently reempha-sized that the responsibility for establishingproper land-use controls around airports restswith local government. He also predicted that


more communities will be establishing localnoise limits by ordinance or statute. f

A local government, whether or not it is theowner of the airport, can exercise some controlover noise, but must do so in a manner that isnondiscriminatory and does not place an undueburden on interstate commerce. For example, acity may select a reasonable noise exposure limitand exclude or fine aircraft exceeding that limit.However, the total ban on jet aircraft in SantaMonica, Calif., was overturned by the courts asunduly discriminatory against one class of air-craft (some new jets are quieter than propeller-driven aircraft).

ATC Equipment and Procedures

Improvements in aircraft surveillance, naviga-tion, and communication equipment over thepast decade have greatly increased the ability ofpilots and controllers to maintain high capacityduring all weather conditions (see ch. 5). How-ever, there are still ATC-related limits on airportcapacity. Clearances used in the en route air-ways and the terminal airspace are frequentlycircuitous, routing aircraft through intermediate“fixes” or control points rather than allowingthem to travel directly from origin to destina-tion. While this places aircraft in an orderly pat-tern so that controllers can better handle them, italso reduces capacity and consumes time andfuel.

“’Helms Places Airport Noise Problems on Operators, Commu-nities, ” Alliatiou Daily, Sept. 29, 1981, p. 154.

The limitations in the accuracy of surveillanceequipment also can influence how airports areconstructed and how they may be used. For ex-ample, the spacing requirement between inde-pendent IFR runways was developed based onthe limitations of surveillance, navigation, andcommunications equipment. Improvements inequipment and procedures have allowed thisminimum to be reduced over the years.

Constraints on capacity can arise whenairspace near one airport must be reserved toprotect operations at another airport. This is anespecially pressing problem in some busy areas.There is such an airspace conflict between LaGuardia and Kennedy in certain weather condi-tions, for example.

Demand Considerations

The daily pattern of demand is characteristicof the airport and the travel markets it serves.Air travelers prefer to travel at certain times ofthe day—midmorning and late afternoon, forexample—and air carriers wish to accommodatethem. Heavy scheduling at peak hours makes iteasier for passengers to transfer to other planesor other airlines, yet (as will be discussed short-ly) peaks in demand can be major causes of de-lay. Even at airports with a high percentage ofscheduled traffic it is not possible to predict theactual number of aircraft which will appear at aparticular hour of a given day, as nonscheduledtraffic volume can vary substantially. At quotaairports, the quota is set at a value between theVFR and IFR capacity, resulting in a built-indelay situation whenever weather conditions de-teriorate.

DELAY AND DELAY REDUCTION

Airport delays received a great deal of public- peaks, there may be delays even when the num-ity during the late 1960’s and they continue to be ber of aircraft using the airport is less than thea major waste of time, money, and fuel. Delay capacity for that peak time period. Somecan be expected whenever instantaneous traffic amount of delay arises every time two aircraftdemand approaches or exceeds the airport’s are scheduled to use a runway at the same time.capacity. When traffic occurs in bunches or The probability of simultaneous arrivals in-

108 . Airport and Air Traffic Control System

creases rapidly with traffic density, so that aver-age delay per aircraft increases exponentially

well before traffic levels reach capacity levels.

A typical variation of delay with operationrates is shown in figure 28. When the traffic levelis above capacity, the accumulation of aircraftawaiting service is directly proportional to theexcess of traffic over capacity. For example, ifthe capacity of a runway system is 60 operationsper hour and traffic rates are averaging 70 opera-tions per hour, then every hour will add an aver-age of 10 aircraft to the queues for service, and10 minutes to the delay for any subsequent ar-rival or departure. Even if the traffic level dropsto 40 operations per hour, delays will persist forsome period since the queues will be depleted ata rate of only 20 aircraft per hour.

The principal delay-reporting systems of FAAcurrently measure only the occurrence of largedelays. The National Airspace CommunicationsSystem (NASCOM) delay reports record in-stances of delays of 30 minutes or more at 46participating airports. The Performance Measur-ing System (PMS) records delays of 15 minutesor more at 15 major airports. The PMS also at-tempts to estimate “average delay per aircraftdelayed.” Both NASCOM and PMS rely on con-troller’s manual recording of instances andcauses of delays during periods when he is al-ready busy. Weather is listed as the primarycause for these delays, ranging from 76 percent

Figure 28.—Typical Distributions of Delay

50 percent 100 percent

Traffic rate capacity

SOURCE: Office of Technology Assessment.

of the 30-minute delays in 1976 to 84 percent in1979 in the NASCOM system. The total numberof delays reported also increased, from approxi-mately 36,200 in 1976 to approximately 61,600in 1979. It must be emphasized that whileweather may indeed be the primary cause, theability of the system to anticipate, adjust to, andrecover from weather-related problems is de-pendent on a number of the other determinantsof airside capacity.

Another major delay-reporting system issponsored by FAA and three airlines—Eastern,United, and American—which have been pool-ing their operational flight-time data since 1976.This Standard Air Carrier Delay ReportingSystem (SACDRS) covers 36 airports and meas-ures taxi times, gate holds, and flight timesagainst standard values in an attempt to deter-mine delay. Unfortunately, an error in thismethod causes an overestimation of delay: forexample, the standard times used for taxi in andout are based on the average over all runways ata given airport, but at some airports there iswide variation in taxi times for different run-ways and terminals; some percentage of theselonger taxi times are always counted as delayunder the SACDRS. FAA recognizes the defi-ciency in this system, but no correction has yetbeen devised. Estimates of the annual cost ofdelay based on SACDRS have ranged as high as237 million gallons of fuel and $273 million ofadditional operating costs to the three airlinesinvolved, although these costs too are overesti-mated. 2 The PMS and NASCOM systems, onthe other hand, because they only count long de-lays, probably underestimate delay. The truevalue of delay lies somewhere in between andhas not been determined with accuracy. Thus,estimates of the cost of delay based on any ofthese reporting systems have to be viewed withsome caution. However, all observers agree thatdelay is a serious and expensive problem at someairports, especially in light of the high cost offuel in recent years.

One method of dealing with delay is to con-strain traffic to manageable levels. This is the

‘Virginia C. Lopez (cd.), Airport and Airway Congestion, A Se-n“ous Threat to Safety and the Growth of Air Transportation(Washington, D. C.: Aerospace Research Center, July 1980).


origin of the quota systems which have been im-posed at a few major airports. Each carrier hasrepresentatives on the “scheduling committees”to negotiate the carrier’s share of allowed peakhour operations. FAA, through its flow controlcenter, also works to ameliorate the costs ofdelays by forewarning air carriers when delayconditions develop at major airports. For ex-ample, when weather deteriorates and capacitygoes down in Chicago, FAA may advise aircraftscheduled into Chicago to delay their arrival

there by waiting on the ground at other cities.Waiting on the ground is much less wasteful offuel than waiting in holding patterns in the air.

Although the lengthy delays of the late 1960’sare no longer typical, delay remains a majorproblem at many airports. Further, the numberof operations will increase as air traffic grows,and additional airports may experience thisproblem. Some possible approaches to dealingwith delay are discussed below.

DEMAND= RELATED ALTERNATIVES

Delay problems tend to be concentrated at theNation’s major airports, and even at these loca-tions the problem is most acute during certainhours of the day (usually midmorning and lateafternoon). If operations could be shifted fromthese peak hours to less busy times, delay couldbe reduced and the overall capacity of the air-port better utilized. Variable user fees or quotasduring peak hours are tools which have beensuggested, and tried at some locations, to reducepeak demand and increase operations in non-peak hours. All these mechanisms,duce the ease of transferring fromanother at hub airports, makingachieve ideal airline economics.

however, re-one flight toit harder to

Peak-Hour Pricing

Most airports now charge a landing fee basedon the weight of the aircraft. This fee schedule isdesigned to recover construction and operatingcosts of the airport, not to ration capacity. How-ever, when the use of an airport is nearing ca-pacity it could be more economically efficient tobase landing fees on the marginal costs imposedby each additional aircraft served. This meansthat the user should pay not only for use of theairport, but for the delay caused other users whowant to use it at the same time. This methodallows users who value access to the airport atpeak times to pay for their preference; thosewho do not wish to pay the higher fee would usethe airport at other times, or perhaps useanother airport.

In general, peak-hour pricing would have lit-tle effect on air carrier operations unless theprice changes are very large. Airlines scheduleflights when they think passengers will want tofly, and they would probably be willing to ab-sorb moderate increases in user fees in order touse the airport at those times. Even a landing feeof several hundred dollars would be small com-pared to the total operating costs of a largejetliner, and such an expense could be passed onto the passenger by a relatively small increase infares. Commuter air carriers, with their smallernumber of passengers, would be unable to paylanding fees quite as high as the larger carriers.

General aviation (GA) users on the otherhand, especially student and personal flyers, aremore sensitive to increases in landing fees. ThePort Authority of New York and New Jersey’s1968 decision to increase minimum landing feesfrom $5 to $25 during peak hours brought aboutan immediate decline of about 30 percent in GAoperations during peak hours at its three air car-rier airports (JFK, La Guardia, and Newark) anda noticeable decline in aircraft delays of 30 min-utes or more.3 In 1979, a $50 surcharge added topeak-hour landing fees at Kennedy and La Guar-dia resulted in a further decrease in GA traffic atthose airports.4 The remaining GA users were

3Airport Quotas and Peak Hour Pricing: Theory and Practice(Washington, D. C.: Federal Aviation Administration, 1976), pp.54-60.

“Port Authority of New York and New Jersey, Aviation Depart-ment, interview, Oct. 23, 1981.


primarily high-performance turboprop aircraftused for corporate travel; corporations, like theairlines, may be willing to absorb a fairly largeincrease in fees in order to use specific airportsduring peak hours.

One problem with a peak-hour pricing systemis that it is difficult in practice to determineprecisely what the marginal cost of an airportoperation is; several years of trial and errorwould be necessary to settle on a pricing schemewhich both controlled delay and allowed the air-port to cover its costs. However, if the same feewere charged to air carrier, commuter, and GAaircraft, peak-hour pricing might be stronglyresisted. Proportionately different fees for dif-ferent categories of users might therefore benecessary.

QuotasAn alternative method for managing demand

is to set a quota on the number of operationswhich can take place during a peak hour. Thequota can be placed on total operations, or acertain number of operations can be allocated todifferent classes of users. The quota levels areusually set between the IFR and VFR capacity ofthe airport; thus, in VFR conditions, additionalaircraft could easily be accommodated. Whencapacity is reduced, users without reservationshave to use the airport at another time or useanother airport.

Although reservations (slots) for GA or evenair taxis might be allocated on a first-come, first-served basis, slots for scheduled carriers presenta more complex problem. At major airportswhere quotas have been in effect for some time(O’Hare, JFK, La Guardia, and Washington Na-tional) representatives of the air carriers are al-lowed (with antitrust immunity) to meet asscheduling committees to negotiate how manyslots will be allocated to each carrier. Althoughnew entrants are able to participate in these ne-gotiations, quota systems do tend to favor thestatus quo. Since the air traffic controllers strikein August, 22 airports have been brought undera quota system designed principally to easepeaks of demand on the en route ATC system.The methods for assigning slots to new entrants

or allowing existing carriers to exchange slots arestill under development.

One objection to quota systems is that the al-locations are made without any price signals toshow that the capacity is being used efficiently.Thus, although the quota may provide somestop-gap congestion relief, it does not provideany long-run guide for allocating resources asthe system grows or changes. It has been sug-gested that this problem could be overcome byauctioning the reservation slots among the carri-ers or by combining the quota system with somesort of peak-hour pricing scheme.

Balanced Use of MetropolitanArea Airports

Many major metropolitan areas are served bytwo or more large airports. Where one or moreof these airports is underutilized, possibilities ex-ist for increasing airside capacity through a morebalanced use of the region’s airports. Examplesinclude: Newark Airport, which is underutilizedcompared to Kennedy and La Guardia; OaklandAirport, which could relieve San Francisco;Midway Airport, which is practically emptywhile Chicago-O’Hare has delay problems; andBaltimore-Washington and Dunes Airports,which might relieve Washington National. Theproblem of balancing use of metropolitan air-ports presents a chicken/egg dilemma: airlineswon’t serve the underutilized airport becausethere are so few passengers, and passengersdon’t go there because there is so little service. Itis difficult to foresee when congestion in itselfwill become great enough to cause redistribu-tion, or to what extent the process can or shouldbe managed by local or even Federal authorities.In some cases, better transportation between air-ports might make it easier to transfer betweenflights and to attract passengers to underutilizedairports.

The Washington, D. C., area is illustrative ofthe problems of imbalance airport use. Wash-ington National Airport, operating since themid-1940’s, is convenient to the downtown area.National has three runways (all under 7,000 ft)and does not accept wide-body jets. Both its air-


side and landside capacity are severely limitedand a quota system and airline scheduling com-mittee are used to ration peak-hour operations.Expansion is difficult due to surrounding devel-opment and the Potomac River. Complaintsabout the airport’s noise have led to a 10 p.m.curfew among other noise abatement policies.From time to time some groups even call for theairport to be closed.

Many of these problems could be alleviated ifsome operations were transferred to Dunes In-ternational Airport, 26 miles from Washington.Dunes, opened in 1962, has two 11,500-ft run-ways, one 10,000-ft runway and capacity tospare. FAA (which operates both airports) hasrepeatedly attempted to induce carriers to useDunes more; for example, only Dunes canreceive international and long-range domesticflights. Despite the constraints of the quotasystem, the curfew, and the restrictions on wide-body and long-range flights, however, Nationalhandled nearly 4 times the operations and 4½times the passengers that Dunes did in 1980. Fur-ther, National generated a net profit of $10 mil-lion that year, while Dunes incurred a net loss of$3 million. ’ The principal problem is ground ac-cess; it is more convenient to fly from Nationalthan from Dunes.

Some new airlines beginning service since de-regulation have sometimes deliberately chosento operate out of underutilized airports to avoidcongestion and delay. One example is MidwayAirlines, which uses the nearly abandoned Mid-way Airport for its Chicago service. Midway’sproblem is also related to ground access: con-gested highways make trips to the airport longeven though Midway is closer to downtownChicago than O’Hare. Another example is Peo-ple Express, which serves the New York areafrom Newark. The Port Authority of New Yorkand New Jersey has been offering incentives topassengers as well as airlines to increase the useof Newark Airport: improved ground access bytrain and express bus allows New York City pas-sengers to get to Newark without paying high in-terstate taxi fares, and new airlines are offered

51nterviews, FAA, Metropolitan Washington Airports, July 6,1981.

more and better space for future growth atNewark. In addition to People Express, NewYork Air has located part of its operation atNewark. Now that permission has been gainedto use Newark as a international airport, severalestablished airlines are also bidding to offertransatlantic service from there.

Restructuring Airline Service Patterns

When delay becomes intolerable at busy hubairports, users themselves may voluntarily movetheir operations to another facility. This move-ment might be to an underutilized airport near-by (e.g., Newark), but it could also be to a medi-um or small hub located at some distance fromthe congested hub. This is especially likely fortransfer traffic. (See ch. 4 for a discussion of thegrowth and capacity impacts of this redistribu-tion scenario. )

Many major airports currently serve as hubsfor a large amount of transfer traffic. Three-fourths of the arriving passengers at Atlanta andabout one-half the passengers at O’Hare, Dallas-Fort Worth and Denver pass through these air-ports only to change planes for somewhere else.Carriers choose to establish their hubs at thesebusy airports so that passengers can choose frommany transfer flights. However, when the trans-fer airport becomes too congested the disadvan-tages of delay may begin to outweigh the advan-tages of convenience, for airlines as well aspassengers. Hence carriers may decide to locatetheir new transfer operations, and even movetheir existing hubbing activities, to other citiesthat have more room for growth.

Redistribution of operations appears to be oc-curring under the new routing freedom availableunder the Airline Deregulation Act of 1978. Car-riers are finding it easier to change their routesand establish new “second-tier” hubs at less con-gested airports. Between 1978 and 1980 the num-ber of large hubs (handling more than 1 percentof total U.S. passenger traffic) fell from 26 to 24,while the number of medium hubs (handling0,25 to 0.99 percent) increased from 33 to 36—a

market shift reflecting the distribution of opera-tions over more airports. This trend may accel-erate as regional carriers modify their patterns of


service, and even the busiest airports such asAtlanta and O’Hare, may see actual declines inboth enplaned passengers and operations in thenext 10 years. A similar decline in operations oc-curred at Kennedy Airport when internationalflights were allowed to enter the United States atother gateway cities.

Reliever AirportsIn metropolitan areas where there is conges-

tion at the main airport and excess capacity atsurrounding airports, diversion of GA trafficwould be effective in improving the use of air-side capacity in the whole region. It would allowa higher level of service for both air carrier andgeneral aviation, and in most metropolitan areasthere are smaller airports which might potential-ly attract some GA traffic away from the mainairport. For example, FAA lists 27 airports in theChicago area, 51 around Los Angeles, and 52 inthe Dallas-Fort Worth metropolitan area. How-ever, most of these airports are quite small, andonly a few have runways long enough to accom-modate business jets or instrument landingequipment for bad-weather operations.

The FAA’s National Airport Systems Plan(NASP) designates 155 airports as “satellites” or“relievers” to major airports, and NASP pro-vides for separate Airport Development AidProgram (ADAP) funding to be set aside for re-lievers. Publicly owned reliever airports may useADAP funds for construction, installation ofsafety equipment, and other eligible expendi-tures. The 25 or so privately owned reliever air-ports, although they presumably provide thesame benefit in terms of diverting traffic fromcongested air carrier airports, are not eligible foraid. Local and State governments may, how-ever, use ADAP funds to help purchase private-ly owned reliever airports, and at least fivereliever airports have changed from private topublic ownership since 1973. One privatelyowned reliever, Chicagoland (a reliever forO’Hare) closed in 1978. Although the FAA re-liever program was initiated largely to segregatetraining activities from major commercial air-ports in the interests of safety, it also providesadditional airport capacity for a certain type oftraffic—namely, personal GA aircraft with ori-

gins or destinations in the local region; businessand commercial GA (i.e., corporate aircraft andair taxis) delivering or picking up airline passen-gers will probably continue to use the majorcommercial airport.

The process of diverting the personal GA traf-fic has already occurred at the Nation’s largestmajor commercial airports. The fraction of GAactivity at Atlanta, O’Hare, Kennedy, Los An-geles International, etc., is very small (about 10percent) because these regions have good alter-nate secondary airports with high levels of traf-fic. In fact, some of the large relievers such asVan Nuys and Long Beach, Calif., Opa Locka,Fla., and Teterboro, N. Y., are among the busiestairports in the country in terms of annual opera-tions. This trend toward establishing a system ofreliever airports is underway and has been en-dorsed by many user groups and observers,most recently the President’s Task Force on Air-craft Crew Complement. e

To be of maximum benefit the reliever airportshould be located so that approach and airspaceconflicts between the reliever and the commer-cial airport do not place capacity limits on both.In the New York area, for example, instrumentoperations at Linden and Teterboro reliever air-ports must alternate with operations at the New-ark Airport. In addition, the noise consequencesof increasing operations at the reliever airportmust be considered. Most reliever airports have,or will soon have, IFR landing aids and runwaysystems capable of handling sophisticated GAaircraft. To be most attractive to users, airportsshould also have commercial services for aircraftservicing, repair and maintenance, groundtransportation, and flight crew amenities. Withsufficient amenities, such an airport might evenattract some commuter airline service, althoughtransfers and interlining would be difficult un-less the airport is served by several carriers orhas excellent ground access to a major hub. Insome cases, however, the provision of better fa-cilities may not be sufficient to divert additionalGA traffic away from major hub airports. In-creased landing fees at the major airport can

‘Report of the President’s Task Force on Aircraft Creu) Comple-ment (Washington, D. C.: July 2, 1981).


provide additional incentives for this shift, looked upon as a complement to the Federal pro-and such pricing policies—the domain of local gram of investment in satellite airports.government and airport authorities—could be

AIRPORT DEVELOPMENT ALTERNATIVES

Expanding Existing Airports

Because runway availability is the major con-straint on airside capacity, one way to increasecapacity is to add more runways. A new longrunway, properly equipped for independent IFRoperations can increase an airport’s capacity by20 to 50 percent depending on the original run-way configuration.

Adding another runway, however, requires alarge amount of land. One 11,000-ft runway forlarge jet operations with its basic safety areascovers 130 acres, and when other necessary“clear zones” are considered, an area three tofour times that size would be directly affected.Further, the additional operations enabled bythe new runway would probably require land-side additions such as new gates, terminal space,and parking for more passengers. Few airportshave the necessary land for this kind of expan-sion, which could add approximately 10 percentto their present area, and for some airports likeWashington National and La Guardia, the pros-pect is especially bleak. Even for larger airports,obtaining proper spacing from other runwayswould be extremely difficult.

A 1977 report by the Department of Trans-portation (DOT) studied the possibility of majorexpansion at 24 airports to meet projected needsfor 1985-2000. Expansion was found to be “feasi-ble” in only four of these cases, and none ofthese four airports (Detroit, Houston, Minne-apolis, and Pittsburgh) are among those whichare experiencing the greatest capacity problems.In 9 other cities the DOT study found expansion“feasible within major constraints,” and in 11cases it was considered “not feasible.” Botheconomic and environmental reasons were citedfor preventing the land acquisition. ’ Airport de-

‘Establishment of New Major Public Airports in the UnitedStates (Washington, D. C.: Federal Aviation Administration,August 1977), p. 6-5.

signers foresaw the need for growth and mostmajor airports were built where land was plenti-ful, but sites that were on the edge of town in1925 or 1948 are now in the middle of urban de-velopment. In some cases the airport itself at-tracted businesses; in other cases developmentsimply resulted from good highways, suburbani-zation, and all the other forces which havecaused urban areas to expand over the years.Developed land tends to be expensive to buy: arecent study of the cost to acquire and clear landaround some major air carrier airports estimatedthese costs at between $100,000 and $200,000per acre.8 Noise is among the largest environ-mental obstacles to airport expansion. Chicago-O’Hare has sufficient land for an additional run-way, but the runway has not been built in partbecause it would cause unacceptable noise expo-sure in nearby neighborhoods. JFK Airport inNew York is surrounded by intensive develop-ment on one side and a National Park and Wild-life Sanctuary on the other, making expansionunlikely. Dallas-Fort Worth, on the other hand,is planning an additional major new runwaythat is expected to ease some of the capacity lim-itations imposed by noise abatement proceduresand airspace conflicts with nearby Love Field.

Development of SecondaryRunway Operations

At some airports where major expansion isunlikely it may still be possible to add one shortrunway for smaller, slow-moving commuter andGA aircraft. This could improve airport capaci-ty by diverting traffic from the longer runwaysand may also provide a partial solution to thewake vortex problem (previously discussed).Many airports routinely use short runways, orsections of long runways, for small aircraft dur-

t‘Louis H. Mayo, Jr., “Noise Compatible Land Uses in Airport

Environments, ” Environmental Comment, March 1979, p. 9.


ing good weather, but because of inadequatelanding aids or spacing these runways cannot beused during bad weather; all-weather operationswould require additional navigational and ap-proach guidance equipment.

One study found that the use of short IFR run-ways for small aircraft was feasible at 11 of 30major airports. Of these 11, suitable runways al-ready existed at 3 airports, existing runwayscould be extended for use at 2 others, and at 6airports space was available for short runwaysto be constructed. The study estimated that thevalue of reduced delays brought about by theaddition of such runways might be $450 millionto $810 million in current dollars between 1980and 1990 at the airports shown in table 10. Thebenefits would be unevenly distributed: Chi-cago, Atlanta, Philadelphia, and Denver wouldreceive 80 to 85 percent of the estimated savings;among the users, 86 to 89 percent of the savingsfrom reduced delays would accrue to the air car-riers. 9

A detailed study of the airfield and airspace ateach airport would be needed to see if the shortrunway could really be constructed. Such stud-ies done at Denver revealed two possible loca-tions for a short GA runway. Construction ofeither one could lead to a 35 to 70 percent in-crease in hourly operations, depending onweather conditions. Total cost was estimated atabout $10.8 million. 0

Building New Airports

Another way to increase airport capacity is tobuild a completely new airport to replace or sup-plement the existing one, an alternative that isespecially attractive where landside facilities(terminals, baggage equipment, parking) arealso outmoded or inadequate. A new site wouldprovide the opportunity to design and build run-ways, terminals, and parking space to meet fu-

‘John D. Gardner, Feasibility of a Separate Short Runway forCommuter and GeneraZ Aviation Traffic at Denz~er, prepared forthe Federal Aviation Administration by The Mitre Corporation,McLean, Va., May 1980, pp. 1-1.

‘“John D. Gardner, Extensions to the Feasibility Study of a Sep-arate Short Runway for Commuter and General Auiation Trafficat Denuer, prepared for the Federal Aviation Administration byThe Mitre Corp., McLean, Va., September 1980, pp. 4-3 and 7-1.

ture needs, rather than making do with what hasevolved over time. Sufficient land could be pur-chased to allow for future growth and properland-use controls could be applied so that noisecompatibility problems do not arise again. Insome recent airport relocations, however, thisdid not work as well as hoped. For example, atboth Dallas-Fort Worth Regional Airport andKansas City International Airport, built in themid-1970’s, encroachment by other land uses isagain leading to complaints about airport noise.On the other hand, Montreal’s new Mirabel Air-port seems to have little problem with noiseincompatibility; the airport itself covers 17,000acres, and is surrounded by an additional 21,000acres controlled by a specially created municipalauthority. However, its distance from the citymakes access a problem.

Building a new airport also provides an op-portunity to add a large amount of new airsidecapacity to a region. The opening of Kansas CityInternational, for example, more than doubledthe available capacity in that hub from the esti-mated 195,000 operations at the old municipalairport to about 445,000 with the new airport.Love Field in Dallas handled 410,000 operationsin 1972; in 1977, after air carrier operations weretransferred to Dallas-Fort Worth Regional Air-port, Love Field still had 310,000 operations(mostly GA), while the new airport had 385,000.

A 1977 investigation by DOT found that any-where from 2 to 19 new airports might be neededin the United States by the year 2000, dependingon the growth rate assumed. When the studyexamined the feasibility of new airport construc-tion for 10 hub areas, it found it to be “feasible”in four instances, “doubtful” in four, and “notfeasible” in two. The reasons for the “doubtful”and “not feasible” findings are related primarilyto site location, land acquisition, funding prob-lems, and the difficulty of providing adequateground access to a remote location. The FAA’s1980 NASP foresees the possibility of a new air-port opening at Palmdale, Calif. (near Los Ange-les), within the next 10 years; some initial workon new airports at Atlanta and San Diego mightalso be expected within the next decade. ’ 2

‘ ‘Establishment of New Major Public Airports, op. cit., p. 7-16.“National Airport System Plan, Revised Statistics 1980-1989

(Washington, D. C., Federal Aviation Administration, 1980) p. vi.

Ch. 6—Airport Capacity Alternatives • 115

Table 10.—Operational Characteristics of Airports With Potential Benefits From aSeparate General Aviation Runway

Parallel Parallel Nonparallelindependent dependent dependent

Modification operations operations operations

New runway . . . . . . . . . . . Chicagoc, Atlantac, Philadelphia,Dallas-Ft. Worthc, Denver Pittsburgh ab

Existing runway Portland,or taxiway. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Detroit a,b St. Louis

Extension of New York (JFK)a,Existing runway. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Indianapolis

aThe ~enera] aviation runway is Independent of 1 of 2 air Carrier runways for departures.bGeneral aviation runway handles departures OnlYc_friple parallel runways.

SOURCE: J. D. Gardner, “Feasibility of a Separate Short Runway For Commuter and General Awatlon Traffic at Denver, ”prepared for the Federal Aviation Administration by Mltre Corp., McLean, Va., May 1980.

Building a new airport is a huge undertaking.A new air carrier airport can represent an invest-ment of $5 billion, to be shared among the air-port sponsor (and local taxpayers), airport con-cessionaires, the airlines (through their landingfees), and the Federal Government. Even a mod-est-sized GA airport would cost several hundredmillion dollars. The length of time required forplanning and construction of a large airport—up to 10 years—can also add substantially tocosts. Political and institutional factors can alsopose substantial difficulties. Building an airportrequires agreement from existing air carriers tomove to the new facilities, but while a new air-port can reduce delays it will also increase airlinecosts, and they must be convinced that the bene-fits will outweigh the costs. Further, approvaland support of a number of State, county, andmunicipal governments, not to mention high-way districts, zoning commissions, and variouscitizens’ interest groups, must also be secured.

In some cases the divergent interests of dif-ferent governments and constituencies can snarlthe process. In St. Louis, for example, a site for anew airport was selected across the MississippiRiver in Illinois. The Illinois State governmentwas a major supporter of the project, as were theSt. Louis city government and FAA. The oppo-nents included citizens groups of the countywhere the new airport would be located (whoobjected on environmental grounds), the Stateof Missouri (which did not want the airportmoved out of the State), and groups in St. Louis

(which did not want the city to give up the close-in Lambert Airport). The project was debatedfor several years, but it was shelved after achange in the St. Louis city government.

Photo credit: Federal Aviation Adrninistration

The design of a modern airport: Dallas-Fort Worth

116 • Airport and Air Traffic Control System

ATC IMPROVEMENT ALTERNATIVES

As mentioned earlier, existing ATC proce-dures and equipment can represent constraintson the airside capacity. Improvements in theseareas can increase the number of aircraft opera-tions.

Airfield/Airspace ConfigurationManagement

The ATC team at an airport decides how therunway and ATC equipment should be usedbased on wind, visibility, traffic mix, ratio of ar-rivals to departures, noise-abatement proce-dures, and the status of the airport (which runways or landing aids are under repair, etc.). In alarge air carrier airport like O’Hare, there maybe 40 or 50 ways in which the runways can beused, so deciding which one offers maximum ca-pacity for any particular set of conditions is acomplex task. The problem is compounded bythe interdependence of runway use and the con-figuration of the surrounding airspace. For ex-ample, changing which runway is used for land-ings may change the route that approaching air-craft must take through the terminal airspace,which may in turn affect or be affected by activ-ity at other airports in the vicinity.

One FAA analysis of capacity and delay prob-lems in Chicago suggested that proper manage-ment of airfield and airspace could have a largepayoff:

Optimized management of the air traffic con-trol system . . . could achieve now, at mini-mum investment cost, savings comparable tothose that will be achieved much later at muchhigher cost when third generation ATC hard-ware is deployed. This highlights the impor-tance of FAA management exploration of op-portunities for improved system efficiency byplacing emphasis on optimization of operationsat least equal to that given development of ATChardware. 13

After study of the runway system of O’Hare air-port, the task force found that a computerized

“Delay Task Force Study, Volume 1: Executive Summary,O’Hare International Airport (Chicago: Federal Aviation Admin-istration Great Lakes Region: July, 1976), p. 4.

airspace/airfield management system could beused to assist the controller team in selecting thehighest capacity and most energy-efficient run-way use for each set of circumstances.

Such a system could have several levels ofcomplexity. In its basic form it would aid in se-lecting the preferred runway configuration for agiven set of conditions; this basic system isunder development by FAA. The intermediateform would update this assessment as changes inweather or traffic conditions arise, and thenselect the most efficient means of making thetransition from the one configuration to an-other. (This is important because the transitionperiod is often a time when airspace and airportcapcity are wasted. ) The advanced versionwould have the ability to make longer term stra-tegic decisions. The 1978 Chicago task force sug-gested that savings of $11 million to $16 millionannually in reduced delay costs might be ex-pected from the basic system alone.14

Wake Vortex Prediction

Alleviation of the wake vortex problem offersthe possibility of a substantial potential payoffin increased capacity without large capital ex-penditures for new runways. Research over thepast decade has shown some possible ways ofdoing this. For example, it has been found thatcertain wind conditions can quickly dissipate avortex or remove it from the path of oncomingtraffic. If wind conditions can be accuratelymonitored and quickly analyzed, then the likeli-hood of wake vortex danger can be known on aminute-by-minute basis.

FAA has been testing such a system at O’HareAirport since 1977. Wind sensors are located on50-ft towers near the runway ends. A computeranalyzes wind conditions and when persistentvortexes are unlikely it gives the controller teama “green light” to permit reduced separations onfinal approach. To have maximum effect (e.g.,to allow all separations to be reduced to 3 nmi),an advisory system would have to be able to————

“Ibid.


predict the likelihood of wake vortexes atgreater distances and higher altitudes than theChicago system now does. However even thisprototype system has been credited with allow-ing reduced average separations, and thus moreoperations per hour, at O’Hare. There are nocurrent FAA plans for implementing full scalewake vortex advisory systems at other airports.

Microwave Landing System (MLS)

As discussed in chapter 5, MLS allows air-space to be used more efficiently than the cur-rent ILS, since aircraft would be able to ap-proach the airport on curved paths, as they dounder visual conditions, and turn onto theirfinal approach much closer to the runway.”Variable MLS glide slope angles could also pro-vide a partial solution to the long separationsrequired to avoid wake vortex; with MLS, thetrailing aircraft could avoid the vortex by ap-proaching the runway at a steeper angle than thelead aircraft.

Models suggest that where the traffic mix con-tains a variety of fast and slow aircraft, the useof variable glide slopes could allow some capaci-ty improvements—perhaps around 10 to 15 per-cent. However, where aircraft have similar per-formance characteristics, MLS landing proce-dures would offer about the same capacity ascurrent ILS procedures. MLS would also allowthe restructuring of airspace at some airports, sothat small aircraft can approach the airport in aseparate arrival stream from jets and make useof a separate short runway. The Dash-7 aircraftin Ransome Airlines’ Washington-Philadelphiaservice use MLS equipment to land on short run-ways.

MLS equipment has been developed, tested,and accepted for international use. Field evalua-tion is taking place at such airports as Washing-ton National, and FAA has published a plan forfull-scale implementation beginning in themid-1980’s and to continuing into the next cen-

“An Analysis of the Requirements For and the Benefits andCosts of the National Microwave Landing System, Volume 1(Washington, D. C.: Federal Aviation Administration, June 1980),p. 2-3.

tury. One reason for this delayed schedule maybe the international agreement to maintain ILSuntil 1995; and another reason is the reluctanceof users, principally the airlines, to install MLSavionics in aircraft already equipped with ILSavionics.

Reducing Separation orSpacing Minimums

Several studies have suggested that wherewake vortex is not a problem (for example,where aircraft have similar performance charac-teristics) it maybe possible to reduce separationsfrom 3 nmi to as little as 2.5 or 2 nmi. Theamount of time each aircraft spends on the run-way is another constraint in reducing separa-tions, and depends on such factors as the num-ber and spacing of the exits, visibility, runwaysurface conditions, and the performance charac-teristics of the aircraft. In general, small, lightaircraft spend less time on the runway thanlarge, heavy ones. According to surveys, mostairports have an average runway occupancytime of between 41 and 63 seconds for landing,although these figures do not include the raresnowy or icy days when separations might haveto be extended to allow time for aircraft to brakesafely and exit the runway. Where the averagerunway occupancy time is so seconds or less, ithas been suggested that the minimum separationcould safely be reduced to 2.5 nmi instead of 3nmi. Greater reductions might be possiblethrough automated metering and spacing.

Another way of increasing airfield capacity isto reduce the required spacing between run-ways. For example, runways must be 4,300 ftapart for simultaneous IFR operations to takeplace. Reduction of this minimum to 3,500 or3,000 ft would enable some airports to make useof more of their runways during IFR conditions.Minimum spacing standards have been reducedbefore (e.g., from 5,000 to 4,300 ft for independ-ent parallel IFR runways in the early 1960’s) as aresult of improvements in surveillance equip-ment and procedures.

“William J. Swedish, Evaluation of the Potential for ReducedLongitudinal Sparing on Final Approach, prepared for the FederalAviation Administration by The Mitre Corp., McLean, Va., p.4-1.


FAA is also investigating the possibility of al-lowing instrument approaches to triple parallelrunways during poor visibility. Currently tripleparallels can be used only during good visibility.One of these three runways might be a shortrunway for commuter or GA aircraft. Efficientuse of triple parallels would require redesign ofthe airspace and approach patterns, a higherdegree of coordination between approach con-trollers than is currently the case, and possiblemodifications to the ILS. MLS, with its greaterflexibility and navigational precision, might beuseful in bringing this procedure into practicaluse. Use of triple parallels could make it possibleto make use of more existing runways duringpoor weather, as at O’Hare, or even to allowconstruction of new runways which are infeasi-ble under current procedures. Capacity im-provements would depend on traffic mix and onwhether the runways had sufficient spacing toallow independent operations. Models indicatethat triple parallel runway systems might handleup to 50 percent more IFR operations than dou-ble parallels with traffic mixes typical of today’smajor airports.l 7

A number of airports have been identifiedwhich might benefit from either reduced spacingstandards or from use of triple parallel ap-preaches.18 However, site-specific analyses ofthe airfield and airspace of each candidate air-port are needed to measure the capacity benefits,costs, and safety effects of these proposedchanges.

Automated Metering and Spacing

The controller’s ability to meter aircraft—todeliver them to a specific point at a specifictime—is based on aircraft speed and position asshown on the radar screen and the controller’s

‘ ‘T. N. Shimi, W. J. Swedish, and L. C. Newman, Requirementsfor lnstrurnent Approaches to Triple Parallel Runways, preparedfor the Federal Aviation Administration by The Mitre Corp.,McLean, Va., 1981, p, E-7.

‘“L. C. Newman, T. N. Shimi, and W. J. Swedish, Sur-ocy of 101U.S. Airports for New Multipfe Approach Concepts, prepared forthe Federal Aviation Administration by The Mitre Corp., McLean,Va., 1981, p. xxiv, 5-4, 6-2; and A. L, Haines and W. J. Swedish,Requirements for Independent and Dependent Parallel instrumentApproaches at Reduced Runway Spacing, prepared for the FederalAviation Administration by The Mitre Corp., McLean, Va., 1981,passim.

instructions to change speed or direction inorder to arrive at the runway threshold at theproper time. Using this manual system the con-troller’s training and experience allow him todeliver aircraft to the runway threshold with anerror (standard deviation) of about 18 seconds.l9

It has been suggested that an automated sys-tem could provide more accurate metering andspacing. In such a system, the ATC computercould analyze radar and transponder data di-rectly and compute future aircraft location withgreat accuracy, then generate commands de-signed to deliver each aircraft at a specific timeand thereby optimize the use of the runway’scapacity. It has been suggested that an auto-mated system could reduce the delivery error toabout 11 seconds.20 The automated concept hasbeen under development at FAA for about 10years but has not yet been approved for imple-mentation. FAA states that the computerizedmethods developed so far are not as reliable as ahuman controller. In addition, FAA believesautomated terminal metering and spacing willnot be of much value unless it can be tied in withen route metering and other aspects of ATCautomation now under development (see ch. 5).

Cockpit Engineering

Advances in technology are in fact changingthe basic character of the cockpit. Electrome-chanical instruments are being replaced withelectronic displays that present full-color imageswith a very high degree of resolution. Comput-ers are also expanding the range of functionsthat can be performed by aircrew. Advancednavigation aids such as area naviation (RNAV)make it possible to navigate from point to pointwithout following established airways. The FAAhas suggested the use of a data link to improvethe quality of the information available in thecockpit. A cockpit display of traffic information(CDTI), currently under investigation at the Na-

“New Engineering and Development initiatives—Polic y andTechnology Choices, coordinated by Economic and Science Plan-ning, Inc. (Washington, D. C.: Federal Aviation Administration,March 1979) p. 107.

20 Parameters of Future A TC Systems Relating to Airport Capac-ity/Delay (Washington, D. C.: Federal Aviation Administration,June 1978).


tional Aeronautics and Space Administration airport and airway facilities are used. Therecould show pilots the locations of nearby air- have been suggestions that the distribution ofcraft, thus reducing their dependence on ground the decisionmaking function in the ATC systemsurveillance. Both RNAV and CDTI offer pilots must or should be changed to take advantage ofsignificant independence from controllers, and the capabilities these technological advancesthis could increase the effectiveness with which have made possible (see ch. 5).

SUMMARY OF ALTERNATIVES

The alternatives discussed above all make use the other hand, rely on the application and en-of some combination of economic, regulatory, forcement of regulatory measures to deal withor technological tools to reduce delay or increase the delay problem. Automated metering andairside capacity. For example, peak-hour pricing spacing is a technological tool, but its use will re-is an economic alternative—allowing the market quire changes in existing rules and standards.to allocate scarce airport capacity. Quotas, on Table 11 summarizes the alternatives discussed

Table 11 .—Summary of Alternatives

EconomicAlternative incentives Regulation Technology Comments

Change of airlineservice patterns

Reliever airports

●

●

●

Demand-relatedPeak hour pricing ● Could be implemented by local airport authority.

Devising and managing the pricing scheme maybe complex, but it could provide a substantiallong-term payoff in reduced delay.

Quotas Could be implemented by local authority or FAA.Would provide some short-term relief for con-gestion and delay problems but is an inefficientlong-term solution. FAA has already imposedquotas at 4 airports since 1969.

Balanced use of ● Could be implemented by local authority whichmetropolitan airports might use economic incentives, improved ac-

cess, and better facilities to encourage use ofunderutilized airports; or could use regulation toimpose it.

Airlines may voluntarily shift some of theirhubbing activities to less congested airports tosave delay. (This trend seems to already beunderway.) The FAA might also be able toachieve this redistribution by regulation. Thiswould make better use of airport capacity na-tionwide, but might do little to reduce delays atcongested airports.

FAA has already designated reliever airports.Many are well used by GA traffic. Localauthorities encourage this trend with pricingstrategies, better facilities, or regulations requir-ing use of relievers by certain classes of users.Relievers have been and will continue to be suc-cessful in providing capacity for GA operationsaway from congested commercial airports.

Airport developmentAirport expansion ● Responsibility of local authorities, possibly with

Federal aid. Could greatly increase capacity,but is unlikely in many locations because of sur-rounding development or environmental prob-lems.


Table Il.–Summary of Alternatives (Continued)

EconomicAlternative incentives Regulation Technology Comments

Addition of short ●

runway

New airport construction

ATC alternativesAirfield/space

management

Wake vortexprediction

Microwave landingsystem

Reduced separationor spacing standards

Automated meteringand spacing

Cockpit engineering

●

●

●

●

●

●

●

●

●

●

●

●

Possible in several airports to provide a separatetraffic stream for GA and commuter aircraft.Increases capacity for both small and large air-craft. Responsibility of local authority withpossible Federal aid. Cost estimate for Denverwas $10 million to $11 million.

Responsibility of local authorities with Federalassistance. Could have a major impact on localairside capacity, but is unlikely in many areasdue to expense, lack of close in suitable land.Good high-speed ground access might makemore distant airports likely in long range.

Allows modest capacity gains by making betteruse of the runways available. Computerizedsystem has been tested in Chicago. Similarsystem could be developed and implemented inother areas by local authorities and FAA.

FAA would be responsible for installing vortexdetection or advisory equipment. FAA hastested one wake vortex advisory system whichprovides some capacity benefits, but is still inthe experimental stage.

Benefits are more efficient use of airspace andavailability of variable glide slopes which,among other things, can allow aircraft to avoidwake vortexes. Fairly substantial increases incapacity available where traffic mix is diverse.The technology now exists and FAA will prob-ably install ground equipment in the 1985-2000period. FAA’s installation costs are estimated tobe $300,000 to $500,000 per airport. Users costsfor avionics will range from $1,500 to $30,000per aircraft.

Responsibility of FAA. Reduction of thesestandards could offer large capacity increases,but FAA’s first priority is safety of the system.Reduction of standards is unlikely without sometechnological change—elimination of wakevortex problem or improved navigation orsurveillance.

increased accuracy of metering could optimizerunway use, offering modest capacity increases.FAA has not yet developed a program which itfeels ready to implement. FAA wants to in-tegrate terminal automated metering and spac-ing with the automated en route system, im-plementation might not be possible until afterthe replacement of the en route computersystem.

RNAV technology is already available. Users mustbuy the avionics, FAA is responsible fordeveloping RNAV procedures which mightreduce delays somewhat. Cockpit displays oftraffic information are being developed andtested by the FAA but will not be available inthe near future.

SOURCE: Office of Technology Assessment.


above and indicates generally what types oftools—economic, regulatory, or technological—would be required to implement them. The com-ments in table 11 touch on several points—whocan implement the change, whether it wouldmake a large or small change in capacity, andhow likely it is to take place in the short or longterm.

In general, the demand-related alternatives donot increase capacity; rather, they reduce delayby molding traffic activity to fit existing capaci-ty. Modest capacity gains are available throughATC improvements that increase the efficiencywith which airfield and airspace are used, espe-cially under IFR conditions, but the benefitavailable to each airport is heavily dependent on

local conditions of runway configuration andtraffic mix. The addition of new runways isclearly effective in increasing capacity, but thisoption is available to only a limited number ofairports. In a few cases, short runways could beconstructed to increase capacity by separatingjet and propellor traffic. New airport construc-tion also offers large capacity gains, but theywould likely be further from cities and thereforeface the problem of ground access. Reliever orsatellite airports to move GA out of air carrierairports are necessary unless the growth of bothuser groups is to be severely limited, but relieverairports will also be constrained by land prices,noise impacts, and community acceptance.

FUTURE RESEARCH NEEDS

Several areas offer possibly fertile ground forfuture research on means to increase airport air-side capacity.

Wake Vortex Avoidance

The FAA’s wake vortex advisory system hasbeen discussed, but more research is needed todevelop operational versions of this systemwhich can predict vortex problems at greaterdistances from the runway ends—say, back tothe ILS middle marker or outer marker. FAA hasalso studied the use of acoustical radar and lasersto detect actual vortexes. Although some prog-ress has been make in understanding the natureof vortexes, these techniques are far from opera-tional. However, with further research this lineof inquiry may be the basis for a ground-basedor airborne wake vortex detection system.

Wake Vortex Alleviation

Also important is the possibility of modifyingor minimizing vortexes at the source. NASA re-search has shown that certain combinations offlaps, spoilers, or protrusions on the wings ofaircraft can cause the wake vortex to be unstableand therefore to dissipate more quickly. Trailingaircraft can then follow closer in safety. These

methods, however, also tend to increase thenoise level and decrease the energy efficiency ofthe aircraft. More work needs to be done to de-velop a system which minimizes the vortex withan acceptable price in terms of noise and fuel.

Noise

Many current noise abatement procedures re-quire a tradeoff in terms of reduced airspace andairport capacity. As long as aircraft remainnoisy, however, there is little alternative to rout-ing them away from noise-sensitive areas. Somenew and re-engined jet aircraft are much lessnoisy than their predecessors, but it has beensuggested that technology may have gone as faras it can, and that administrative solutions arethe only alternative. In any case, a great deal offurther research is needed to develop creative so-lutions to the noise problem.

Airport Design

The scarcity of suitable land for expanding ex-isting airports or building new ones means thatnew research is needed on basic concepts of howan airport and its access system should be de-signed. For example, it may be possible to re-design the runway-taxiway system in a manner


that is less profligate of land. Research is neededinto the safety and capacity questions raised bythis type of design. In some locations where littleland is available for a new airport, it may bepossible to locate an airport on a nearby lake orbay. Such an airport would be expensive tobuild, even when the necessary technology hasbeen developed, but in some cases it might bethe most cost-effective alternative.

Ground Access

Airport access is a major area of concern.Research is needed not only to alleviate the ac-cess problems plaguing some of today’s majorairports, but also on cost-effective means to getpassengers out to new airports which may haveto be constructed at distances of 30 to 50 milesfrom the city center.

chapter 6 airport capacity alternatives

Documents