A Study to Optimise the Environmental Capacity of Amsterdam Airport Schiphol

This paper was presented at the World Conference on Transport Research, July 1998, by the Civil Aviation Department, The Netherlands.

By: Jaap de Wit,  Jan Veldhuis,  Peter Uittenboogaart, Thalicia Wei-Yun

1. Context

The last five years Amsterdam Airport has experienced an extremely high growth. However this growth can be attributed only partly to the growth of the airline market itself. The economic growth of the Netherlands and the Western European region has been moderate in the period 1992-1995. Only 1996 and 1997 have shown some economic recovery, at least in the Netherlands. A factor that cer-tainly has contributed to airline market growth concerns the air fares, which have dropped considerably, especially through the introduction of new promotional fares. But even taking the fares into consideration, the contribution of market growth to Schiphol’s growth is moderate. The main factor has been the market share of KLM and its partners. A number of factors can be mentioned in this context. During the first half of the 90’s KLM has extended the co-operation with Northwest Airlines, mainly by codesharing on the North Atlantic route, and by offering through connections in the USA by the Northwest network, and in Europe by the KLM network. An important year was 1992, when the Netherlands – as the first European state – signed an Open Skies Agreement with the United States. In this agreement an anti-trust immunity for KLM/Northwest was included, which made it possible to closely integrate both airline networks. This stimulated traffic at Schiphol further. Also during that period KLM started to build up a new wave system at Schiphol, by concentrating European arrivals and departures (in addition to the Euro-pean and intercontinental) in such a way that connectivity via Schiphol improved considera-bly, which mainly boosted the connecting traffic via Schiphol. A summary of this growth may be found in table 1:

TABLE  1: Traffic volume at Amsterdam Airport Schiphol,  1992 – 1997

Passengers (mln)
Cargo (000 tons)
Movements (000)

Although KLM’s strategy has always focused on a bigger market share as a partner in one of the few remaining global networks, the resulting growth of this strategy has surprised policymakers, especially from the point of view of environmental protection. In 1995 it was agreed that the development of Schiphol would be subject to environmental constraints. An important element in these constraints are the noise contours. According to these constraints a maximum of 15.000 houses within these contours are allowed to be affected by a certain noise level until 2003. After 2003, when the fifth runway is expected to be operational, this contour must shrink even to 10.000 houses. The fast growth how-ever has led to a situation, that within these noise limits, after 1997 only a limited traffic growth is possible. Certainly if traffic demand is continuing to increase fast, drastic measures are needed to obey the noise limit. Measures are already in operation, but their effect is apparently not sufficient enough. Chapter 2- aircraft have already disappeared al/most completely, although this was only supposed to be effective by as long as 2002. The airport has discouraged night operations by differenti-ating landing fees in certain periods of the day. And finally the air transport policy is – when granting access to Schiphol – taking into consideration also the noise performance of the aircraft that is going to be used, when using Schiphol. Nevertheless, some of the measures only become effective after a certain period, and their effectivity is still limited for the short term.

During the second half of 1997 it became clear that – despite of the measures that have already been taken – the noise limit would be violated. It has led to a situation that from April 1st, 1998 Schiphol is fully slot-coordinated. Moreover it was agreed that from 1998 onwards,  20.000 extra slots will be granted annually, enabling a further growth from 360.000 movements in 1997 to 460.000 movements in 2002. Comparing this to the actual market demand, which – for 1998 – which had been estimated already at a level of 420.000 movements, this slot-coordination means a severe restriction. It also implies economic damage not only for the airlines, but possibly also for the surrounding regions. The economic damage will certainly emerge when restrictions continue over the longer term. This situation at Schiphol is however quite unique. Not by the im-plementation of the slot co-ordination, which is existing already at several European airports. But the fact that the co-ordination has been imple-mented because of environmental reasons, and not because of an operational capacity con-straint, like runway capacity. Estimations of long term runway capacity at Schiphol may for instance – depending on the assumptions of peak patterns – increase to a level as high as 600.000 movements, indicating that the operational limits of the airport has not yet been reached.

These considerations have brought up the question, how to implement a mixture of policies, to minimize the economic damage of restricted traffic volumes, within the limits that have been set. The existing slot co-ordination has a strong regulatory orientation. The co-ordination committee is using strict operational rules, and grandfather rights are an important element in using these rules. This may however not necessarily be efficient. It does not guarantee that the best economic performers are given access to Schiphol. It may even lead to a situation that the environmental improvement of the measures is low, but the economic damage is considerable. And even more important: In-centives to improve environmental performance are not inserted into the system. When these incentives are available, a situation may be created by policymakers, that environmental (noise) costs are internalized in the system, stimulating airlines to use the most noise friendly aircraft at the most noise friendly times of the day. It may even mean that the effect of these incentives would be a growth in 2002 beyond the limit of 460.000 movements, within the noise limits.  A mixture of measures leading to this, would be considered as efficient, as it limits economic damage, but has a considerable environmental performance.

2. Unrestricted Scenarios

To address the effects of measures to be taken, it is necessary to separate and define a restricted situation versus an unrestricted situation. Within each of both situations different future growth patterns for Schiphol are assumed by introduc-ing two alternative economic scenarios. One so-called ‘cautious’ scenario based on a moderate economic growth until 2003, including a stabilized market share for KLM. Another so-called ‘favourable’ scenario based on a more optimistic economic growth, combined with the assumption that KLM would gain further market share. This results in various combinations to contrast the differences in traffic development, e.g. two economic scenarios within an unrestricted situation for Schiphol airport as well as two economic scenarios for various restricted airport scenarios. (See figure below)

To trace the differences between the resulting alternatives an Integrated Model System has been developed by the Netherlands Civil Aviation Department (RLD). The two economic scenarios used in this model are based on two macro-economic mid term scenarios of the Dutch Central Planning Bu-reau (CPB). These scenarios describe the economic developments of the Dutch as well as the world economy until the end of the year 2002. For both economic scenarios aviation industry scenarios are developed in case of unrestricted capacity on Schiphol airport. The full development of the KLM five waves system at Schiphol Airport fits in these scenarios.


Unrestricted  Traffic  Scenarios , 1996-2002
Year  1996
Year  2002
Year  2002
Pax (mln.)
Cargo (mln. ton)
ACMs (1000)

A continuation of the aeropolitical selectivity policy of the Dutch government is another assumption made in the aviation scenarios. This implicates that even in the unrestricted scenarios of the model an un-bridled growth of aviation is excluded from the results.

For both unrestricted scenarios the following impacts are analysed or will be included in the Integrated Model System for the period until the year 2002:

traffic and transport on Schiphol; that is both the passengers- and freight volumes and the number of aircraft movements.
environmental effects in the vicinity of Schiphol Airport; that is both the number of hindered houses and the surface of the so-called 35 Ke noise con-tour.
the economic effects of increasing traffic volumes for the vicinity of Schiphol airport; that is the contribution to the local employment and the added value.

Later on monetary preferences of ‘noise consumers’ in the various noise zones around the airport will be included to study possible trade-off effects between positive economic effects and negative external effects in the different areas around the airport.
The corresponding traffic volumes for Schiphol are displayed in next table. These figures are generated by the Schiphol Competition Model, which will be discussed in more detail below.

It is clear that the favourable scenario may have far going implications for Schiphol. Not only with respect to the necessary airport extension like terminals, and the new runway, but particularly with respect to the noise contours that may – very likely – be exceeded in this scenario without additional measures. The necessity of these measures is probably much less in the cautious scenario. It may even mean – depending on the noise emissions – that no additional measures have to be taken as the volumes of 450.000 movements are just within the limits set by the slot-coordination.

3. The Schiphol Competition Model

Starting with 1996 volumes at Schiphol, the growth in passenger demand is forecasted by this model based on economic growth by world region, on trade, and on airfare levels. The demand elasticities vary by region, and between business, independent leisure, and inclusive tour travel. Passenger demand GDP elasticities vary between +0.6 (European destinations) and +2.0 (intercontinental destina-tions, southern hemisphere). These elasticities are assumed to decrease in time. Fare elasticities are -1.0 (leisure passen-gers) and, in addition, there is – ceteris paribus – an underlying annual traffic growth of up to 1% per annum.

This conventional approach is enhanced and extended to re-flect on Schiphol’s throughput the effects of competition from other airports. This will primarily affect transfer traffic, a particularly important segment at Schiphol where the O-D traffic is generated from a rather limited domestic market. A hierarchical logit model, calibrated on existing information for passengers’ air route choices, is used to forecast how passengers choose between alternative air routes via competing airport hubs.

Unlike many air traffic forecasting systems, which assume unlimited airport capacity, the Schiphol Competition model is able to simulate and forecast the traffic volume consequences of a constrained capacity situation, arising from potential policy measures. This enables the model user to assess the impacts of these policies, compared to unconstrained forecasts. This provides a basis to both evaluate the impacts and the robustness of alternative government policies.

The model has been designed to address the following categories of policy measures to reduce airport throughput until 2002:

quotas, or slot control, limit the number of aircraft movements and can be applied to particular traffic categories, for example a ban on night traffic; quotas can also be imposed on total passenger or cargo volumes, although the mechanism to achieve these political targets is less clear in practice; levies or landing surcharges may be applied uniformly or on specific aircraft types or groups of aircraft movements; passenger surcharges may be applied uniformly or on specific passenger market segments, although dis-crimination has to be avoided.

One of the policy instruments that directly affect passenger demand are the departing passenger taxes and passenger demand quota. The passenger tax is added to the airlines’ fare and suppresses demand through the price elasticity. Where passenger demand is constrained to a policy target (a quotum), the model calculates a ‘shadow tax’ as the cost that would be required to constrain demand to meet the policy target. The underlying principle is that, when constraints are effective, the airlines intend to serve the highest yield passenger segments, and these are likely to be those with low price sensitivity and those with relatively limited choice for alternative routes for the journey. The shadow tax gradually reduces the most price sensitive passenger segments (leisure passengers and those with alternative routes like transfer passengers, or those having access to high-speed rail services) and thereby reflects the envisaged airline response.

4. Summary of the results

4.1. The unrestricted scenario

The Schiphol Competition Model as described above is one of the modules in the Integrated Model System available at the Dutch Civil Aviation department. Also a noise contour module is connected to the level of aircraft movements. This enables the policy makers to analyse environmental consequences as far as the number of affected houses in various noise zones is concerned. Another module is the employment and value added module, which derives figures for these variables based on the forecasted traffic and transport volumes at Schiphol Airport. Finally an additional module will be added to translate the physical units of affected houses by noise into monetary values assuming welfare decreases from this noise emissions. 1) Only then a trade-off will be possible between positive effects and negative external effects.

The effects now available for the short term forecasts (the year 2002) from the integral model system are summarized in table 4.1:

TABLE  4.1

The results from table 4.1 emphasize that in an unrestricted growth situation the number of passengers in 2002 will vary between the 40 and 50 million depending on the scenario used. Freight volumes and aircraft movements respectively vary between 1.5 and 1.8 million tons cargo and 450 and 535 thousand movements. Furthermore, the environmental effects reveal that in a favourable unrestricted scenario the maximum of 15000 hindered houses within the 35 Ke zone is exceeded by over 8000, namely 23300 hindered houses. While in the cautious unrestricted scenario the maximum is almost reached already.

One of the reasons why the relative noise impact in the favourable scenario is considerably higher then in the cautious scenario, is the use of extra landing or take-off runways in peak hours. The approach and taking off routes of these extra runways are situated over densely populated areas. In off-peak periods these runways are avoided as much as possible, of course depending on weather conditions.

4.2. Restricted Scenarios

In the unrestricted scenario policy measures are inevitable to bring down the noise emission levels. Therefore the next starting point of analysis is some kind of restriction in case of the favourable scenario. A few measures have however already been taken, and we will first evaluate their effects if these would be continued to be implemented until 2003.

In the Physical Planning Document (PKB) for Schiphol the traffic growth has been restricted up to a maximum of 44 million passengers and 3.3 million tonne freight per year. In addition to these restrictions, recently the government has decided that Schiphol is not allowed to accommodate more than 460,000 aircraft movements in the year 2002. Therefter a new runway is planned to be available which provides new opportunities for further traffic growth.

This implicates that Schiphol now is a slot-coordinated air-port. In principle two slot-allocation mechanisms deserve consideration. First of all, the present allocation system based on grandfather rights. This system does not automatically lead to an optimal economic allocation of slots, because a system based on grandfather rights does not take into account any value added. An alternative is a slot-allocation system where slots are free to allocate and (maybe even) tradable. Although this system is not (yet) approved by EU law it would lead to an optimal economic allocation of scarce capacity. The reason for this is that tradable (scarce) slots will push upward the price. Therefore slots will only be allocated to those parties in the market willing and able to pay the higher fares that are inevitably the result from this slot-trading.

The effects of these restrictions on the economic and environmental loss and gain based on the favourable scenario are summarized in table 4.2:

Summary of efforts of unrestricted and restricted development, 2002

Table 4.2 demonstrates that in comparison to the favourable unrestricted scenario, a restriction of 460,000 market allocation system results in a reduction of  3.1 million passengers (e.g. respectively 48.3 and 45.2 pax) and a substantial loss of cargo. However, it should be taken in consideration that the uncertainties concerning price elasticities of freight are large. Although price sensitivity relations are well known for passengers, detailed information is almost completely missing for freight. Nevertheless, it is expected that the price sensitivity of freight is larger than that of passengers.

Therefore the model assumes that in case of a market allocation system freight will relatively be hit stronger by the price increase for available scarce slots.

In comparison to the favourable unrestricted scenario the environmental effects of this ‘restricted’ analysis are positive, since the number of hindered houses within the 35 Ke zone is 8 to 11 thousand less and within the PKB constraints of a maximum of 15000 hindered houses. In this analysis the number of hindered houses varies between 12000 and 15000 houses. The extent to which it will be closer to 12000 or 15000 is dependent on the reaction of the full freighters on the price increase.

The third effect is the economic effect. The direct and indirect employment will decrease, varying from a loss of 8000 to 11000 full time equivalents. This is also caused by the substantial reduction in freight volume, but similarly depends on the behaviour of the full freighters and whether they will avoid Schiphol or not. Furthermore the (shadow-) costs of mobility turns out to be 1,15 billion higher then in a unrestricted situation. The major part of this increase is on account of passenger transport since freight transport can be and is outplaced at relatively low costs. The increase in costs for the passengers still travelling via Schiphol is 24 guilders per passenger, while the increase for forwarders is 88 guilders per tonne.

So, while the slot restriction based on market principles is effective in terms of environmental effects (<15000 hindered houses) it is not efficient in terms of economic effects, considering the substantial increase in social costs and the considerable loss in terms of employment and added value.

A more regulated slot restriction will not hit the freight segment as hard as indicated above, because this category also has the disposal of grand-father rights. In this analysis the passenger segment is affected to a larger extent than the freight segment. As can be seen in table 4.2, five to seven million passengers will no longer travel via Schiphol. The consequence is an immediate increase in mobility (shadow) costs, which will be much higher than the mobility (shadow) costs in the alternative slot allocation mechanism. However it is difficult to judge which segment will be affected the most, because there is no clear insight in the allocation of grandfather rights in the year 2002 2) and the model is not capable of simulating this. That is why the amount of mobility costs is very uncertain. Nevertheless it can be expected that the mobility costs are larger than 1.15 billion guilders.

In environmental respect this regulated allocation system is more ineffective than the market system, because in the former case full freighters are accommodated. These transporters usually fly older aircraft and operate at night-time, late in the evening or early in the morn-ing. In economic respect the results are even worse compared with the market allocation mechanism, because fewer passengers are accommodated on Schiphol and the costs associated with avoiding Schiphol is much higher for the passenger segment than for the freight segment. As to the employment effects the difference between the two allocation systems is less clear and is probably mainly depending on the reaction of the freight segment when everything is left to the market.

Table 4.2 also clarifies that a restriction of 44 million passengers (PKB Schiphol) is not only more inefficient than each of the two slot allocation systems, but is also worse for the environment (when noise is considered). The social costs are increasing by 2.6 billion guilders and the number of hindered houses is well above the maximum of 15000, namely 20 to 22 thousand hindered houses. This is caused by the fact that the number of aircraft movements is more than 490.000 and full freighters are accommodated on Schiphol.

It must be pointed out that both the increase in costs as a consequence of growth restrictions on the aviation industry in  the Netherlands and the related loss in employment and added value are consistently underestimated. This is based on the following two aspects:

  • Because of a more stringent policy the growth potential in recent years has not been fully exploited intentionally.
  • Although the extra growth-potential spinning of the alliance between KLM and Alitalia has been taken into account in the model not all extra growth-potential of (future) alliances are included (e.g. the effects of the alliance between Continental and Northwest or maybe an alliance with an East-European carrier).

4.3. Other measures

The above measures have the intention to reduce traffic volume to certain levels, either by regulatory or market mechanism. Nevertheless the relation with the noise level is an indirect one. In taking these measures it is aimed that also noise levels are reduced in line with this, although the extent to which noise is reduced varies considerably.

Other measures, more directly related to noise emissions, may therefore be considered. In this analysis the efficiency and effectiveness of three types of measures are worked out.

  • price control measures; The aim of these type of measures is to give an incentive to airlines to use the environmental capacity more efficiently via the price mechanism. That is, to encourage air carriers to use the most quiet aircraft, preferably not in the night or early morning and late evening period.
  • measures concerning flight scheduling; The aim of these kind of measures is to spread the flight pattern of carriers in such a way that the use of extra runways in peak times is less necessary.
  • technical operational measures; The aim of these kind of measures is to optimize the use of airspace- and runway capacity. In this way the number of aircraft movements within the 15.000 houses contour can be maximized without negative consequences for safety.

A surcharge may be considered according to the use of environmental capacity. The use of environmental capacity is depending on the noise emissions of the aircraft and the time of the day when the aircraft is taking off or landing. In noise computations time penalty factors are used for different hours of the day. Both measures have been evaluated in their effects. A surcharge has been set on aircraft movements according the use of environmental capacity. The surcharge has been set in such a way, that – on average – the resulting fare increase for passengers will be Dfl. 50 per return ticket. However passengers flying in the noisiest aircraft are supposed to pay Dfl. 100, whereas passengers in the most noise friendly fleet are not supposed to pay any surcharge. Furthermore night flights do use considerable more environmental capacity, and therefore the surcharge on passenger movements is assumed to be Dfl. 250, when taking off or landing is in the night period. The charge is lower according to the time of day of taking off or landing.

Furthermore some analysis have been done assuming peak spreading of the latest evening arrivals. These arrivals are not followed by departures on the same day, and consequently there is no connectivity loss to be expected if this peak is spreaded in such a way that the use of a second landing runway is not necessary. As mentioned before the use of this second runway does affect many residential areas, at the evening period when a noise penalty factor is also applied.

Finally some analysis has been done assuming improvements in technical operational procedures. These improvements refer to landing and take-off procedures and to runway use conditions with respect to weather conditions. It is expected that these procedures may also help in reducing perceived noise levels.

In table 4.3 the effects of the above mentioned types of measures are summarized.

Summary of efforts of unrestricted and restricted development, 2002



Cargo(mln. tons) Aircraft
in 35 Ke
Employment(000’s) Shadow
(bln. Dfl)
Cautious 39.2 1.5 452 14.100
Favourable 48.3 1.8 533 23.300
44 mln. pax quotum 44.0 1.8 491 20 to 22.000 -3000 2.64
Slot Regulation 41 to 43 1.6 to 1.7 460 16 to 18.000 P.M. >> 1.15
Slot Trading 45.2 0.5 to 1.0 460 12 to 15.000 -8 to -11 1.15
Levies by noise cat. 46.5 1.6 509 15 to 17.000 -3.000 1.20
Levies by time of day 46.4 0.8 to 1.2 498 14 to 17.000 -4 to -7.000 1.20
Evening peak spr. 48.3 1.8 533 18 to 20.000 P.M.
Tech. Operational Meas. 48.3 1.8 533 14 to 17.000

Table 4.3 puts forward that each of these measures results already in a considerable reduction of the number of hindered houses in relation to the unrestricted favourable scenario:

A surcharge on noise levels per aircraft results in 6 to 8 thousand less hindered houses or a reduction of approximately 30%. This charge particularly encourages airlines either to shift within their fleet to operate more noise friendly aircraft on Schiphol or to invest in a more modern fleet, of course depending on the surcharge set. Similar argumentation holds when surcharge is set on operations for different times of the day. Both types of measures have been set in such a way that the total shadow costs are similar. Both measures do not differ in effects for passenger, but when charges are set for night flights, we see cargo being hit more severely, as the cargo segment is using Schiphol relatively more in the night period than the passenger segment. Consequently also total employment effects are more negative in case of levies by time of day. Although also with these type of measures the shadow costs are high, we must conclude that the efficiency of these measures with respect to noise reduction is higher when compared to the regulatory restrictions.
An adjustment of flight schedules, where the last evening arrival wave is levelled off results in a reduction of hindered houses of 3 to 5 thousand or 15% less. As the costs and traffic loss is expected to be low, if not zero, we must conclude that the efficiency of this measure is high. Finally also the effects of technical operational measures are considerable in the effects on noise reduction, although their implementation costs and efforts are high. It is generally felt that the effects of this type of measures are more promising on the longer term compared to the short term period observed in this analysis.

The preceding findings show that separate implementation of the latter type of measures already leads to considerable reductions in noise levels. This may give an idea about the potential reductions when these measures are implemented as some kind of mix between them. With a combination of measures it is expected to be possible to achieve still a considerable growth to levels of more than 500.000 aircraft movement and still to respect the environmental limits that have been set.


For the time being the following conclusions can be drawn:

Capacity restrictions on Schiphol do not offer the final solution for the noise problems at Schiphol. These restrictions are in economical respect inefficient, because of the high social costs, and the reductions in noise emissions are insufficient

The three other optimization measures are however more promising. There are certainly economic costs involved, but on the other hand the efficiency with respect to noise reduction is considerable higher. With some kind of combinations of the measures mentioned, it is expected that growth level – respecting the environmental limits – may be reached beyond the slot control limits that have currently been set.

End notes

At this moment the noise nuisance (noise contours) is the only environmental effect in the model. Other environmental effects like emissions are not yet addressed in the integral modelsystem, because these effects are not on the critical path of the capacity restrictions which Schiphol Airport is dealing with on the short term (= optimilization of the existing capacity at Schiphol until 2002).
The government has granted Schiphol an annual growth in aircraft movements of 20.000, starting on January 1st, 1998. This means that in the period till 2002 the available capacity will reach a maximum of 460.000 (= 360.000 + 5 *20.000

About Admin
Welcome to my Personal Blog! Just feel free to come back regularly and read some updates or subscribe to an RSS feed!

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

%d bloggers like this: