Inleiding Na de ingebruikname van de Polderbaan in 2003 ontving het Informatie & Klachtenbureau van de Commissie Regionaal Overleg luchthaven Schiphol (CROS) veel klachten over grondgeluid uit Hoofddorp-Noord. Werden in 2002 vanuit Hoofddorp 84 aan grondgeluid gerelateerde klachten geregistreerd, in 2004 was dit aantal opgelopen tot 5.257. Grondgeluid is in dit onderzoek gedefinieerd als het laag frequente geluid van vliegtuig operaties, wat niet wordt meegenomen in de traditionele geluidsmodellen of meetsystemen; beide presenteren data in de vorm van zogenaamd A-gewogen geluidsniveaus, welke meer geschikt zijn voor het beschrijven van geluid van overvliegende vliegtuigen. In de Milieu Effect Rapportage Schiphol 2003, die ten grondslag ligt aan het gebruik van de Polderbaan, is naar nu blijkt de hoeveelheid hinder veroorzaakt door grondgeluid en de intensiteit hiervan niet voorzien ofwel onderschat. Uit gesprekken met gehinderde omwonenden bleek dat zij vooral hinder ondervonden van geluid en trillingen in huis bij startende vliegtuigen vanaf de Polderbaan. Deze hinder kon niet worden verklaard vanuit de verwachte geluidbelasting, zoals die was onderzocht in de Milieueffect Rapportage Schiphol 2003 en deelstudies1 naar de milieu-impact van de nieuwe Polderbaan. Na overleg met de Bewonersvereniging Hoofddorp-Noord en de gemeente Haarlemmermeer heeft Schiphol Group begin 2004 besloten om een onderzoek te starten met als doelstellingen om de oorzaak van de overlast vast te stellen, inzicht te krijgen in de mate van ervaren hinder en mogelijke maatregelen om deze hinder te kunnen beperken te onderzoeken. Met andere woorden, doel van de onderzoek was om eerst te begrijpen wat er aan de hand was en vervolgens te onderzoeken wat er aan gedaan kan worden. Dit onderzoek, uitgevoerd door TNO, NLR en het Amerikaanse Wyle laboratories werd gedurende de hele onderzoeksperiode gevolgd vanuit een technische begeleidingsgroep waarin behalve medewerkers van Schiphol en de onderzoeksbureaus ook bewoners en vertegenwoordigers van de gemeente zitting hadden. Het onderzoek heeft bevestigd dat de bewoners in Hoofddorp-Noord bij bepaalde omstandigheden hinder hebben van startende vliegtuigen vanaf de Polderbaan. De hinder wordt voornamelijk veroorzaakt door de lage frequenties die grondgeluid typeren en kan, afhankelijk van een aantal factoren vervelend/hinderlijkk of zelfs zeer hinderlijk/zwaar belastendd zijn. Uit het onderzoek van TNO, NLR en Wyle laboratories blijkt bovendien dat
1
het laagfrequente geluid veroorzaakt door vooral grote vliegtuigen hierbij een belangrijke rol speelt. Dit fenomeen (geluid en trillingen) wordt extra versterkt bij bepaalde weersomstandigheden en temperaturen. Na toetsing van een scala aan maatregelen in het onderzoeksmodel hebben de overlegpartners van de gemeente Haarlemmermeer, Bewonersvereniging Hoofddorp-Noord en Schiphol Group gezamenlijk maatregelen gewogen op basis van onder meer effectiviteit, haalbaarheid en de invloed op de mainport. Uit een technische analyse van deze afweging blijkt dat een hoge geluidswal (of een hoog gebouw), het verbeteren van grondabsorptie en operationele maatregelen de meest kansrijke maatregelen zijn om de huidige overlast te verminderen. Grondgeluid was tot nu toe in (internationale) onderzoeken naar geluidbelasting rond luchthavens gedefinieerd als een combinatie van het geluid van taxiënde vliegtuigen en het geluid dat wordt veroorzaakt bij “reverse thrust”, waarbij een vliegtuig na de landing wordt afgeremd via de motoren. De invloed van deze vorm van grondgeluid werd over het algemeen als verwaarloosbaar aangeduid ten opzichte van het geluid van een startend of landend vliegtuig. Laagfrequent geluid van vliegtuigen vormt een relatief nieuwe hinderbron en er is nog relatief weinig over bekend. In de bestaande geluidsnormen Lden en Lnight wordt met het laagfrequente karakter van grondgeluid dan ook onvoldoende rekening gehouden, vanwege de dB (A) maatstaf van die normen (A-gewogen geluidsniveaus). De nu vastgestelde overlast voor bewoners in HoofddorpNoord als gevolg van het voornamelijk uit laagfrequent geluid (geluid en trillingen) bestaande grondgeluid vraagt om effectieve oplossingen. Een aantal maatregelen wordt op dit moment nader uitgewerkt. Gezien de ervaren overlast is er urgentie hierin te investeren en de bewoners hierbij te blijven betrekken. Om die reden is dit rapport ook aangeboden aan de instanties die een belangrijke rol spelen in hinderbeperking rond de luchthaven Schiphol en zich momenteel buigen over de verbetervoorstellen in het kader van de evaluatie van het Schipholbeleid: de Commissie Regionaal Overleg luchthaven Schiphol en de Staatsecretarissen van de ministeries van Verkeer & Waterstaat en VROM. Een overzicht van het gehele onderzoek is te vinden als centraal deel (Engelstalig) van dit rapport. Vervolgens wordt door de initiatiefnemers in een afsluitend Nederlandstalig hoofdstuk conclusies getrokken en vervolgstappen aangekondigd. Het onderzoek is uitgevoerd in opdracht en op kosten van Schiphol Group.
Rapport ML 447-1 d.d. 3 juli 2001 “Geluid vanwege het taxiën van vliegtuigen op de Luchthaven Schiphol”, uitgevoerd door Adviesbureau Peutz
Wyle Report
WR 06-02 (J/N 52611) February 2006 Wyle Contract No. 2000008725/0
Groundnoise Polderbaan Overview of Results
Prepared for
Schiphol Nederland BV Postbus 7501 1118 ZG SCHIPHOL Nederland
Prepared by
Ben Sharp – Wyle Toon Beeks – TNO Henk Veerbeek - NLR
Table of Contents 1.0
2.0
3.0
4.0
5.0
6.0
7.0
INTRODUCTION
2
1.1
Project Goals
2
1.2
Project Background and Phases
2
INITIAL MODELING
2
2.1
Definition of Possible Mechanisms
3
2.2
Noise and Vibration Modeling
3
2.3
Conclusions from the Initial Modeling
4
NOISE AND VIBRATION MEASUREMENT SURVEY
5
3.1
Selection of Measurement Sites
5
3.2
Data Acquisition and Processing
6
3.3
Data Analysis and Conclusions
6
MODEL VALIDATION
8
4.1
Source Strength and Directivity
8
4.2
Noise Propagation
8
4.3
Conclusions from Model Validation
10
MITIGATION GOALS
11
5.1
Noise Criteria
11
5.2
Selection of Noise Goals
12
EFFECTIVENESS OF POSSIBLE MITIGATION MEASURES
13
6.1
Selection of Possible Measures
13
6.2
Barriers
13
6.3
Ground Absorption
14
6.4
Sound Insulation
14
6.5
Operational Procedures
15
6.6
Long-Term Mitigation Measures
16
CONCLUSIONS
17
7.1
17
Summary – Conclusion-matrix
REFERENCES FOR LOW FREQUENCY NOISE CRITERIA
20
1
1.0
INTRODUCTION This report “Groundnoise Polderbaan, Overview of Results” can be considered as the technical management summary of the whole project. More detailed information, including description of research methods, data sources, analysis and detailed results is available (only in digital format) in a separate technical annex of all interim reports and presentations.
1.1
Project Goal To understand and define the phenomenon Groundnoise and investigate cost-effective mitigation measures. Groundnoise is defined as the lowfrequency noise generated by aircraft operations that is not accounted for by conventional noise models or noise monitoring systems, both of which provide data in the form of A-weighted noise levels that are more suited to describing the noise from aircraft overflights.
1.2
Project Background and Phases In 2003, after the new Polderbaan at Schiphol Airport was first used, an unexpected type and number of complaints was reported from HoofddorpNoord. The reported nuisance correlated with take-off operations from the runway, but could not be explained by traditional noise-modeling. December 2003 Schiphol Group and the municipality of Haarlemmermeer, in consultation with the Community Group Hoofddorp-Noord, decided to start a project “Groundnoise”, based on agreed upon project outlines. After distribution of work packages between the three involved research organizations, various contracts were awarded by Schiphol Group based on separate Statements of Work. The overall project to investigate Groundnoise at Schiphol Airport has been conducted in three main phases. Phases I and II consisted on noise surveys in Hoofddorp-Noord. In Phase I, noise data taken from noise monitoring system NOMOS (designed to measure fly-over noise) did not correlate with nuisance reported by the community, possibly because NOMOS data is presented in terms of A-weighted noise levels which deemphasize noise at low-frequencies. In Phase II, conducted in November 2004, more detailed measurements of noise and vibration were taken at two test houses – a brick/concrete house in Vrijschot, and a wooden-type in Houtwijkervelt - with emphasis on low-frequencies. The results showed that during departures of large aircraft from the Polderbaan significant nuisance may occur under certain conditions. The study also concluded that additional information was required on the low-frequency noise/vibration source and propagation mechanisms in order to develop mitigation measures. Phase III, the current phase, was designed to develop a better understanding of the source and propagation mechanisms, and to evaluate possible mitigation measures.
2
Phase III was conducted in four steps, namely: • Step 1: Initial modeling - to provide an initial estimate of contributions. • Step 2: Noise and vibration measurements – to develop the data necessary to define the contributions, and provide data for model validation. • Step 3: Model validation – to confirm that the model replicated the measurement data, and to provide a tool for evaluating mitigation measures. • Step 4: Select and evaluate mitigation measures.
2.0
INITIAL MODELING To be successful in reducing community noise levels, mitigation measures must be applied to the dominant noise and vibration source and propagation mechanisms. It was therefore necessary to identify all possible mechanisms and develop initial (approximate) estimates of their contribution so that the measurement study could be designed most efficiently.
2.1
Definition of Possible Mechanisms The low-frequency noise and vibration levels generated in the local community by aircraft departing on the Polderbaan may be the result of one or more source and propagation mechanisms. The possible mechanisms are: • Jet noise generated by the engines and propagated through the air. • Vibration of the runway and surrounding area caused by excitation from the jet turbulence and jet noise, and propagated through the ground. • Vibration of the runway caused propagated through the ground.
by
aircraft
movement,
and
Of major importance are the relative contributions from airborne and groundborne propagation as this will determine the type of mitigation measures required. 2.2
Noise and Vibration Modeling Models were developed for the two possible propagation mechanisms – groundborne vibration and airborne noise. At this stage of the study the detailed parameters, such as ground absorption, wind and temperature conditions etc., were unknown, so the modeling was conducted using a range of parameters to cover all possible conditions. For the modeling of groundborne vibration, the two types of waves (known as Raleigh waves and surface waves) with the lowest attenuation with distance were considered, and typical attenuation rates were obtained from the published literature.
3
It was shown that, to be perceptible at houses in Hoofddorp-Noord with the predicted attenuation from the runway to the houses, the vibration levels at the runway would have to be unrealistically high. For the modeling of airborne noise, typical noise data for a DC10 was taken as the source. Since long-range noise propagation is very strongly dependent on wind and temperature distributions (gradients) and ground absorption, the calculations were performed for several atmospheric conditions experienced at Schiphol Airport, and for two ground conditions, namely soft absorbing grass and a hard non-absorbing surface. An advanced calculation method was employed using a Parabolic Equation (PE) scheme that allows detailed atmospheric effects and ground absorption to be taken into account. The modeling results showed that the dominant frequencies of the noise experienced in Hoofddorp-Noord were in the range of 31.5Hz. When propagation conditions are favorable (temperature increasing with height above the ground – inversion - and/or a wind from the runway to the houses) it was determined that the noise levels could be a nuisance. For unfavorable propagation conditions (temperature decreasing with height, and/or no wind or wind from the house to the runway) the noise levels probably would not cause a nuisance.
2.3
Conclusions from the Initial Modeling The first important conclusion to be drawn from the modeling is that the transfer of ground-vibrations is too weak to produce observable vibration levels at the test house, and that it is not likely that the vibrations measured in the houses in Hoofddorp-Noord are induced by the transfer of ground-vibration. It is most likely that the vibrations are the result of excitation of the house structure by airborne noise. Secondly, the airborne sound transfer path may produce nuisance under favorable propagation conditions (with wind from the runway to the houses). These tentative conclusions were then validated through a measurement survey.
4
3.0
NOISE AND VIBRATION MEASUREMENT SURVEY The purpose of the noise measurement survey was twofold, namely: • To validate the conclusions of the initial modeling and establish the main source and propagation mechanisms. • To collect the parameters required effectiveness of mitigation measures.
3.1
for
the
evaluation
and
Selection of Measurement Sites Noise and vibration measurements were conducted in August 2005 at the following locations – see Figure 1:
3 2 4 1
5 6
8
10
7
Figure 1: Measurement survey locations • Close to the runway, to document the source levels: - To the side and rear of the aircraft at the start-of-takeoff, to define the noise and vibration source levels - Alongside the runway, to obtain source directivity data • In the farmland, between the runway and the test house, to gain a better understanding of propagation. • In front of the dike, to determine the attenuation provided by the dike. • At the test house (the same brick/concrete house from Phase II), to document community levels and the attenuation provided by the test
5
house structure: - Outside the test house at two elevations - Inside the test house 3.2
Data Acquisition and Processing Measurements were conducted over a period of two days under different wind conditions. On the first day there was light rain, but no significant wind at the site. There was no rain on the second day, but there was a 6 to 10 kph wind from the north-west. The forecast for the following days did not include winds from the north-east (in the direction from the aircraft to the test house) and so no measurements were taken after the second day. Before the measurement period on each day all microphones and geophones were calibrated and recording equipment was synchronized in time so that data from different locations could be subsequently compared. Atmospheric conditions, namely wind speed and direction, temperature, and relative humidity, were continually monitored throughout the measurement period. A video camera was set up close to the runway to document all aircraft departure conditions, such as aircraft type, start-of-take-off-roll location, rolling start vs stationary start, etc. Also, highly accurate aircraft position information was obtained from radar data measurements.
3.3
Data Analysis and Conclusions The first step in the data analysis was to develop a table of noise levels measured at the test house for all the events (departures) and relate them to the aircraft type. As expected from the results of the November 2004 measurements, the highest noise levels were associated with the larger and heavier aircraft, such as the B747-400, DC10, MD11, and A330. Of these, the DC10 produced the highest noise levels at low frequencies, with the B744 and MD11 lower by less than 5 dB. Examination of the ground vibration data measured at the farmland site between the runway and the test house showed that the levels were very low – too low in fact to be responsible for any significant house vibration. Noise and vibration data measured at the runway and at the test house were superimposed to compare noise and vibration levels and the times at which maximum levels were received. This is illustrated in Figure 2 showing the time histories of the unweighted noise and vibration levels at the runway and at the test house for a single MD11 departure.
6
110
100
90
80
70
60
50
40
30
20
10
0 13:16:30
G eophone v elocity dB [re 1 V olt + 94 dB ]
S ound pressure lev el [dB re 2 0 m icroP asc al]
120
-10 13:17:00 time [hh:mm:ss]
Figure 2: Time history of unweighted noise and vibration levels for an MD11 departure – noise level close to the runway (top); noise level outside the house (center); vibration level at the house (bottom).
Comparing the time histories in Figure 2 it can be seen that there is a time difference of 6 to 10 seconds between the noise generated at the runway and the noise measured at the test house. This corresponds to the time for noise to travel over the intervening distance. It can also be noted that the house vibration levels, although low, occur at exactly the same time as the noise levels at the house. Since it is likely that the propagation time for ground vibration is much less than that for airborne noise, it can be concluded that the house vibration is the result of airborne noise excitation and not by ground vibration. This result is common to all of the analyzed events. These experimental results confirm results obtained from the initial modeling, namely that the transfer of ground-vibrations is too weak to produce observable vibration levels at the test house, and so it is not likely that the vibrations measured in the houses in Hoofddorp-Noord are induced by ground-vibration transfer. Noise data measured on the two days were compared to determine the effect of wind on the noise levels at the test house. Wind conditions on the second day (6 kph from the northwest) were not the most favorable for noise propagation from the runway toward the house, yet produced an average increase of 4 dB in the measured sound levels compared to no-wind conditions. It can be expected that with more favorable north-easterly winds or other atmospheric conditions, levels could be increased by 5 to 10 dB.
7
4.0
MODEL VALIDATION One of the purposes of the measurement survey was to provide data which could be used to improve and validate the model used for the initial estimates, so that it could be used to predict the performance of various mitigation measures. The main elements for which inputs and validation were required are as follows: • Noise source strength and directivity • Meteorological conditions • Ground absorption The modeling process involved simulating the aircraft take-off procedure (source position vs time), and calculating the noise level time history at the test house using the PE modeling scheme developed in the initial modeling task. This time, however, model input parameters were updated using the data obtained in the measurement survey.
4.1
Source Strength and Directivity The noise data used to define the source in the initial modeling was replaced by measured data taken for the noisiest aircraft (B747, DC10, A330, and MD11) during the survey. The source strength and directivity were determined from pass-by measurements using microphones close to the side and behind departing aircraft. As expected, the directional radiation characteristics show a 10 to 20 dB increase at 45 degrees from the rear of the aircraft compared to the front at the most important frequency (31.5 Hz) depending on the aircraft type. These higher radiation levels are directed towards Hoofddorp when the aircraft starts its takeoff roll.
4.2
Noise Propagation (ground absorption, wind effects) Using the measured noise source characteristics of the noisiest aircraft types, together with best estimates of the ground absorption, the noise level time histories were calculated for two different locations – an intermediate position at the farm, and at the dike separating the farm area from the community in Hoofddorp – and compared with the measured time-histories. The results are shown in Figures 3.
T h e c a l c u l Figure 3: Modeled and measured noise level time histories at the farmland location (left) and in front of the dike (right) 8
The calculated levels at the farm, approximately 1000m from the end of the runway, are in very good agreement with the measured levels for the entire event. At the dike, approximately 2000m from the runway, the agreement is not quite as good, as would be expected, but still well within acceptable limits. Thus the model can be used to evaluate the effect of propagation parameters and of mitigation measures. The most important parameters that affect noise propagation are wind speed and direction and ground absorption. The effect of wind direction is illustrated in Figure 4, where it can be seen that wind from the north-east produces higher noise levels in the south-west direction.
Figure 4: Effect of wind on noise directivity of aircraft: no wind (left); wind from the NE (right)
The interrelationship between these parameters is illustrated in Figure 5 where sound level is calculated for different wind speeds for two different values of ground absorption. Since the purpose of this graph is to demonstrate the relative effects of the propagation parameters no absolute levels are indicated along the vertical axis.
9
Figure 5: Sound level as a function of wind speed for two different values of ground absorption
For the case where the wind direction is from the south, from Hoofddorp to the runway, (represented by negative wind speed in Figure 5) the aircraft noise is attenuated by up to 20 dB. Under this condition the noise is less likely to cause a nuisance in the community, as proven during the measurements in August 2005. For positive wind speeds, representing wind from the north, the situation is quite different depending on the ground absorption. If the absorption is high, as would be the case with soft grassland, the levels will be increased by less than 5 dB for all wind speeds. If the absorption is low, as would be the case with a hard surface such as concrete or water, the levels will increase (with 10 dB or more) with increasing wind speed. 4.3
Conclusions from Model Validation Low-frequency noise levels have been predicted using the long-range noise propagation model and the experimentally determined values of sound power and directivity patterns of aircraft at the runway. Aircraft departures on the Polderbaan with even a slight tail wind (from the south) are less likely to cause a nuisance in the community. Under favorable propagation conditions (wind from the north-east) the noise levels in the community will be increased by an amount dependent on the absorption of the ground along the transfer path. • For ‘soft’ ground the increase due to wind is small (< 5 dB) • For ‘hard’ ground the increase due to wind may be larger than 10 dB A soft ground surface will reduce the influence of wind on community noise levels.
10
5.0
MITIGATION GOALS
5.1
Noise Criteria Before mitigation measures can be evaluated it is necessary to establish goals for acceptable low-frequency noise levels in the community, and then determine the amount of noise reduction required to minimize nuisance. Criteria for acceptable low-frequency noise levels were established based on available information from a wide range of published research data (see the list of references at the end of this report). These criteria were then compared to the noise levels measured at the test house under different meteorological conditions, and to the reports of nuisance. A summary of criteria for low-frequency noise is shown in Figure 6 in terms of noise levels versus frequency. The region marked ‘inaudible’ represents noise levels below the threshold of hearing. In this region the exposure is not perceived by the ear, although it might be experienced by the body as a vibration. The other three regions in the figure represent noise levels that are perceived to be ‘detectable’, ‘annoying/objectionable’, and ‘oppressive chest vibration feeling’.
120
110
Oppressive Chest Vibration Feeling
Sound Pressure Level, dB
100
90
Annoying/Objectionable
80
Inaudible
70
Detectable
60
50 8
16
31.5
63
125
250
1/3-octave Band Center Frequency, Hz
Figure 6: Effects of low-frequency noise on people with noise levels from measurement survey – highest outdoor levels per 1/3-octave band, no wind, no nuisance (solid line); highest outdoor levels, wind, nuisance (dashed line).
11
Also included in the figure are the results of noise measurements that were conducted at the test house in Hoofddorp-Noord. The solid black curve represents the highest outdoor noise levels per 1/3-octave band measured over the two-day period of the survey in August 2005 when no nuisance was reported. According to the description of effects shown in the figure, these noise levels are detectable but not annoying or objectionable, which is consistent with the response noted. This then represents a baseline condition that should not be exceeded. The dashed black curve represents the highest outdoor noise levels per 1/3-octave band measured in the previous survey of November 2004 when nuisance was reported for some events. This response is consistent with the description of effects which states that the noise can be annoying and even in some cases causing oppressive chest vibration. 5.2
Selection of Noise Goals In summary, the measured data from Phases II and III combined with reports on nuisance are consistent with the effects shown in Figure 6 derived form the published literature. What should be prevented is an increase of sound levels compared to the situation measured in August 2005 when no nuisance was reported. The data in the figure shows that 5 to 10 dB higher sound levels can occur under favorable propagation conditions (see also Section 4.3). Therefore, the project goal should be to provide an additional attenuation of 5 to 10 dB. The most important frequencies to be considered are in the range of 31.5 Hz.
12
6.0
EFFECTIVENESS OF POSSIBLE MITIGATION MEASURES
6.1
Selection of Possible Measures The project team and community representatives from Hoofddorp selected the following possible mitigation measures for study: • Sound barriers (single, multiple, with and without vegetation), and buildings. • Ground absorption mechanisms and materials, including vegetation. • Operational procedures (aircraft type exclusions, weather exclusions, runway alternatives, take-off procedures) • Extra sound insulation of dwellings • Anti-noise and air-path disturbing elements (low priority) The goal is a 5 to 10 dB reduction in noise levels at 31.5 Hz.
6.2
Barriers Barriers provide attenuation by eliminating the direct line of sight between the noise source and the receiver. They don’t work quite as well as might be expected, however, because the sound diffracts, or bends, over the top of the barrier, and propagates into the shadow zone behind it, thereby reducing the attenuation. The higher the barrier, and the higher the frequency of the sound, the less the bending, and the higher the attenuation. Sources close to the barrier are better attenuated than those farther away. Similarly, receiver locations near a barrier experience a higher attenuation than those further away. The presence of wind can seriously reduce the acoustic performance of a barrier by increasing the diffraction of sound over the barrier and increasing the levels on the receiver side. The loss of attenuation increases as the wind speed increases. Using the validated noise propagation model, the effectiveness of barriers close to the runway and close to the Hoofddorp community were evaluated with the following results: • Barriers 10m high near the source may reduce community noise levels at low frequencies about 6 dB in case of sound-absorbing ground, for the aircraft at the starting position. During take-off, the barrier performance will decrease owing to the increasing distance to the barrier and rising aircraft. • Barriers near the dwellings are less effective. Heights of at least 15m are required to achieve a similar attenuation of 6 dB. • The performance of a barrier depends critically of the combination of ground absorption and barrier height. In conclusion, the evaluation of noise barriers shows that to be effective at low-frequencies, barrier heights must be at least 10 to 15m high.
13
6.3
Ground Absorption According to the validated model, the absorption of the ground between the runway and the community has a large influence on the noise propagation, particularly in the presence of wind. To examine this effect more fully, and to understand the influence of ground absorption on the noise levels in Hoofddorp, in-situ measurements of absorption were conducted in the area between the runway and the dike. An impulsive noise source was used so that the direct and ground reflected signals could be separated and compared, resulting in an accurate measurement of the absorption provided by the ground surface. Measurements were conducted in areas where the surface was covered with corn-stubble, partly covered with water, and the results compared with similar measurements conducted over a grassland surface. The measured data was then inserted into the noise propagation model to determine the effect on community noise levels. It was found that there were large local variations in noise absorption properties, and locally strong reflecting ground, particularly where there was standing water. For a sufficiently large area of strong reflecting ground, the noise level variations induced by wind variations at Hoofddorp-Noord can exceed 10 dB in the 31.5 Hz octave band. To prevent such an increase of noise under favorable propagation conditions (in the presence of wind) by maximizing the noise absorption quality of the ground, a soft grassland-type surface would be sufficient, such as a well-drained golf course-type grass. The goal would be to provide such a low ground impedance under all weather conditions. Examination of data in the literature indicates the following possibilities: • Membrane-type absorbers as used in concert halls. Although effective, this measure would not be feasible in outdoor conditions. • An artificial surface, such as a gravel pit (layer of gravel with a total depth of about 1.5m). • Natural vegetation, such as a forest with a thick layer of ground humus (after several years). The lateral and longitudinal extent of the latter two measures requires further study to establish the optimum combination of ground absorption when used in conjunction with realistic barrier designs.
6.4
Sound Insulation The noise levels inside the houses in Hoofddorp-Noord could be reduced by increasing the sound insulation provided by the house structures. It is generally not difficult to achieve 5 dB or more improvement in noise reduction in most houses exposed to aircraft flyover noise. Standard techniques include installing acoustical windows and doors, interior wall treatments, increased insulation in the attic space, and generally sealing up or baffling any open paths of noise transmission. However, these treatments are more effective at medium and high frequencies than they are at low frequencies.
14
At frequencies greater than about 160 Hz the noise reduction of building elements tends to increase with increasing frequency, and is dependent on three parameters – mass, spacing, and decoupling between elements. The higher the mass, the greater the spacing between interior and exterior wall panels or window panes, and the higher the decoupling between panels, then the higher is the noise reduction. Below about 160 Hz, the noise reduction is compromised by numerous structural resonances which can be shifted in frequency by careful selection of mass and spacing, but they are difficult to eliminate entirely. It is not difficult to increase the noise reduction of a house at medium and high frequencies, but unfortunately, the same treatments have very little effect at frequencies below 160 Hz. Examination of the sound insulation characteristics of typical Dutch brick houses indicates that the glazing weight would have to be tripled to achieve a 5 dB increase, and this is only true for older houses with existing single glazing. Increasing the low-frequency insulation of houses with double glazing would require significantly heavier and thicker window assemblies. Houses with light-weight facades would require replacement of wooden elements with brick, (additional information on designs is contained in a technical annex with this report; available on request). 6.5
Operational Procedures Modifying the aircraft departure procedures is potentially an effective mitigation measure that is achieved by moving the noise source relative to the community. Three possibilities were examined: • Shifting the departure point • Moving the aircraft to a different runway • Implementing a rolling take-off procedure Depending on the runway length required for take-off, an aircraft could change the departure point where maximum thrust is applied further away down the runway and further from the community. The noise levels in the community would be reduced slightly as the distance to the aircraft increases by 250 to 500m. However, this reduction may be negated by an increase in noise level due to the directional characteristics of the noise radiation. The overall effect would be very small at best. Moving the aircraft to a different runway would remove the noise almost completely, and would certainly meet the goal of achieving a 5 to 10 dB noise reduction. Airport operational requirements would limit the extent of such a measure, but it could be implemented for certain aircraft types (larger aircraft) and under specified weather conditions (wind from the NE, or with standing water on the farmland). However, this option is in conflict with the current Schiphol Act which specifies that the Polderbaan is the only runway for night departures in a northern direction. There exists no flexibility in the current system to move aircraft to other runways, since this will also displace the noise load and will cause breaches of the legal noise restrictions.
15
Finally, the departure operation was examined in more detail to determine if modifications to the start-of-takeoff-roll would provide any reductions in noise levels. Using the video recordings taken during the measurement survey, it was noted that some aircraft applied takeoff thrust from a stationary position whereas others employed a rolling start. Although there was evidence that the rolling start generated slightly lower noise levels in the community for a few events, there was not statistical evidence to conclude that this effect was significant. 6.6
Long-Term Mitigation Measures Included in the list of selected mitigation measures were two methods that have the potential for reducing noise levels, but were considered of low priority because they are conceptual in nature and are not considered proven technology. Active noise control, or “anti-noise”, involves the use of loudspeakers to generate noise in anti-phase to the noise from the Polderbaan. Initial experiments conducted under a limited range of conditions have shown that noise levels can be reduced by 5 to 10 dB. However, the extent of the region over which the reduction is obtained has not been fully studied, and there is always the possibility that levels in some regions could be slightly increased. Further study is required to develop this technique for field application. The modeling conducted in this project has shown that wind speed and direction are critical in determining noise propagation. In particular, it has shown that noise levels in the community are reduced significantly when the wind is from a southerly direction. This result leads to the concept of artificially producing a southerly wind using large fans or wind turbines. A large number of fans would be necessary to generate a sufficient wind over the entire community, and unfortunately, these devices produce lowfrequency noise of their own which would not make them ideal neighbors in the community. Further study is required to investigate the feasibility of this measure.
16
7.0
CONCLUSIONS In the evaluation of mitigation measures the following factors were taken into consideration: • Magnitude of the attenuation • Relative costs • Feasibility • Proven technology
The conclusions from the evaluation of mitigation measures are as follows: • Barriers
•
-
For limited barrier heights (10 m for barriers near runway, 15 m near Hoofddorp-Noord) the estimated attenuation is 6 dB which does not meet the project goals over an extended area.
-
Project goals can be approached at houses very close to a barrier.
-
A high barrier close to the runway would impact on aircraft safety, and would not provide attenuation when the aircraft is airborne.
-
Barriers are proven technology and can be effective for mitigating road noise.
-
A barrier in excess of 10m in height would need to be structurally massive to withstand wind loads.
-
A barrier of 10 to 15m in height could be added to the dike separating the houses from the farmland.
-
The acoustic performance is reduced in the presence of wind.
-
High-rise buildings located between the runway and the community are a possible alternative to barriers near the runway or the houses.
Ground absorption control -
Increasing the ground absorption in the farmland could provide an additional attenuation of 5 to 10 dB.
-
The noise absorption of the ground can be increased by artificial measures, such as an extended gravel pit 1 to 1.5m in depth. The effectiveness of this measure has been demonstrated in small-scale experiments only, and hence is not proven technology.
-
Alternatively, ground absorption can be increased by adding natural vegetation, such as a forest which after many years develops an accompanying thick humus ground layer. This is proven technology.
17
-
The lateral and longitudinal extent of these measures necessary to meet project goals needs to be studied.
-
Further study is required to determine the optimum combination of ground absorption and barrier design/location.
• Operational measures -
Moving traffic to other runways is a most effective measure that could provide 5 to 10 dB noise mitigation.
-
Traffic could be diverted on the basis of aircraft type, weather conditions that support favourable propagation, and time of day.
-
The impact of noise from communities must be studied.
-
Modifications to the Schiphol Act would be required to implement this measure.
diverted
aircraft
on
other
• Improvement of isolation at dwellings -
Requires three times heavier glazing for single-glazed windows (not realistic for any modern Dutch house).
-
Requires the replacement of the light-weight facades by heavy facades for the houses in Houtwijkerveld.
-
The modifications would be expensive and in many cases would affect the exterior appearance of the houses. This measure is not cost/effective in existing houses.
-
Additional sound isolation would not affect the outdoor noise environment.
• Long-term measures -
The implementation of active noise control (antinoise) could conceptually meet project goals, but the technology has not been demonstrated on a large scale, and hence is not proven.
-
The effectiveness of creating a wind to mitigate noise has never been demonstrated on a large scale. This is not proven technology.
18
7.1
Summary – Conclusion-matrix The table below provides a matrix of conclusions on these factors for each measure, with color-coded qualitative results for the criteria Effective, Feasible and Mainport. Green indicates “good”, orange “potential” and red “bad” results from the qualitative analysis as described in this report.
Workpackage
Effective (accoust.)
Ground Absorption
Gravel/Grass Forest floor
Barriers
Near Rwy a/c safety Near Houses Office blocks?
Insulation
In theory
Operational Measures
Feasible
tbd
Not in existing structures
Remove a/c Legal (2010) adverse cond CROS pilot
Long term measures
PM
Tbd
– to be determined
a/c
– aircraft
PM
Mainport
tbd tbd “polluter pays” tbd PM
cond – conditions PM
– Pro Memory: not yet analyzed in detail
These conclusions can be used to define actions for the next implementation phase.
19
References for Low Frequency Noise Criteria x
Sutherland, L.C., Sharp, B.H. and Mantey, R.A., “Preliminary Evaluation of Low Frequency Noise and Vibration Reduction Retrofit Concepts for Wood Frame Structures”, Wyle Research Report WR 83-26, June 1983
x
Findings of the Low-Frequency Noise Expert Panel of the Richfield-MAC Noise Mitigation Agreement, Annotated Report, September 2000
x
Stephens, D.G., et al., “Guide to Evaluation of Human Exposure to Noise from Large Wind Turbines”, NASA Tech. Memo 83288, March 1982
x
Tokita, Y. and Nakamura, S., “Frequency Weighting Characteristics for Evaluation of Low Frequency Sound”, Proceedinga, 1981 Inter. Conf. on Noise Control Eng., Amsterdam, The Netherlands, 6-8 Oct. 1981
x
Yeowart, N.S. and Evans, M.J., “Threshold of Audibility for Very LowFrequency Pure Tones”, J. Acoust. Soc. Am. 55, 814-818, 1974
x
Kamigawara, K., Yue, J., Saito, T. and Hirano, T. “Publication of Handbook to Deal with Low Frequency Noise (2004)”, Proceedings of Low Frequency 2004, Maastricht, The Netherlands, 157-161, 2004
x
ISO 226: Acoustics – Equal Loudness contour for otologically normal listeners, Part 1.
x
NSG: Nederlandse Stiching Geluidhinder: Guideline for Low Frequency Noise (Richtlijn Laagfrequent Geluid), Delft 1999 (in Dutch)
x
Frits (G.P.) van den Berg, “Low Frequency Sounds in Dwellings: A Case Control Study”, Journal of Low Frequency Noise, Vibration and Active Control 19, 2, 59-71, 2000
20
Conclusies, vervolgstappen en aanbevelingen Conclusies In fase I, uitgevoerd 1e helft 2004, bleken gemeten geluidsdata uit het NOMOS meetsysteem van Amsterdam Airport Schiphol niet overeen te komen met de zelf gerapporteerde en de door CROS geregistreerde hinder van de bewoners. De verklaring hiervoor is dat NOMOS vanwege de standaard geluidsinstellingen het laagfrequente geluid niet meet. In fase II, uitgevoerd in november 2004, zijn meer gedetailleerde metingen van geluid en trillingen met de nadruk op lage frequenties uitgevoerd in een tweetal woningen in Hoofdorp-Noord. De resultaten toonden aan dat tijdens starts van de Polderbaan met grotere vliegtuigen (waaronder DC10, MD11, B747, A330) aanzienlijke hinder kan optreden onder bepaalde omstandigheden. Er werd geconcludeerd dat nadere informatie nodig was over de laag frequente geluid/trillingenbron en het overdracht mechanisme, alvorens maatregelen voor reductie van geluidniveaus onderzocht konden worden. Uit de 1e modellering in fase III, uitgevoerd in voorjaar 2005, bleek het mogelijke overdrachtspad voor trillingen door de grond veel te zwak om de waargenomen trillingen in de testhuizen te kunnen verklaren. Hieruit werd geconcludeerd dat de waargenomen geluid- en trillingsniveaus, veroorzaakt werden door het laagfrequente motorgeluid van (grotere) startende vliegtuigen, welke voornamelijk door de lucht worden overgedragen. De hoogst gemeten waarden in het laagfrequente gebied lagen rond de 31.5Hz. Naast de geluidsfrequentie werd ook geconstateerd dat de mate van hinder sterk beinvloed wordt door factoren zoals weersgesteldheid (windrichting en lucht temperatuur en de bouwtype van de huizen (stenen- of houtbouw). Tenslotte is er een directe correlatie geconstateerd tussen de intensiteit van het gebruik van de Polderbaan en de hinder die wordt ervaren. Met uitgebreide geluid- en trillingsmetingen in augustus 2005 zijn deze voorlopige conclusies vervolgens gevalideerd. Er is gemeten op diverse locaties en hoogtes in en nabij de startbaan, in het tussengebied en nabij huizen in Hoofddorp-Noord. Ook is in een woning gemeten. De verwerkte resultaten van starts van een verschillende types grotere en zwaardere vliegtuigen bleken zeer goed overeen te komen met de voorspelde waarden uit de modellering. Van alle gemeten waarden bleek de DC10 ruimschoots de hoogste belasting bij lage frequenties te veroorzaken, waarbij de MD11 en B747-400 tot 5dB lager scoorden. Ook werd geconcludeerd dat onder bepaalde omstandigheden, zoals matige (noord)oostelijke wind en overige atmosferische condities, de geluidsniveaus in Hoofddorp met 5 tot 10 dB kunnen toenemen ten
opzichte van andere condities, bij gelijke bronsterkte van het vliegtuig. Vervolgens is deze dataset in oktober 2005 gebruikt in stap 3 om het eerder ontwikkelde model te valideren. Daarbij is aandacht besteed aan de sterkte en richtingsgevoeligheid van de geluidsbron, de meteorologische omstandigheden en de dempingeigenschappen van de grond. Het gevalideerde model toont aan dat de luidste uitstraalrichting 45 graden naar achteren is, ofwel direct richting Hoofddorp bij starts van de Polderbaan. Tevens werd het eerder gemeten effect van windsnelheid en richting bevestigd. Het model heeft bij “wind mee” condities het dempend vermogen van de grond in kaart gebracht. Bij een akoestisch zachte grond blijkt het effect van “wind mee” beperkt (< 5 dB), maar bij een akoestisch harde grond is het geluidniveau hoger, dit kan zelfs meer zijn dan 10 dB. Wereldwijd is beperkt onderzoek naar hinder door laagfrequent geluid beschikbaar. Bij dit onderzoek is door TNO, NLR en Wyle laboratories een literatuurstudie gedaan naar deze internationale onderzoeken. De uit deze eerdere onderzoeken bekende relaties voor hinderbeleving per individueel ‘event’ zijn als basis gebruikt voor drie frequentie afhankelijke criteria: waarneembaar/acceptabel, vervelend/hinderlijkk en zeer hinderlijk/zwaar belastend. Hiertegen zijn vervolgens de in dit onderzoek werkelijk gemeten geluidsniveaus onder diverse omstandigheden uitgezet, waarbij ze ook zijn vergeleken met de door bewoners zelf gerapporteerde hinder. De metingen uit fase II en III gecombineerd met de zelf gerapporteerde hinder blijken consistent met deze criteria. Om hogere geluidsniveaus dan die gemeten in augustus 2005 (geen gerapporteerde hinder) te voorkomen dient in de meest ongunstige omstandigheden (vanuit hinderbeleving bewoners) een reductie van 5 tot 10 dB te worden gerealiseerd, waarbij de meest dominante frequenties in de buurt van 31.5 Hz liggen. Hiermee is een meetbare reductiedoelstelling voor effectieve maatregelen afgeleid. In de laatste stap van het project is in een gezamenlijke brainstorm van de technische begeleidingscommissie een hele serie aan mogelijke hinderbeperkende maatregelen opgesteld. Deze zijn getoetst op hun effectiviteit door te bepalen of zij een reductie van (minimaal) 5 tot 10 dB kunnen verzorgen. Hierbij zijn diverse types geluidswallen en gebouwen, maatregelen die demping door de grond en/ of begroeiing in het tussengebied bevorderen, extra isolatie van woningen, operationele procedures en mogelijke toepassing van anti-geluid en luchtverstorende maatregelen in beschouwing genomen. Deze zijn tevens op hoofdlijnen getoetst op de criteria “haalbaarheid” en “effect op mainport”.
Op basis van deze criteria zijn de oplossingsrichtingen als extra isolatie, toepassing van anti-geluid en luchtverstorende maatregelen als niet effectief, en/of (voorlopig) niet haalbaar afgevallen als realistische maatregelen voor de korte termijn Meest kansrijke maatregelen zijn te vinden in: verbeteren grondabsorptie van het tussengebied en het plaatsen van hoge obstructies (wal of gebouwen) van vijftien tot dertig meter of hoger tussen de Polderbaan en de woonkernen. Operationeel kan het beperken van starts van de Polderbaan onder bepaalde (weers)omstandigheden mogelijk worden “getest” tijdens de pilots van de Commissie Regionaal Overleg luchthaven Schiphol (CROS). Verplaatsing van het vliegverkeer kan wel worden uitgezocht, maar niet ingevoerd zolang het huidige wettelijke milieusysteem voor Schiphol geldt. Het verbeteren van grondabsorptie is op beperkte schaal in het model getoetst. Grote oppervlakten, met grind of beplanting bijvoorbeeld, zullen nog nader op effect moeten worden bekeken, wellicht in combinatie met een wal of hoge gebouwen.
Vervolgstappen en aanbevelingen In het regulier overleg tussen Bewonersvereniging Hoofddorp-Noord, gemeente Haarlemmermeer, Schiphol Group en het CROS-bestuur is door partijen besloten dat de meest positief scorende maatregelen verder uitgewerkt gaan worden, op zichzelf staand dan wel in combinatie, waarbij tevens een primaire probleemeigenaar is aangewezen die de volgende (implementatie) fase zal gaan trekken. De gemeente Haarlemmermeer zal binnenkort de initiatieffase van een project starten om de resultaten van het onderzoek naar grondgeluid en de meest kansrijke maatregelen (grondabsorptie en wal/gebouwen) in te brengen in lopende projecten in het betreffende gebied. Uiteraard zal bij de vervolgstappen ook rekening gehouden moeten worden met een wetenschappelijke validatie achteraf van het effect van de genomen maatregelen en de bereikte hinderreductie. Inwoners van Hoofddorp-Noord worden aan hinder van startende vliegtuigen vanaf de Polderbaan blootgesteld. Laagfrequent geluid (geluid en trillingen) door vliegtuigen speelt hierbij een belangrijke rol. Uit het onderzoek blijkt dat het laagfrequente geluid van starts met bepaalde grote vliegtuigen vanaf de Polderbaan extra wordt versterkt bij bepaalde atmosferische omstandigheden.
De Commissie Regionaal Overleg Luchthaven Schiphol zal de Staatssecretaris van Verkeer en Waterstaat melden dat zij de werkgroep Hinderbeperking verzoekt in het kader van pilots c.q. experimenten maatregelen uit te werken die in combinatie met een gewijzigd baan– en routegebruik het grondgeluid bij Hoofddorp-Noord effectief kan verminderen. Het is belangrijk dat de Bewonersvereniging Hoofddorp-Noord hier goed bij aangesloten blijft. Dit onderzoek heeft een gevalideerd model opgeleverd dat de invloed van laagfrequent geluid door individuele vliegtuigen op een specifieke omgeving in kaart kan brengen. Op basis van dit model kunnen klachten en meldingen over grondgeluid rond andere banen van de luchthaven Schiphol nader worden onderzocht. Ook daarvoor kunnen via het model eventueel maatregelen op effectiviteit worden berekend en nader uitgewerkt. Instanties die een rol spelen bij hinderbeperking van het vliegverkeer op de luchthaven Schiphol kunnen hier nu en in de toekomst hun voordeel mee doen
