Noise mapping on a European scale as a basis for cost benefit analyses of different noise control options Paul de Vos MSc, AEA Technology Rail bv, P.O. Box 8125, NL 3503 RC Utrecht, The Netherlands Summary Strategic noise mapping for major railways will become compulsary under the new EU legislation, i.e. Directive on the Assessment and Management of Environmental Noise. With the right system in place noise mapping is an easy job, and moreover it can provide very usefull information to the decision makers in developing railway noise strategies. The Eurano tool which was used for e.g. UIC funded studies is a good example of such a decision support system. Strategic Noise Mapping The European Commission has published it s draft Directive on the Assessment and Management of Environmental Noise(COM (2000) 468 final). This Directive, among others, aims to determine exposure of the European citizens to noise through strategic noise mapping. Such strategic noise maps will have to be drawn using noise indicators (L den and L night ) and assessment methods common to all Member States. For so-called major railways, i.e. railway links with an annual number of more than 30,000 trains passing (NB. this relates to an order of magnitude of 4 trains per hour!), strategic noise maps shall be prepared and provided to the Commission within 5 years after the Directive has come into force, and every 5 years after that date. For the first deadline, only railway links with more than 60,000 trains passing need to be considered as major railways.
Figure1
Figure 1 present the major railway lines in The Netherlands according to this definition. These maps will allow assessment of an existing, historical or future noise exposure, the excess of a limit value, the estimated number of dwellings, schools and hospitals in a certain area which are exposed to certain values of the noise indicators, and the estimated number of individuals in a certain area exposed to noise. The noise maps will serve as a first step towards noise action plans, where noise abatement on the basis of national noise reception limits will enter into the discussion. The obligation to prepare and supply (or alternatively to approve of) the maps rests on the competent authorities. The total cost for initial noise mapping of major roads and railways is estimated by the Commission to be between 50 and 75 million EUR, which is assumed (again: by the Commission) to represent an increase in cost of 10 to 20 million EUR compared to the present situation, where noise maps are already being made in several Member States. Preparing noise maps Noise maps as specified in the previous section shall be prepared on the basis of computational methods only; measured noise levels will be considered irrelevant. The computational methods to be used will be harmonised methods, common in all Member States. A fifth Framework project aiming to develop such harmonised methods has started recently [1]. Untill these methods are available, preferred interim methods shall be applied. For railway noise, the existing Dutch method is suggested as interim method. Alternatively Member States can continue to use their own national methods, provided that these are able to produce the harmonised indicators. These indicators represent long term average values of the noise reception level. Figure 2
In principle, all these computational methods start with three input databases (figure 2): a traffic database, containing all the traffic data which is relevant for the calculation of the noise creation at a certain point, describing the type of trains and their intensity (i.e. the number of vehicles passing the receiver point under concern per hour), and their operating conditions (speed, brakes on or off), an infrastructure database, containing all the infrastructure data which has an influence to the noise creation, i.e. the type of track, the dynamic properties of the substructure, the maintenance condition of the rail head, number of joints and switcher per m, etc. a noise attenuation database, containing all the relevant data on the surroundings of the track, such as: the distance between track and receiver, the condition of the ground surface, the presence of screening or scattering elements, the weather conditions (downwind or upwind or average). Once the three sets of data as described above are available for three periods of time covering a timespan of 24 hours (i.c. average day, average evening and average night), the harmonised indicators L den and L night can be assessed and the results plotted on a map or in a table (output database). The origin of these three databases may differ significantly; the traffic database for railway lines is usually based on the time table or service plan. Compared to noise mapping for major roads, the traffic data for railways is much easier to collect and access than that for road traffic. However, due to disturbances substantial differences with reality may occur. Also, data may vary significantly from one year to another. The infrastructure database is usually more stable. Reliable information on the maintenance condition of the track may be difficult to acquire. The information on the surroundings is often available from digital topographical maps and is not subject to large or frequent variation. Responsibilities for railway organisations The Directive does not define any specific responsibility for railway bodies such as infrastructure managers, railway regulators or operators. In the old constitution, when monolithic railway companies were controlled by national governments, it was obvious that the authorities competent for strategic noise mapping (i.e. the environmental national Ministeries) could dispose of the necessary data from these state railways. At present, railway operators may be reluctant to provide the information, since it may be sensitive to their competitors. In some Member States even the infrastructure management is turned into a private enterprise (or many competitive suppliers exist) and there even the infrastructure data may be classified. In any case it is up to the national authorities to appoint responsible bodies who will collect, transfer and make available all the necessary data for noise mapping. It is to be expected and to be preferred - that the railway organisations will be involved in that. Wherever railway organisations operate as private enterprises, it is also to be expected that they will require to be financially reimbursed for the effort related to their involvement. The organisational set up required for strategic noise mapping in the first instance and for the definition and execution of noise action plans in the second phase will become clear in the next few years and may differ from one Member State to another. The Dutch Noise Emission Register Since the early nineties AEA Technology Rail has composed and maintained a database under the title Dutch Noise Emission Register, which conforms almost completely with the input
databases for traffic and infrastructure required for strategic noise mapping. This Register was set up and is maintained under assignment of Railinfrabeheer, the Dutch state controlled body responsible for infrastructure management. Railinfrabeheer owns and provides all the data on track construction types. Railned, the Dutch state controlled capacity regulator, owns and provides the data on actually achieved traffic intensities and planned services. These two datasets are converted, de-bugged, linked and form the basis for the Register. The Register is analysed on a yearly basis by AEA Technology Rail and the findings are reported to the national Ministery of Environment, under shared responsibility of Railinfrabeheer and Railned. Thanks to the high amount of detail of the data, it is possible to indicate noise increases or reductions per single line. In addition to this, the converted input database is made available (on a CD-ROM) on a cost price basis to the public. Any user, such as noise consultants, municipalities, urban planners and building contractors, is obliged to use this particular database whenever he needs to assess an actual or future noise level anywhere near a railway line. In practice it is most efficient to avoid discussion on the reliability of the input data. The database is sold as Acoustic Time Table, but it has been developed further so to allow indicative noise and barrier efficiency calculations, geographic front end (Figure 3) and other features. Figure3 Figure 3 presents the geographic front end with access to the track database. The combination of obligatory data collection of the Register on the one hand and supply of the database to a wide
range of users on the other hand, both under responsibility of state controlled organisations but carried out by a commercial organisation linked closely to the railways, has proven to be advantageous to all parties involved. Particularly the price to quality ratio can be kept high thanks to the chosen division of responsibilities. The data is available for a range of different datum years in history and future: 1987 data forms the reference for noise legislation in The Netherlands, being the year when the noise control act for new and significantly altered railway lines entered into force, yearly averages and three year averages (smoothening the effect of drastic changes from one year to another) are available from 1994 onwards, realisation figures are available up to the year 2000, future estimates are available for 2010/2015 (prognosis). The realisation figures are presently under discussion, because these data are based on planned operational paths rather than actual train movements. An operator may have claimed a path for a certain link and time, without actually using it. As a consequence the calculated noise exposure may be somewhat overestimated. Real monitored traffic data is envisaged to be included into the system soon to overcome this discussion. Manipulating the databases An additional advantage emerges for those parties who have access to the databases; apart from the range of datum years mentioned above, one then has the option to simulate certain scenarios and study their effects with respect to noise exposure. Such scenarios may refer to the traffic intensity and speed, but also to retrofit noise abatement in existing fleets or replacement of noisy vehicles by quiet ones. Figure4
The link of that sort of manipulated input data with the geographic information that is acquired from plotting the output noise contours onto a map, allows quick and easy assessment of the effects of such scenarios. The special version of the Register which allows manipulation of the input databases and mapping of the results was developed by AEA Technology Rail in the mid nineties under the name Gerano (GEographic RAilway NOise tool). In recent years the tool has been improved and extended, focussing on a wider range application. Railway traffic and infrastructure data for a range of railway lines in France, Germany, Switzerland and Italy were added, and the computation method for the train type characteristics and noise propagation calculation were adapted to the existing methods in those countries. The result of these extensions is Eurano. Figure5 Figure 5 presents a geographic section of Eurano, showing urban areas (pink), isolated houses (bleu), noise contours (red, blue and green) and barrier heights. Like it s predecessor Gerano, this tool is able to provide geographically oriented output on the basis of manipulated input data. The output data involved could for instance be: the length and height of noise barriers required to comply with certain noise regulations, the number of residents or dwellings above a certain maximum noise exposure, the surface of the area above a certain maximum noise exposure.
The combination of manipulated input data and geographically oriented output data allows the execution of cost to benefit comparative studies. This is an application that has found great interest and wide spread over the last years. Cost benefit studies Cost benefit studies have been carried out during the past years, focussing on different magnitudes of scale. For one particular line in The Netherlands, i.e. Amsterdam-Utrecht undergoing an extension from 2 to 4 tracks, a first study was carried out in 1998, indicating that high investments in noise barriers could be avoided if noise reduction at source (at the vehicle and the track) were to be achieved. This first study showed an important complexity of this type of studies: since rail transport is to a certain extent international transport (particularly freight), when simulating the cost of retrofit low noise vehicles one has to come up with a representative estimate of the total number of vehicles to be treated in Europe. Hypothetically, in order to achieve a guaranteed noise reduction at one particular site in The Netherlands by retrofitting vehicles for international use, one would have to treat every single vehicle homologated for the European network. Thus the cost of noise reduction at source would be astronomic, but the effect would also be extended to a far larger area than just the one site under concern. It is therefore fair to assume that the cost estimates get more reliable the larger the area of concern involved in the study. A second study [2] in 1999 covered the whole Dutch network (2800 km of track, of which 1800 km more than 1 track). Assuming a certain development in national noise legislation and a certain growth in rail transport, the study showed that 1010 km of noise barriers would be required and additionally 5000 properties would need to be treated (façade insulation or demolition) in order to comply with the legislation, involving a total amount of LCC of 1.7 billion EUR over a 40 year timespan. In the same study, different alternatives were then considered, mainly focussing on noise reduction at source. For instance: one scenario that was studied assumed a replacement of cast iron brake blocks (either by disk brakes or by composite brake blocks), which was estimated to represent LCC of appr. 250 mio EUR, but could at the same time introduce savings in barrier cost of appr. 630 mio EUR (40 year basis).
Figure5 As an example, figure 5 presents cost for noise reduction at source and noise reduction in propagation (barriers and façade insulation), adding up to total life cycle cost for a 40 years period of time, for 7 scenarios. Similar studies were carried out under assignment of UIC [3], focussing on larger parts of the European railway network in order to increase the representativity of the results. At present, cost benefit studies are prepared and carried out in the fifth Framework project STAIRRS [4]. In this project an even larger scale is obtained by extrapolating results for a wide range of different, typical railway line sections throughout Europe which differ in parameters like traffic intensity, freight percentage, density of built up area near the line, etc. The results of this study will influence the definition of the future railway noise control policy, both of the European Commission and of the railway related organisations. This emphasizes the importance of this type of study, as it may affect the profitability and credibility of rail transport in Europe. Other features In the previous section it was demonstrated that it is advantageous to the organisations related to railways to be involved in the activities of strategic noise mapping and to be able to dispose of the tools. The execution of scenario studies for cost benefit analyses certainly helps decision making units in these organisations. But other advantages may emerge. First of all, noise mapping is a powerfull and convincing method to compare the pro s and con s of different
routings when planning new railway lines. The cost for noise barriers and the number of additional noise annoyed residents may be important criteria for the preference of one particular route over the other. Noise mapping is a quick and easy way to assess these criteria. Moreover, once the mapping tool is available, other adverse environmental effects may be included, such as exhaust gas emissions (diesel traction), ground borne vibrations (leading to annoyance and/or damage to properties) and safety risks emerging from transport of hazardous goods. Conclusions The draft European Directive on the Assessment and Management of Environmental Noise imposes the obligation upon Member States to provide, on a regular basis, strategic noise maps for major railway lines. Although railway related organisations, such as infrastructure managers, capacity regulators and operators, are not officially involved in the process of preparing these maps, it is to be expected that some involvement will be required, e.g. for the assessment of traffic and infrastructure data needed for the noise computations. The effort of collecting and maintaining this information in a well structured and organised way is certainly of value. The ability to manipulate that data enables the user to carry out what-if studies, particularly for cost benefit analyses, which help him to make the right decisions and base his general policy on it. Strategic noise mapping is therefore an activity that is worth setting up in the right organisational structure. The experience in The Netherlands, illustrated here, may serve as an example. References [1] Vos, P.H. de, Opportunities for improved harmonised noise prediction methods in Europe, Proc. Internoise, Den Haag, 2001 [2] Beek, A.J. van and Boer, L.E. de, Savings by noise abatement at source for trains and tracks (in Dutch, English translation available), Proceedings National Congress on Noise and Vibration, Rotterdam 1999 [3] Haaren, R. van, Economic study on railway noise: Effectiveness of different source reduction configurations, Proc. Internoise 2000, Nice, August 2000 [4] Oertli, J., The STAIRRS project, A cost benefit analysis of railway noise reduction on a European scale, 7th International Workshop on Railway Noise, Portland (Maine), 2001