CONTROL OF LARGE SCALE NOISE IMPACT OF RAILWAY LINES: OVERVIEW OF RESULTS IN THE NETHERLANDS AND EUROPE

 

 

Frank Elbers

NS Technisch Onderzoek

PO Box 8125

3503 RC Utrecht, The Netherlands

Phone: +31 30 2358128, fax: +31 30 2359150

e-mail: f.b.j.elbers@geluid.nsto.ns.nl or eurano@ns.nl

 

INTRODUCTION

Transport policy tend towards an increase of rail transport. There is a need for (a more stringent) noise legislation because of a growing awareness of local communities to railway noise. When noise legislation is being enacted the cost for noise measures becomes a major economic factor for the railways. Optimisation of noise control strategy will lead to an increase of large scale environmental benefit and a decrease of cost for noise measures.

From 1995 different large scale noise control studies are carried out in The Netherlands, Switzerland and for Europe. All studies lead to the conclusion that noise control by barriers only lead to tremendous cost. Mostly these cost are not within available budgets of national governments and railway companies. A change in railway policy is needed to support an increase of railway transport and to limit the large scale noise impact.

In this paper results are presented of six large scale environmental studies carried out in The Netherlands, Switzerland and for Europe.

 

OUTLINE OF THE STUDIES

1. Noise creation quotas will limit the environmental impact (The Netherlands)

In 1998 a NS Railinfrabeheer (Infra management), Ministry of Environment and Railned (Capacity management) supported study was carried out by NS Technisch Onderzoek in co-operation with RIVM [1]. The goal was to quantify the effect of noise creation quotas on the entire Dutch railway network with a total line length of 2.800 km. An adaptation of the Dutch noise legislation with noise creation quotas is preferred, to prevent an actual allowed limited growth of transport by about 20 % per year. Calculations where done with Gerano for the entire Dutch railway network [2, 3] with a total line length of 2.800 km. Results are given in figure 1.

The studies concludes that an increase of rail transport to the year 2010 with noise creation quotas based on the level of 1997 will cost 350.000 Euro per km (net present value). This is needed for 600 km of noise barriers and window insulation of 1780 houses. Source control measures like a 7 dB(A) improvement of tread braked national passenger and freight trains and an early replacement of wooden sleepers by concrete sleepers (- 2 dB(A)) will reduce barrier and window cost to 32.000 Euro per km.

The number of people annoyed by railway noise in 1997 is 241 per km. Without additional noise measures an increase of rail transport to 2010 will lead to 286 people annoyed per km (+19 %). Calculated noise control measures will limit this number to 198 (-18 %). A future scenario with source measures without any additional barriers will result in 168 annoyed people per km (-30 %). Additional barriers to prevent an increase of noise relative to 1997 will lead to 160 people annoyed per km (-34 %).

This study concludes that noise measures at the source will save enormous cost for barriers and window insulation. An additional benefit (not included in the Dutch noise legislation) is a decrease of surface with a noise impact above 55 dB(A) (-30 %).

Figure 1 Net present values for noise barriers and window insulation for future noise legislation in The Netherlands. The range indicate the cost for short term (disc brake) and long term (K-block) implementation.

 

2. Noise reduction at the source will save cost (The Netherlands)

In 1999 a Ministry of Environment supported study was carried out by KPMG BEA and NS Technisch Onderzoek in co-operation with and RIVM [4]. The goal was to quantify the total future cost of noise measures for different noise control scenarios. Starting point was the actual noise creation, modified legislation with noise creation quotas, the transport volume up to 2010 and additional measures to prevent noise reception above 57 dB(A). Calculations where done with Gerano for the entire Dutch railway network. Results are given in figure 2.

If no source measures are introduced until 2010, noise barriers and window insulation will cost 600.000 Euro per km (net present values). A 7 dB(A) reduction for all passenger trains with cast iron tread brakes will save about 55.000-100.000 Euro per km on noise measures. The range indicate the cost for short term (disc brake) and long term (K-block) implementation. Cost only include adaptation of national rolling stock. Foreign rolling stock is assumed to be improved also but cost are not accounted. A 7 dB(A) reduction for all freight wagons will save about 90.000-130.000 Euro per km on noise measures. Improvement of both passenger and freight wagons (with cast iron tread brakes) will save 144.000-229.000 Euro per km. Additional measures to these passenger and freight wagons like wheel dampers will have an extra reduction of about 1-2 dB(A) and will save 102.000-187.000 Euro per km. A very large set of source measures like a 7 dB(A) improvement of all cast iron tread braked trains, wheel dampers, track improvement and close to track barriers and bogie shrouds will lead to high cost for source measures. This scenario will cost 195.000-280.000 Euro per km more than the reference scenario without any source measures.

Without additional noise measures railway noise will cause about 286 people annoyed per km in The Netherlands for the year 2010. Above mentioned scenario with barriers and window insulation only will reduce this number to 169. The surface with a noise impact above 55 and 65 dB(A) is 1,1 and 0,2 km2 per km. For the most cost effective scenario with wagon improvement and wheel dampers, the surface with a noise impact above 55 and 65 dB(A) is 0,6 and 0,08 km2 per km.

This study concluded that above mentioned noise legislation (quotas based on 2000 and a 57 dB(A) limit) in combination with the present rolling stock (=no source measures) will lead to cost for noise barriers and window insulation of 600.000 Euro per km. In general noise measures at the source will save enormous overall cost (up to 40%). An additional benefit (not included in the Dutch noise legislation) is a decrease of surface with a noise impact. A 7 dB(A) noise reduction of freight trains outside The Netherlands will save 70.000-115.000 Euro per km on cost for noise barriers and window insulation.

Figure 2 Net present values for noise control measures for future noise legislation in The Netherlands.

 

3. Total barrier cost for noise reduction to 65 dB(A) (The Netherlands)

In 1999 a Ministry of Environment supported study was carried out by NS Technisch Onderzoek [5]. The goal was to quantify the total cost of noise measures for different noise control scenarios. Starting point was the actual legislation to prevent 65 dB(A) and the transport volume up to 2010. Calculations were done with Gerano for the entire Dutch railway network with a total line length of 3.000 km. Results are given in figure 3.

This study focused on 4 scenarios:

  1. No source reduction.

  2. Replacement of all wooden sleepers with concrete sleepers (-2 dB(A)).

  3. Replacement of all rails with wooden sleepers with rails with concrete sleepers and dampers (-5 dB(A)).

  4. Replacement of all wooden sleepers with concrete sleepers (-2 dB(A)) and 7 dB(A) source reduction of all national passenger and freight trains with cast iron tread brake.

Figure 1 gives the results.

Figure 3 Cost for noise barriers to prevent a future noise level above 65 dB(A) based on the entire Dutch railway network (2.800 km).

This study concludes that:

 

4. Total barrier cost for noise reduction to 57 dB(A) (The Netherlands)

In 1999 a NS Railinfrabeheer supported study was carried out by NS Technisch Onderzoek [6]. The goal was to quantify the total extra cost of noise measures for different noise control scenarios when the noise limit will decrease to 57 dB(A). Starting point was the actual legislation to prevent 65 dB(A) and the transport volume up to 2010. Calculations were done with Gerano for a part of the Dutch railway network where future reconstruction is planned, with a total line length of 415 km. Results are given in figure 4.

Figure 4 Cost for noise barriers to prevent a future noise level above 57 or 65 dB(A) based on the Dutch railway network were future reconstruction is planned (415 km).

A general noise reduction to 65 dB(A) at the facade wall of houses will lead to cost of 250.000 Euro per km if no source measures are implemented. These cost will increase with 37-40% if the limit value will be decreased to 57 dB(A).

 

5. Noise control strategy for Switzerland

To comply the Swiss noise ordinance, noise measures (barriers and window insulation) for 3.000 km line length of Swiss railway network, will cost 4.360*106 Euro. With these measures there will be about 269.000 people above the reception limit (55-70 dB(A)) and 26.000 people above the alarm value (65-75 dB(A)) for the year 2005. Calculations were done with the Lärmbelastungskataster.

There was no financial basis to comply the noise ordinance and the foreseen train capacity. In June 1996 the Swiss railway finished a research project [7], where the possibilities are shown to reduce the railway noise with increasing cost. Scenarios with source measures where part of the study also. For the preferred scenario with source measures, the total cost of measures is 1.045-1.455*106 Euro. With these measures there will be about 116.00 people above the reception limit and 800 people above the alarm value for the year 2005. With a referendum on November 1998 the people agreed to spend less measures than forced by the ordinance. Therefore not desired cost will be saved and a general decrease of annoyance will be realised also.

For the scenario with source measures for Swiss trains only, the total cost for noise measures is 1.059*106 Euro [8]. These cost will be made for:

When freight wagons of other European countries will have source measures too, Switzerland will save 409*106 Euro on noise measures (barriers and window insulation).

Until the year 2005 the effort to implement noise measures are directed toward low noise trains. After 2005 noise barriers and facade wall insulation will be placed. The amount of the last two measures is dependent on the success of noise reduction on trains outside Switzerland.

 

6. Cost Benefit analysis for European rail noise abatement (Europe)

A UIC sponsored project studied the cost and benefits of different noise control strategies for a total line length of 1667 km [9, 10]. Two major freight transit lines were analysed: Rotterdam-Basel-Milano and Bettembourg-Lyon. Calculations were done with Eurano [11, 12]. Results are given in figure 5.

The reference wagon scenario tested had no wagon improvement. Two additional wagon scenarios had a 5 and 10 dB(A) freight wagon improvement (for example K-block). Each scenario was tested for 3 different noise legislation scenario. The reference legislation scenario had window insulation above 60 dB(A) only. Two additional legislation scenarios had optimised barriers with a maximum height of 2 and 4 metres when a threshold value of 60 dB(A) was exceeded. Additional to this, cost for window insulation were accounted, when measures would not reduce noise levels below 60 dB(A). Two additional scenarios with 5 dB(A) track improvement only and with track improvement and 10 dB(A) freight wagons improvement were tested.

This study concludes that:

 

Figure 5 Cost for noise measures to prevent future noise levels above 60 dB(A).

 

7. European economic study on railway noise reduction measures (Europe)

A EC/ERRI sponsored project studied the cost and benefits of different noise measures for a total line length of 1177 km [13, 14]. One major freight transit lines was analysed: Rotterdam-Basel-Milano. Calculations were done with Eurano. Results are given in figure 6.

Seven source reduction scenarios were tested:

  1. Reference situation without any additional measures (ORE 920 wheel and tread brakes for freight wagons).

  2. Brake block solution with the EuroSabot LL brake block (-5 dB(A)).

  3. Low cost retrofit solution with ring damper and tuned rail absorber (-6 dB(A)).

  4. Advanced retrofit solution with ring damper and tuned rail absorber.

  5. Maximised retrofit solution with ISVR 860 wheel and tuned absorbers, tuned rail absorber and shield and low barriers.

  6. New track solution with ISVR 860 wheel and tuned absorbers, optimised rail shape and tuned rail absorber.

  7. Maximised new track solution with ISVR 860 wheel and tuned absorbers, optimised rail shape, tuned rail absorber, EuroSabot K block and shield and low barriers.

Each scenario was tested without any legislation that forced noise barriers and window insulation. Additional scenario 1, 2, 3, 4 and 6 were tested with a 55 and 65 dB(A) noise limit above which noise barriers and window insulation was forced.

This study concludes that:

Figure 6 Cost for noise measure to prevent future noise levels for different noise limits.

 

 

CONCLUSIONS

The overall conclusion of the studies described above is:

 

Note: The number of people annoyed between [1, 4] and the other studies is calculated with a different method and different data for population density. Results for comparable situations can vary a factor 3 to 7. This inaccuracy is not important to compare different scenarios within one study or between calculations done with the same method as long as cost for window insulation is relative small.

 

REFERENCES

  1. The effects of noise quotas for railway noise (in Dutch: De effecten van de invoering en hoogte van de eerste vaststelling van emissieplafonds), Elbers F.B.J., NS Technisch Onderzoek, October 1998.
  2. Methods and tools for monitoring and prediction of the environmental impact of railway noise, Elbers, F.B.J., Proceedings of IWRN98 (International Workshop on Railway Noise, Ile des Embiez, France, 1998) (to be published in JSV 1999)
  3. Monitoring and prediction of noise from large scale traffic networks with Gerano (in Dutch: Handhaving en planning van geluid van grote verkeersnetwerken met Gerano), Elbers, F.B.J., Geluid, number 5, December 1999.
  4. Quiet trains in a wide perspective (in Dutch: Stille treinen in breed perspectief), Boer L de, and Beek A. van, NS Technisch Onderzoek and KPMG BEA, November 1999.
  5. Prediction of noise barriers and steel bridge measures to prevent 65 dB(A) (in Dutch: Inventarisatie sanering railverkeerslawaai met schermen en stille bruggen), Elbers, F.B.J., Bol, S., NS Technisch Onderzoek, 24 March 2000.
  6. Cost of noise barriers to prevent 65 dB(A) and 57 dB(A) within line sections for future reconstruction (in Dutch: Kosten sanering 65 dB(A) en 57 dB(A)), Elbers, F.B.J., NS Technisch Onderzoek, 6 October 1999.
  7. Lärmsanierungen von Eisenbahnanlagen, Bericht der interdepartementalen Arbeitsgruppe (IDA-E), Bern, 28 June 1996.
  8. Lärmsanierungen von Eisenbahnen, Bericht der interdepartementalen Arbeitsgruppe (IDA-E2), Bern, 26 June 1998.
  9. Cost-Benefit Analysis European Rail Noise Abatement, Oertli, J. (SBB), Elbers F.B.J. (NSTO), Beier, M. (DB-AG), Joncour, S. (SNCF), Hemsworth, B. (ERRI), UIC Task Force Noise, August 1999.
  10. European Rail Noise Abatement: Cost-Benefit Analysis, Oertli, J., to be published in Proceedings of InterNoise 2000.
  11. Eurano99: Noise policy tool for European Railway Network, Elbers, F.B.J., Proceedings of International Railway Conference: Noise abatement, Warszawa, May 1999.
  12. Eurano99: Policy tool for strategy of railway noise, Elbers, F.B.J., Oertli, J., to be published in Proceedings of InterNoise 2000.
  13. European economic study on railway noise reduction measures, Elbers F.B.J., Haaren , E. van, NS Technisch Onderzoek, January 2000.
  14. Economic study on railway noise: Environmental impact of different, Elbers F.B.J., Van der Stap, P.A.B.J., to be published in Proceedings of InterNoise 2000.