Contents. Page 1 of 14

Transcription

1 Contents Guidance on the application of the interim railway noise computation method 1 Summary of input data for computation method (copy of 3.3 of interim computation method) 1.1 Emission data 1.2 Assessment data 2 Guidance on the definition of emission data 3 Guidance on application of the propagation model 3.1 Modelling the situation 3.2 Computation Emission value Attenuation by distance L GU Attenuation by propagation L OD Attenuation factor for screening L S W Determining rail specific absorption Reduction of levels as a result of reflections L R Similarities & differences of ORM with END and Use of ORM for strategic noise mapping 1 Aim 2 Compatibility with END 2.1 Emission values 2.2 Meteo 2.3 Assessment points 3 Compatibility between rmr 1996 & RMR Comparison of ARM 1.5, ORM & ISO methods 4.1 Emissions 4.2 Geometrical noise propagation 4.3 Meteorological terms D air air attenuation D soil soil absorption D meteo meteorological correction 4.4 Reflections 4.5 Screening 4.6 Urban damping (ARM 1.5 method) 4.7 Conclusions Page 1 of 14

2 Guidance on the application of the interim railway noise computation method prepared by AKRON n.v.-s.a. The aim of the interim noise computation method is both the strategic noise mapping along major railway lines and the strategic noise mapping in urban areas. The proposed interim method is based upon the octave band calculation method (ORM) of the Dutch method RMR of 2002 (equivalent to SRM II of RMR-1996). Based on the facts that: o ORM method already includes extensive information about modelling (see 1.3 Modelling the situation ); o the Dutch text about Guidance on the application (chapter 10 of RMR 2002) has been integrated in the specific introductory par of each paragraph of the proposed interim method; o specific adaptation of the Dutch method to the END-terminology has also been integrated in the interim method ( 3.2); the authors believe that little information on guidance should be added. Further, the interim method contains an additional paragraph giving a summary of input data ( 3.3). And concerning emission data, the interim calculation method ( 2) provides also sufficient guidance on that application. On the other hand, comparison with the ISO 9613 computation method is discussed in task part 2. Therefor, this text only contains: o A summary of the required data. o References to the interim computation text concerning the specific items. 1 Summary of input data for computation method (copy of 3.3 of interim computation method) A list of input data requested for the above defined calculation method is summarised hereafter. 1.1 Emission data Track sections If the characteristics of the track, of the rolling stock or of driving conditions depend on the position along the track, different straight track sections are defined by the positions of their outer points. The characteristics and conditions should be virtually homogeneous along a section. Track type (1-9) and density of rail joints (1-4) are specified for each section. Track type (1-9) 1. track with concrete single block or twin block sleepers in ballast bed; 2. track with wooden or zigzag concrete sleepers in ballast bed; Page 2 of 14

3 3. track with ballast bed and rails with joints; rails with not more than two crossings with joints within 50 m; 4. track with blocks; 5. track with blocks and ballast bed; 6. track with controllable rail fixation; 7. track with controllable rail fixation and ballast bed; 8. track with poured-in rail; 9. track with level crossing. Density of rail joints (1-4) 1. jointless rail (fully welded) with or without jointless switches and crossings; 2. rails with joints; 3. switches and crossings with joints, 2 per 100 m; 4. more than 2 crossings per 100 m (the number of crossings can be stated). Vehicle specifications Category: 1. passenger train with tread brakes; 2. passenger train with both disc brakes and tread brakes; 3. passenger train with disc brakes; 4. goods train with tread brakes; 5. diesel train with disc brakes; 6. underground or express tramway vehicle with disc brakes; 7. intercity train with disc brakes; 8. high speed trains with disc brakes and/or tread brakes. For each category: 9. vehicle intensity (number of passing trains per hour) [1/h]; 10. driving speed (for trains that are passing at constant speed) [km/h]; 11. percentage of braking vehicles [%]; Sound power levels of non-standard vehicles (not fitting within categories 1-7): in decibels re 1 pw, at track height and at 0.5 m above railhead, for the octave bands with centre frequencies Hz. Page 3 of 14

4 1.2 Assessment data Buildings Specified are: o sizes (by the positions of the corners, in the co-ordinate system that has been chosen); o height (ride height in case of a peaked roof) [m]; o reflection factor of façades [%]. Sound barriers Specified are: o height [m]; o shape types: sharp or obtuse top (angle between 0 and 70 or between 70 and 165 ); o reflection coefficients of barrier surfaces (standard or customised). Soil Specified are: o fraction of sound absorbing soil surface between track and observer [-]; or o hard and sound absorbing areas are specified separately by stating the character of each area (reflecting or absorbing) and the positions of the border lines. o height of ground surfaces [m]. Assessment point For each point is specified: o position [m]; o whether or not it is on a façade or in an open field. Maximum number of reflections per sound ray The maximum number of reflections per sound ray is specified. A possible sound transmission path sound ray is only taken into account if it can be constructed with not more than the specified maximum number of reflections. 2 Guidance on the definition of emission data As the emission data are the basics for any noise calculation, those data could be contained in a central national or regional register (in the Netherlands, called "Akoestische spoorboekje"). Considering the fact that these data need to be directly used for acoustical surveys, they need to comply with the minimum requirements for accuracy. For each type of data mentioned above, the minimum requirements are described below (reprint of 3.1 of Interim Computation Method). Map The map must state a unique link between the gathering of data and the track route. A certain scale level is not imposed as it depends on the complexity. In most cases, a scale of 1/ is sufficient, Page 4 of 14

5 but in some urban areas 1/ is necessary. A stepless adjustable electronic version must -for each route- provide the link with the data. Tracks The start and end of each track must accurately be stated in metres. For a multi track route, the type of track must be stated. For the position of stations, a global indication with an accuracy of 100 m and the name is sufficient. Vehicle intensity Use of the track must be stated per track, in units per hour, rounded up to 0.1 unit. The statement is done per vehicle category according to 2.1, over day, evening and night period. Speed profile Speeds on the route, averaged over a year, are stated per vehicle category, including indication where the vehicles at normal conditions in the service use their brakes. If several speed profiles need to be used, an indication of which part of the vehicles use which profile is necessary (see also intensities). Speeds are to be rounded up to the nearest 5 km/h. Track The position beginning and end of the constructions described in 1 are indicated with an accuracy of 1 m. In very complex situations (several switches over distances less than 100 m) an indication of the number of joints over the complex situation is sufficient, depending on the total number of switches. Barriers (not mandatory) If the position of barriers is included in the register, the following data should be stated: o beginning and end [m]; o track along which it is placed; o indication whether it is placed on the left or the right side of the track; o height [dm]. Height (not mandatory) The height must be given per at least 100 m of track in dm above NAP. 3 Guidance on application of the propagation model 3.1 Modelling the situation See 1.3 of Interim Computation Method including: 1. Source lines 2. Composition of Ground 3. Ground height differences 4. Standard Embankment 5. Level crossing 6. Screening slabs Page 5 of 14

6 7. Barriers 8. Platforms 9. Bridge constructions 10. Noise absorbent constructions 11. Reflections 3.2 Computation The equivalent sound level L Aeq in db(a) is calculated as follows: where L eq.i.j.n L eq.i.j.n includes following values: with: L Aeq 8 = 10lg i= 1 J N 10 j= 1 n= 1 Leq,i, j,n /10 specifies the contribution in an octave band (index code i) of a sector (index code j) and a source point (index code n). Leq,i, j,n = LE + LGU LOD LSW LR 58.6 L E emission value per source height and octave band L GU attenuation due to distance L OD attenuation due to propagation L SW screening effect, if present L R attenuation due to reflections, if present Emission value Emission value is the source power of one source point. Source point is part of sector: opering angle of 5 ; number of source points of one sector depends on how often the source line intersects the sector area. Each sector is one homogeneous part (~100m) of one source line: same vehicle, same speed, same track, same conditions Attenuation by distance L GU Data In order to calculate the geometric propagation factor the following data is necessary: r ν φ distance between source and assessment point, measured along the shortest connection line [m]; angle between sector area and section of the source line [in degrees]; opening angle of the sector [in degrees]. Page 6 of 14

7 Calculation The calculation of L GU is as follows: L GU φsin ν = 10lg r Remark If the angle ν takes on a value smaller than the opening angle of the sector concerned, further examinations must be carried out to determine L GU Attenuation by propagation L OD Losses on the transmission path L OD are composed of the following factors: where: L OD = DL + DB + CM D L D B C M air attenuation ground attenuation meteorological correction factor. Air attenuation D L The given values for δ air are derived from the third band spectrum ISO DIS 3891 at 10 C and relative humidity of 80%. Specifically in the case of the high frequency bands, certain compensations for the intense dispersion character of the absorption have been added. Data In order to calculate D L the following data is necessary: r the distance between source and assessment point, measured at the shortest connection line [m] Calculation Calculation is as follows: where: DL = rδair δ air air absorption coefficient Ground attenuation D B When determining the ground attenuation D B, the horizontally measured distance between source and assessment point (Symbol r o ) is divided into three areas: source area, assessment area and middle area. The source area has a length of 15 m and the assessment area a length of 70 m. The remaining section of the distance r o between the source and assessment point forms the middle area. The term acoustically hard here refers to: pavement, asphalt and other sealed surfaces, water surfaces etc. The term acoustically non-hard refers to: grass surfaces, agricultural ground with or without vegetation, sandy surfaces, ground without vegetation etc. Detailed information on modelling and equations can be found in of the proposed interim method. Page 7 of 14

8 Meteorological correction factor C M Calculation of ground attenuation is based on downwind noise propagation. The C M correction factor corrects to long term average conditions based on 45 % downwind conditions. C M = 0 calculates downwind noise propagation. Data In order to calculate the meteorological correction factor C M, the following information is necessary: r o h b h w Calculation horizontally measured distance between source and assessment point [m] height of the source point above the average terrain level inside the source area [m] height of the assessment point above the average terrain level inside the assessment area [m]. The calculation is as follows: h for r o > 10(h b + h w ) C M = 3,5-35 for r o 10(h b + h w ) C M = 0 + b r 0 h w Attenuation factor for screening L SW If objects found inside a sector have at least a viewing angle that corresponds with the opening angle of the sector concerned and if we can presume that these objects interfere with sound transmission, the attenuation factor L SW is taken into account, along with reduced ground attenuation. The formula for calculating the attenuation contributed by an object of variable shape contains two factors. The first factor describes the screening by an equivalent idealised barrier (a thin, vertical plane). The height of the equivalent barrier corresponds to the height of the obstructing object. The upper edge of the barrier corresponds to the highest edge of the obstacle. If it is possible to place the barrier in various positions, the position at which the highest attenuation occurs is chosen. The second factor is of importance only if the profile, deviates from that of the idealised barrier. The profile is defined as the cross-section of the sector plane of the attenuating object. The attenuation of the object is equal to the attenuation of the equivalent barrier minus a correction factor C p depending on the profile. If several attenuating objects are present in a sector, only the object that - in the absence of the others - would cause the most attenuation is taken into account. Detailed information on modelling and equations can be found in 1.6 of the proposed interim method Determining rail specific absorption The absorption coefficients α will be averaged using a weighting factor. As weighting factor the averaged A-weighted 1/3 octave spectrum of the traffic spectrum is used Reduction of levels as a result of reflections L R Data In order to calculate level reductions as a result of absorption caused by reflections, the following data is necessary: Page 8 of 14

9 N ref number of reflections (see also 1.3) between source point and assessment point [-] - type of reflecting object Calculation The calculation is as follows: where: δ ref LR = N ref δref level reduction by means of reflection. Results For buildings δ ref = 0.8 is valid for all octave bands. For all other objects δ ref = 1 is valid for all octave bands, unless the object is proven to be sound absorbing. In this case, δ ref = 1 - α per octave band is valid, where α is the sound absorption factor of the object in the octave band concerned. The highest value for N ref is 3: reflections of 4 th and higher orders are not taken into account. Page 9 of 14

10 Similarities & differences of ORM with END and Use of ORM for strategic noise mapping prepared by AKRON n.v.-s.a. 1 Aim The aim of the proposed computation method is strategic noise mapping for railway noise. This means the realisation of global overview of noise levels in a certain area/zoning. As imposed by the EU directive, the method should be based on the Dutch RMR method of On the other hand, following aspects have to be taken into account: o publication of Dutch government of RMR 2002; o proposal of Dutch government of ARM 1.5 model for strategic noise mapping; o integration of information of EU directive: END terminology; o co-ordination between different EU Interim Methods. The choice for the ORM method out of RMR 2002 with adaptation to END terminology is justified hereafter. 2 Compatibility with END 2.1 Emission values The calculated noise levels at the assessment points fulfil the criteria for strategic noise mapping if the emission register contains corresponding data: for L day yearly average of vehicle density during day period for each vehicle category; for L evening yearly average of vehicle density during evening period for each vehicle category; for L night yearly average of vehicle density during night period for each vehicle category. For the Netherlands, this has been realised in a publication named "Akoestisch Spoorboekje", where an estimate of the 2010 average vehicle load is described for each track section. The global evaluation parameter L den is then calculated according to the procedure in EC-document 6660, annex I. with: L day L evening L night L den = 10lg 1 24 Lday /10 Levening + 5 /10 Lnight + 10 / *10 + 4*10 + 8*10 A-weighted long-term average sound level as defined in ISO : 1987, determined over all the day periods of a year A-weighted long-term average sound level as defined in ISO : 1987, determined over all the evening periods of a year A-weighted long-term average sound level as defined in ISO : 1987, determined over all the night periods of a year Page 10 of 14

11 2.2 Meteo If meteorological conditions similar to the other interim calculation methods have to be applied, following hypotheses should be realised: o down wind: day period: 50%; evening period: 75%; night period: 100%. This should be no problem in this computation method. D soil, the basic calculation of ground effect is based on downwind curve effects (theory of Maekawa); a meteo correction term C meteo is applied for global overall calculation, based on an average 45% downwind propagation. Thus, C meteo = 0, leads to downwind calculation. Combination according to the percentage of downwind and global conditions leads to the required values. 2.3 Assessment points According to END, the assessment points are to be situated: o height: 4 m; o distance to façade: 2 m; o calculation without reflection on the considered façade. This is possible in all Dutch calculation methods (ARM 1, ARM 1.5, ORM): o reflection is not taken into account; o reception point (= assessment point) can be selected freely. 3 Compatibility between rmr 1996 & RMR 2002 The RMR 2002 is the third edition of the Dutch document Standaard Rekenmethode. The first version dated from 1987; a second version was made in 1996; the most recent version dates from 28 March 2002 and will become official in the Netherlands, somewhere during the summer of Whereas the version of 1996 introduced considerable modifications of the emission values, the 2002 version introduces modifications on determination by measurement of emission values and on propagation methods. The most significant modification concerns the addition of measurement procedures for emission values. These methods have been developed by TNO (Technological Research Institute) on demand of the Dutch authorities. Two procedures are added: a simplified procedure to add new vehicles to e- xisting categories (procedure A); a more elaborate procedure to define totally new emission values for new categories of vehicles (procedure B). In addition, a procedure for determining categories for track is provided (procedure C). In the calculation for the emission values in octave bands, the actual roughness of the track and the reduction of sections with permanent smooth rails can be taken into account. In the propagation algorithms, recent knowledge about the type of reflections (based on minimum viewing angle), the number of reflections and the screening effect of different types of barriers has been introduced. Page 11 of 14

12 In chapter 6 of RMR 2002, a new calculation method for strategic noise mapping is added: ARM 1.5. It concerns a method of intermediate accuracy, in between: o the global db(a) calculation method: named: ARM 1 (previously SRM I); o the octave band calculation method: ORM (previously SRM II). Although not included in the text, reference should be made to the Akoestisch Spoorboekje. This is the Dutch emission register based on the 1996 emission values of the vehicles. The most recent version (2000) is available as software ASWIN2000. It contains yearly averages for each section based on a prognosis for Therefor this register is directly useful for the strategic noise mapping as requested by the EC. 4 Comparison of ARM 1.5, ORM & ISO methods 4.1 Emissions Emission is based on an emission model as function of vehicle characteristics, speeds, density of traffic, track conditions, etc In the actual emission model, nine different categories of rail vehicle are defined based on common European rail vehicles. Procedures are defined for other types of vehicles to assimilate them to e- xisting categories or to define new categories. For the ARM 1.5 method, global db(a) emission values can be obtained, situated at 0.25 m above the track for standard trains. For high-speed trains, four emission values at different heights are calculated out of an octave band emission table. The values are then combined in four global db(a) emission values at four different heights. For the ORM method: octave band emission values are used : o at heights of 0 and 0.5 m for standard trains; o at heights of 0, 0.5, 2, 4 & 5 m for HST trains. 4.2 Geometrical noise propagation The combination of L w,n and two first terms of equation 1.6 of ARM 1.5 (T n ) 10 log (4πr²) + 20 log r/d is equal to the emission value and the geometrical propagation term in the ORM, which is based on line sources and linear geometrical propagation. In this, the di-pole directivity is embedded in the term 20 log r/d. 4.3 Meteorological terms D air air attenuation For the ARM 1.5 method, the air attenuation has been determined as a weighted summation of the 250, 500, 1000 and 2000 Hz terms out of the octave band calculation method, with a respective weighting of 10, 35, 35 & 20%. The difference with the octave band calculation method is maximum 0.3 db between 0 and 1000 m. The octave band values are globalised from the 1/3 octave band values of the ISO-DIN 3891 standard at 10 C and 80% humidity. Page 12 of 14

13 4.3.2 D soil soil absorption For the ARM 1.5 method, the equation for soil absorption has been recalled from the ORM method, using a weighted summation of the 500, 1000 and 2000 Hz band with a respective weighting of 20, 40 & 40%. For an average railway emission spectrum, difference with ORM is less than 0.2 db(a) for distances between 0 and 1000 m. For modelling simplification, it is proposed to use an equal sound absorption coefficient of 0.8 for source, middle and receiver areas. This is based on the fact that the source area for railways is almost always absorbent. In special situations, differences up to 2 db can be obtained. Then, the use of the specific absorption coefficients for each of the three areas is required. Modelling is based on the downwind curved rays model. Comparison with ISO 9613 indicates a simplification in the ORM method for the source and receiver area; the equations are identical for the intermediate area D meteo meteorological correction The meteorological correction term corrects for the ground effects calculation, based on downwind propagation. The meteorological correction term in the ARM 1.5 method is identical to the correction term used in the octave band calculation method. The ORM term is independent of frequency. The calculation is similar to ISO 9613 with a C 0 value of 3.5 db: this means favourable wind conditions approximately 45% of the time present. 4.4 Reflections h + = b h C w M C r0 In most situations, reflections do not play an important role in railway noise because reflection surfaces are situated at larger distance of the source line. In specific situations such as entrance/exit of tunnels, crossing bridges or tracks without ballast, this can lead to large differences, but mostly these inaccuracies cause only local effects (in direct area around barrier/reflector) and are not important for noise mapping. In the ARM 1.5 method, reflections are not taken into account. In urban areas, the effect of building has been taken into account by a global diffuse term such as D urban. In the ORM method, reflections up to the 3 rd reflection are taken into account, based on the definition of a mirror receiver with a reflection correction. 4.5 Screening The calculation of the additional transmission length is identical in the ARM 1.5 and the ORM method. Screening is based on a downwind model of Maekawa at 800 Hz (as proposed in Urbis equation for passenger trains). Other corrections (as proposed by Urbis equation) for freight trains are not applied. The curved ray model calculates a correction of the effective barrier height. Page 13 of 14

14 4.6 Urban damping (ARM 1.5 method) This value is based on studies carried out by the Dutch government, on the effects of urban areas. The basic equation calculates noise at 3 m below the ridge. A correction term as function of height is added for other receiver heights. 4.7 Conclusions The ARM 1.5 method can be considered as a simplification of the ORM method. The propagation model of the ORM and the ISO method are very similar. As ARM 1.5 simplifications dilute the link between RMR and ISO and are questionable at several situations, the ARM 1.5 method is not retained. Page 14 of 14