Freight Train Noise Assessments

Transcription

1 PO Box 717 Hillarys WA 6923 T: F: E: W: Freight Train Noise Assessments Reference: Final.docx Prepared for: Freight & Logistics Council Member Firm of Association of Australian Acoustical Consultants

2 Report: Final.docx Lloyd George Acoustics Pty Ltd ABN: PO Box 717 Hillarys WA 6923 T: / F: Contacts Daniel Lloyd Terry George Mike Cake Matt Moyle E: daniel@lgacoustics.com.au terry@lgacoustics.com.au mike@lgacoustics.com.au matt@lgacoustics.com.au M: This report has been prepared in accordance with the scope of services described in the contract or agreement between Lloyd George Acoustics Pty Ltd and the Client. The report relies upon data, surveys, measurements and results taken at or under the particular times and conditions specified herein. Any findings, conclusions or recommendations only apply to the aforementioned circumstances and no greater reliance should be assumed or drawn by the Client. Furthermore, the report has been prepared solely for use by the Client, and Lloyd George Acoustics Pty Ltd accepts no responsibility for its use by other parties. Prepared By: Mike Cake Position: Project Director Verified Terry George Date: 14 September 2015

3 Table of Contents 1 INTRODUCTION 3 2 CRITERIA 4 3 NOISE MEASUREMENTS 5 4 ASSESSMENT OF NOISE MEASUREMENTS 10 5 MAXIMUM VERSUS AVERAGE NOISE LEVEL 12 6 SPP GUIDELINES APPENDIX A 14 7 APPLYING ARCHITECTURAL PACKAGES 16 8 Sensitivity Analysis Analysis of Measured Data Maximum Noise Criteria Glazing to Floor Ratio Upper Limit of Packages 21 9 CONCLUSION 22 List of Tables Table 2-1 Outdoor Noise Criteria 4 Table 3-1 Noise Logging Results: 25m From Railway 5 Table 4-1 Noise Logging Results Adjusted to 1 Movement Per Hour: 25m From Railway 10 Table 6-1 Estimated L AeqNight) Noise Levels Provided in SPP Guidelines 14 Table 7-1 External Noise Spectra for Freight Trains 16 Table 7-2 Typical Apartment Room Dimensions 16 Table 7-3 Calculated Internal Noise Levels Typical Apartment 17 Table 8-1 Calculated Internal Noise Levels Larger Glazing 21 Table 8-2 Calculated Internal Noise Levels Typical Apartment - Upper Limit Package B 21 Table 9-1 Recommended Acceptable Treatment Packages for Freight Rail Noise 24 Table 9-2 Example Construction for Freight Rail Noise 25

4 List of Figures Figure 3-1 Location of Noise Measurements 5 Figure 3-2 Spectral Content of Train Pass-by L Aeq 7 Figure 3-3 Spectral Content of Train Pass-by L Amax 8 Figure 5-1 Comparison of L Amax Versus L Aeq - Up to 2 Train Movements Per Hour 12 Figure 5-2 Comparison of L Amax Versus L Aeq - Up to 4 Train Movements Per Hour 13 Figure 6-1 Comparison of L Amax Versus L Aeq - Measured vs Guidelines 14 Figure 9-1 Recommended Land Use Planning Guidance for Freight Rail Noise 23 Appendices A Acceptable Treatment Packages B Terminology

5 1 INTRODUCTION In 2009 the State Planning Policy 5.4 Road and Rail Transport Noise and Freight Considerations in Land Use Planning (SPP 5.4) was gazetted and provided a whole-of-government assessment approach for noise impacts associated with road and rail transportation. The objectives of SPP 5.4 include protecting people from unreasonable levels of transport noise and protecting major transport corridors and freight operations from incompatible urban encroachment. Early drafts of SPP 5.4 contemplated the inclusion of a maximum noise level criteria (L Amax ), which would apply in addition to the day and night average noise level criteria (L Aeq(Day) & L Aeq(Night) ). However, the L Amax criteria was removed during stakeholder consultation. For road and passenger rail traffic there are generally sufficient numbers of vehicle movements for the L Aeq value to reasonably represent the noise impact. However, freight rail can be subject to a relatively low number of movements, which results in low L Aeq noise levels when averaged over a long time period. Freight rail noise is typically characterised by infrequent train pass-bys, which generally result in sharp increases in noise, at levels that are significantly higher than the background noise environment. it is this sudden increase in noise that can cause disturbance inside dwellings (and other noise-sensitive premises). Sleep disturbance is a primary area of concern, as this can lead to a range of associated negative health effects. L Amax noise levels are often used to assess these impacts, because they are known to have a strong correlation with sleep disturbance and to awakenings during sleep. The SPP 5.4 Guidelines require noise impacts to consider a minimum of one train movement per hour, irrespective of whether these numbers of movements are forecast or not. This is partly to ensure that reasonable, long term utilisation of a freight rail line is not compromised when a noise assessment is undertaken. However, it is Lloyd George Acoustics understanding that this assessment approach was also adopted as a way of taking into account the fact that a low use track might still have a significant acoustic impact on nearby noise sensitive land uses. The intent was that SPP 5.4 would, by default, address L Amax impacts, by artificially increasing the L Aeq value in these cases. A key question for the Freight and Logistics Council is whether the current approach adopted by SPP 5.4 (and the associated Guidelines) adequately addresses L Amax impacts from freight rail. If not, the question is then asked, under what circumstances is the existing policy approach inadequate at addressing L Amax freight rail impacts. This report attempts to answer these key questions. Such questions are critical if freight rail corridors are to be protected from inappropriate urban development, since increasing residential complaints and pressure on freight rail operations is likely to be the result if these noise impacts are not properly ameliorated. Reference: Final.docx Page 3

6 2 CRITERIA The criteria relevant to freight trains in Western Australia is the State Planning Policy 5.4 Road and Rail Transport Noise and Freight Considerations in Land Use Planning (and the associated Guidelines) produced by the Western Australian Planning Commission. The objectives of SPP 5 4 are to: Protect people from unreasonable levels of transport noise by establishing a standardised set of criteria to be used in the assessment of proposals; Protect major transport corridors and freight operations from incompatible urban encroachment; Encourage best practice design and construction standards for new development proposals and new or redevelopment transport infrastructure proposals; Facilitate the development and operation of an efficient freight network; and Facilitate the strategic co-location of freight handling facilities. The Policy s outdoor noise criteria are shown in Table 2-1. These criteria apply at any point 1-metre from a habitable façade of a noise sensitive premises and in one outdoor living area. Compliance with these criteria are to give regard to a year transport horizon. Table 2-1 Outdoor Noise Criteria Period Target Limit Day (6am to 10pm) 55 db L Aeq(Day) 60 db L Aeq(Day) Night (10pm to 6am) 50 db L Aeq(Night) 55 db L Aeq(Night) Note: The 5 db difference between the target and limit is referred to as the margin. In the case of new noise sensitive developments constructed within the vicinity of a freight rail corridor, the objectives of SPP 5.4 is to achieve - acceptable indoor noise levels within residential buildings of 35 db L Aeq(Night) in bedrooms and 40 db L Aeq(Day) in other habitable rooms; and a reasonable degree of acoustic amenity in at least one outdoor living area on each residential lot. In the 2005 draft version of SPP 5.4, the L Amax outdoor criteria were a 75 db target and 80 db limit, which applied equally day and night. Reference: Final.docx Page 4

7 3 NOISE MEASUREMENTS It was considered that the most appropriate way to answer the Freight & Logistics Council s questions was to examine some real measured data of freight trains on a low usage track. Noise level measurements were undertaken alongside a freight railway in the vicinity of Robbs Road Jetty (refer Figure 3-1), where trains travel to and from the North Quay Rail Terminal at Fremantle Port 1. It was proposed to construct multi-residential apartments on this site and similar developments are occurring in this vicinity within the Cockburn area. As part of earlier work by another consultant, a buffer distance of 15 metres from the edge of the railway reserve to noise sensitive development was set (mainly to minimise vibration impacts), representing a distance of approximately 25 metres from the track. So noise measurements were taken at this distance of 25 metres from the track. Noise Level Measurements Figure 3-1 Location of Noise Measurements Table 3-1 provides the results of the noise level measurements recorded by the automatic noise data logger in free-field conditions. Table 3-1 Noise Logging Results: 25m From Railway Date No. of Pass-bys (Night / Day) Time L AFmax, db L Aeq(Day) / L Aeq(Night), db 14 November : November : N/A 14 November : November : November : Currently train movements along this rail line service either the Kwinana Port or the Forrestfield Intermodal Terminal. In the past, trains have also accessed this line for train movements to/from Kalgoorlie. Reference: Final.docx Page 5

8 Date No. of Pass-bys (Night / Day) Time L AFmax, db L Aeq(Day) / L Aeq(Night), db 15 November : November : November : November : November : November : November : November : November : November : November : November : November : November : November : November : November : November : November : November : November : November : November : November : November : November : November : November : Reference: Final.docx Page 6

9 Date No. of Pass-bys (Night / Day) Time L AFmax, db L Aeq(Day) / L Aeq(Night), db 20 November : November : November : November : November : November : November : N/A Note: N/A means incomplete L Aeq(Day) time period at the start and the end of the measurement. It should be noted that there was no train movements on Sunday 17 November 2013 and limited movements on Saturday 16 November Of those train movements measured, 16 occurred during the night time period (10pm to 6am), representing 40% of all movements, with 24 movements, or 60% of all total movements, during the day time period (6am to 10pm). The noise logger also provided the spectral content of the noise and this is shown by Figures 3-2 and 3-3, representing the L Aeq for a train pass-by and the L Amax level respectively. These spectral results reflect the low frequency characteristics of freight rail noise and where vital for later calculations to predict indoor noise levels (see Section 7). Also shown on the charts is the range of measured noise levels, which can be quite significant, particularly in the case of the L Amax levels. Figure 3-2 Spectral Content of Train Pass-by LAeq Reference: Final.docx Page 7

10 Figure 3-3 Spectral Content of Train Pass-by LAmax From the measurements, the L Aeq(Day) and L Aeq(Night) for trains were calculated to be 46.7 db and 48.4 db respectively. For comparison to the criteria, which applies at 1-metre from a habitable façade, the noise levels would be increased by 2.5 db due to reflected noise from the façade. That is, the L Aeq(Day) and L Aeq(Night) levels would be assessed as 49.2 db and 50.9 db respectively. The L Aeq(Day) level complies with the SPP 5.4 target outdoor noise criteria of 55 db. However, the L Aeq(Night) level marginally exceeds the SPP 5.4 target outdoor noise criteria of 50 db. The maximum noise levels from passing trains ranged from 67.7 db to 93.2 db L Amax. The mean L Amax noise level was calculated to be 80.3 db, with the data set having a standard deviation of 7.0 db. Of all the measured train movements, 75% (30 out of 40) exceeded the L Amax outdoor noise target of 75 db from the 2005 draft SPP 5.4. Almost half of all train movements (19 out of 40) exceeded the 80 db L Amax outdoor noise limit. Taking the mean, or average, to represent the wide range of L Amax values somewhat distorts the full picture in respect to noise disturbance. Doing so would mean that close to 50% of all the trains measured had a higher L Amax noise level. Ignoring all these noisier trains would likely result in an under representation of the true noise impact from this rail line. So, for the purposes of the subsequent analysis in this report, the mean plus one standard deviation ( +1 ) has be used 2. This gives a representative L Amax value of 87.3 db. (The impact of this decision to use the mean plus one standard deviation value is further discussed in Section 8.1.) Including the façade correction of 2.5 db, this results in an L Amax noise level of 89.8 db. 2 Data that is normally distributed will be perfectly symmetrical around its mean. For normally distributed data, taking the mean plus one standard deviation means that 84% of all trains would be expected to be less than (or equal to) this L Amax value, with only 16% of all trains recording L Amax noise levels higher than this. In this circumstance, because the measured values are not perfectly symmetrical about the mean, the mean plus one standard deviation captures 78% of all measured train movements. Reference: Final.docx Page 8

11 Note that the calculated L Aeq(Day) and L Aeq(Night) values are around 40 db less than the L Amax value. This difference in L Amax versus L Aeq(Day) /L Aeq(Night) noise levels is much greater than would be expected for major roads or passenger rail. This highlights the problem of using L Aeq values to assess freight rail noise. The infrequency of freight train movements means that long term L Aeq averaging tends to under play the real impact of these freight train pass-bys, thereby under representing the annoyance/disturbance caused. Reference: Final.docx Page 9

12 4 ASSESSMENT OF NOISE MEASUREMENTS The measured noise levels are determined to be 49.2 db L Aeq(Day), 50.9 db L Aeq(Night) and 89.8 db L Amax (including facade correction). It is understood that at the time of the measurements (in 2013) there were nominally 8 train movements per day on this section of the rail line, although operationally this varied somewhat. The SPP 5.4 requires that consideration be given to a year planning horizon. In many cases, obtaining forecast movements for freight trains is difficult, however in this case, the following information was provided: Current freight train movements (2015) are 6 per day, Mondays to Saturdays with a typical train length of 600 metres (50 TEU s per train). Total trade through Fremantle Port is currently 700,000 TEU s (Twenty foot Equivalent Unit), with the freight train modal split currently representing 13% of this total (i.e. 91,000 TEU s per annum via rail). There is a government initiative to increase freight train movements to represent 30% of Port trade (210,000 TEU s). This equates to 7 movements per day, 7 days a week with a train length in the order of 850 metres (80 TEU s per train). Future trade through the Port is expected to increase to 1,200,000 TEU s per annum. Assuming 30% of these are by freight rail (360,000 TEU s), this would require 12 movements per day of the 850 metre long trains. Because the trains become longer in the future scenario, and operate on a 7 day week schedule, the number of train movements is lower than would be the case if the trains remained at the same length. Whilst it would be operationally inefficient if the 360,000 TEU s remained on trains carrying 50 TEU s, for comparative purposes this would equate to 20 train movements per day. As discussed earlier, the SPP Guidelines requires it to be assumed that there is at least 1 train movement per hour (24 movements per day). As the SPP assumption is greater than the forecast number of movements, the 1 movement per hour assumption is used in the analysis. Table 4-1 provides the daily noise levels by increasing the actual number of movements noted in Table 3-1 to 16 during the day and 8 during the night (i.e. 1 per hour). Table 4-1 Noise Logging Results Adjusted to 1 Movement Per Hour: 25m From Railway Date L Aeq(Day) L Aeq(Night) 15 November November November November November November Average Note: L Aeq(Night) is calculated from 10pm on the previous date to 6am on the date shown. Reference: Final.docx Page 10

13 Including the façade correction, and allowing for 1 train movement per hour, the noise levels at 25 metres would be 55.3 db L Aeq(Day) and 55.8 db L Aeq(Night), with the L Amax unchanged at 89.8 db. Assessing the noise levels against the SPP 5.4 criteria shows that during the day, noise levels are compliant, being no more than the target (55 db L Aeq(Day) ) when rounded, with night-time noise levels being slightly above the limit (55 db L Aeq(Night) ) when rounded to 56 db L Aeq(Night). Where noise is 6 db above the target, as night-time noise levels are in this instance, Acceptable Treatment Package B (refer Appendix A) would be applicable. Application of Package B is considered to provide an attenuation from outside to inside of up to 23 db since the external noise level can be up to 58 db L Aeq(Night) and the internal design noise level is 35 db L Aeq(Night). That is, in this instance, the internal noise levels would be 32 db L Aeq(Day) and 33 db L Aeq(Night). Comparing the rounded L Amax noise level of 90 db to the draft 2005 SPP 5.4, indicates an exceedance of 15 db (where the target was set at 75 db). Assuming the same relationship between external and internal acceptable noise levels (15 db difference between outside and inside) applies equally to L Amax as it does to L Aeq, means that the internal design goal L Amax would be 60 db. Applying the 23 db reduction of Package B to the assessed external maximum noise level results in an internal noise level of 67 db L Amax. As such, under the above scenario, following the Policy requirements and requiring Package B, would result in maximum noise levels being 7 db above the design goal. Reference: Final.docx Page 11

14 5 MAXIMUM VERSUS AVERAGE NOISE LEVEL From Section 4, it can be seen that for the example where noise is assessed at 25 metres from the track, and on the assumption that the criteria is 50 db L Aeq(Night) and 75 db L Amax, (as per the draft 2005 SPP 5.4), the L Amax criteria is more critical. To explore this issue further, Figure 5-1 shows how the L Amax noise level varies with distance from the track compared to the L Aeq(Night) noise level, depending on number of movements per hour, in terms of their exceedance above the relevant criteria. Some items to note that influence the analysis are: 0 m equates to the closest line of the track within the rail reserve. The L Aeq is based on the Table 4-1 analysis and is therefore specific to this section of track. Where trains are longer, use different locomotive configurations, in different notch settings or the like, the L Aeq noise level will be different. Train movements of 1 and 2 trains per hour are shown to align with the SPP 5.4 Guidelines. It is assumed the L Aeq will decay at a rate of 4 db per doubling of distance. It is assumed the L Amax will decay at a rate of 6 db per doubling of distance. If the criteria varies (e.g. external maximum noise level is 80 db L Amax rather than 75 db L Amax ), the relationship will also vary. Figure 5-1 Comparison of LAmax Versus LAeq - Up to 2 Train Movements Per Hour The chart of Figure 5-1 shows the following: The L Amax external noise target is not satisfied until a distance of 135 metres from the track. This is shown as the time when the red L max line reaches 0 db exceedance. Reference: Final.docx Page 12

15 The L Aeq(Night) criteria for 1 train movement per hour is not satisfied until a distance of 70 metres from the track, being when the dark blue 1/hr-L eq line reaches 0 db exceedance. The L Aeq(Night) criteria for 2 train movements per hour is not satisfied until a distance of 115 metres from the track, being when the blue 2/hr-L eq line reaches 0 db exceedance. The L Amax criteria is always more stringent than the L Aeq(Night) criteria for both 1 and 2 train movements per hour, since the L Amax line is always higher than the 1/hr-L eq and 2/hr-L eq lines. Figure 5-2 expands on the above analysis by also showing the L eq compliance curves for 3 and 4 movements per hour. 4 train movements per hour (i.e one train every 15 minutes) is assumed to represent a realistic upper limit for operational activity on a freight rail line. Figure 5-2 Comparison of LAmax Versus LAeq - Up to 4 Train Movements Per Hour From Figure 5-2, the following additional comments can be made: Considering a maximum of 4 train movements per hour, the L Amax criteria is more critical than the L Aeq(Night) criteria within a distance of 70 metres from the track. Beyond that distance the L Aeq(Night) criteria becomes more critical. For 3 train movements per hour, the L Amax criteria is more critical than the L Aeq(Night) criteria within a distance of 105 metres from the track. Figures 5-1 and 5-2 indicate that an L Amax assessment will be critical in many circumstances when compared to a standard SPP 5.4 L Aeq noise assessment. Reference: Final.docx Page 13

16 6 SPP GUIDELINES APPENDIX A The outcome of the Section 4 and 5 analysis highlights that, for most scenarios that are likely to exist along the Forrestfield to Fremantle Port track, the L Amax criteria may not be adequately addressed. This is the case even under the assumption that there may, in the future, be up to 1 train movement per hour, which is considered conservative based on forecast movements. Appendix A of the SPP 5.4 Guidelines includes the following table to provide guidance on estimated L Aeq(Night) noise levels from a freight rail line, as part of a screening assessment tool. Table 6-1 Estimated LAeqNight) Noise Levels Provided in SPP Guidelines Movements per day Distance from Track (m) per hour (minimum) per hour * Noted as a typical mix of, S or Q, P and D, AB, DA etc or L Class locomotives, and number of wagons varying from 45 to 90. Notch settings varying from 5-8. The Guidelines give an estimated L Aeq(Night) value at 25 metres from the track of around 60 db for 1 train movement per hour, compared to the 56 db used in this report. Figure 6-1 compares the L Aeq(Night) values from the Guidelines with the measured values calculated in this report, for 1 train movement per hour. Figure 6-1 Comparison of LAmax Versus LAeq - Measured vs Guidelines Reference: Final.docx Page 14

17 The values provided in the Guidelines are certainly higher (more conservative) than those that were measured along the Forrestfield - Fremantle Port track. Therefore the relative difference between the L Amax and L Aeq values is not as great when the Guidelines values are applied. The key points from Figure 6-1 are summarised below: The L Aeq(Night) criteria was satisfied at 70 metres from the track based on the measurements, whereas this is now increased to around 150 metres for the Guidelines values. The L Amax criteria is always more critical than the L Aeq(Night) measured values, whereas at a distance of around 105 metres the L Amax becomes less critical than the Guidelines L Aeq(Night). At a distance of 25 metres, and assuming 1 train movement per hour, the L Aeq(Night) exceedance is noted as 6 db for the measured values. Therefore Acceptable Treatment Package B would be applicable to development at this location. Using the Guidelines L Aeq(Night) values, the exceedance would be 10 db, therefore Package C would be applicable. Whilst the latter is more stringent, it is still insufficient to accommodate the 15 db L Amax exceedance. The above demonstrates that an L Amax assessment will still be critical, in a range of situations, even if the more conservative L Aeq(Night) values mentioned in Appendix A of the SPP 5.4 Guidelines are applied. Reference: Final.docx Page 15

18 7 APPLYING ARCHITECTURAL PACKAGES The Acceptable Treatment Packages within the SPP 5.4 Guidelines (refer Appendix A of this report) are applied as follows: Package A Where noise levels are within the margin (5 db above the target), that is up to 60 db L Aeq(Day) and 55 db L Aeq(Night). Package B Where noise levels are up to 8 db above the target (3 db above the limit), that is up to 63 db L Aeq(Day) and 58 db L Aeq(Night). Package C Where noise levels are up to 10 db above the target (5 db above the limit), that is up to 65 db L Aeq(Day) and 60 db L Aeq(Night). The Packages, however, are a deemed-to-comply standard and are therefore generally considered conservative. For many projects, where a house plan is provided to a consultant, the level of architectural treatment can generally be reduced. As such, the internal noise level provided by the deemed-to-comply Packages may in fact be less than the design goals of 35 db L Aeq(Night) in bedrooms and 40 db L Aeq(Day) in other habitable rooms. From the measurements, Table 7-1 provides the external noise spectra used in the calculations, representing the external noise level at 25 metres from the track and assuming 1 train movement per hour. Note that, for the purpose of these calculations, these values do not include a façade correction. Table 7-1 External Noise Spectra for Freight Trains Parameter Octave Band Centre Frequency (Hz) k 2k 4k 8k 16k db(a) L Aeq(Day) / L Aeq(Night)* L Amax * Assumes 1 train movement per hour. To undertake the calculations, knowledge of the room size is required. Table 7-2 provides these on the basis of typical apartments. Higher end apartments are more likely to have larger glazing to floor ratios. Table 7-2 Typical Apartment Room Dimensions Description Bedroom Living Floor Area 3.4m x 3.6m = 12.24m 2 3.7m x 4.3m = 15.91m 2 Glazing 2.4m x 1.4m = 3.36m 2 (27% of floor area) 2.9m x 2.2m = 6.38m 2 (40% of floor area) Room Volume 31.8m m 3 Reverberation Time 0.5 seconds 0.5 seconds Reference: Final.docx Page 16

19 The computer programme Insul has been used to calculate the internal noise level on the basis of applying the Package B requirements, as outlined in the SPP 5.4 Guidelines. For multi-storey apartments, there may be a number of floors of apartments where the only noise paths are through the walls and glazing, whereas the top floor will also have contribution through the roof/ceiling structure. Table 7-3 provides the calculated internal noise levels. Table 7-3 Calculated Internal Noise Levels Typical Apartment Room Lower Level Apartment Top Floor Apartment L Aeq, db L Amax, db L Aeq, db L Amax, db Bedroom Living The results show the importance of the roof/ceiling contribution, which can dramatically increase noise levels. Currently the SPP 5.4 Guidelines do not specify the type of roof, so either Colorbond or clay tiles could be used. However, because of the low frequency component of freight rail, the use of Colorbond, as assumed in these calculations, will result in significant contribution via the roof/ceiling noise pathway. Taking the worst-case top floor apartment, the calculated indoor noise levels are 34 db L Aeq(Night) in bedrooms and 35 db L Aeq(Day) in living areas, both of which comply by 1 db and 5 db with the design goal of 35 db L Aeq(Night) and 40 db L Aeq(Day) respectively. The L Amax in bedrooms is 6 db above the design goal of 60 db L Amax, whilst in living areas there is a 7 db exceedance 3. Section 4 applied a 23 db reduction from outside (reflected noise level) to inside. For the worst-case top floor apartment, the noise reduction is 22 db L Aeq to bedrooms, 21 db L Aeq to living areas, 24 db L Amax to bedrooms and 23 db L Amax to living areas. This aligns reasonably well with the assumed acoustic performance of Package B. Where there is negligible ceiling contribution (e.g. lower level apartments), the noise reduction can be 5 to 9 db higher. As expected, the analysis shows that the deemed-to-comply construction is adequate, in this instance, for L Aeq noise levels, but inadequate for L Amax noise levels. As discussed, the major weakness is the roof/ceiling construction, due to the assumed Colorbond roof sheeting. Standard Colorbond roof/ceiling construction would have an acoustic performance of approximately R w + C tr 44 + (-9) = 35. By comparison, a roof/ceiling construction with a clay roof tile would have a typical performance of R w + C tr 37 + (-2) = 35. Although the overall acoustic performance appears to be the same (R w + C tr is 35 in both cases), the lesser C tr number indicates a much better low frequency performance for the clay tile roof. To make this point more clearly, standard roof/ceiling construction with Colorbond sheeting has a transmission loss of 7 db at 63 Hz, whereas with clay roof tiles the transmission loss increases to 22 db. Given that freight rail contains a significant amount of noise at this frequency, the low frequency characteristics of the building become critical. To achieve an equivalent level of acoustic performance with a Colorbond roof 3 There may be some argument that the L Amax in a living area is not as critical for the protection of residential amenity, since maximum noise impacts are often considered, primarily, a sleep disturbance issue. Reference: Final.docx Page 17

20 would require 4mm compressed fibre cement sheeting to be installed between the Colorbond and the roof purlins and the ceiling construction to comprise 2 layers of 10mm plasterboard. With a clay tile roof (or equivalent) in place, the calculated internal noise levels reduce by 5 and 7 db, to 59 db L Amax in bedrooms and 62 db L Amax in living areas. To achieve no more than 60 db L Amax in living areas would require the glazing to be upgraded, so that it was equivalent to the requirements for the bedrooms. Recommended improved architectural treatment packages for freight rail are provided in Section 9 of this report. Reference: Final.docx Page 18

21 8 Sensitivity Analysis To present the analysis of Sections 3 to 7, a number of assumptions have had to be made, each of which will result in different effects if changed. Some of the critical assumptions are discussed in this section of the report. 8.1 Analysis of Measured Data The process of measuring the L Aeq generally involves the following method: Setting up a noise logger on the subject site for a period of 4-7 days; For each complete day and night period, the L Aeq(Day) and L Aeq(Night) of trains is calculated, with these calculated values being the logarithmic average of each time period; and Each L Aeq(Day) and L Aeq(Night) value is then arithmetically averaged to obtain an overall value for the entire measurement period. Using the above approach for the Section 4 data meant that the average noise level was 53.3 db L Aeq(Night), however there was a worst-case level of 55.8 db L Aeq(Night) being 2.5 db higher. In this case, the change in L Aeq(Night) would still have resulted in Acceptable Treatment Package B being applied, however there would be some cases where taking a worst-case value would change the architectural treatment package required. Analysis of the L Amax is even more critical, as the range of noise levels is much larger. For instance, in the data set of Section 3, the range of values was 67.7 to 93.2 db L Amax, with a mean of 80.3 db L Amax. Of the 40 measured trains, 19 had a maximum noise level above the mean, with 9 having a maximum noise level of more than 87.3 db(a), being the mean plus one standard deviation ( + 1 ). The latter was considered reasonable for the analysis in this report, however using the mean or the highest maximum value (or any other statistical value) would significantly affect the results of this analysis. For instance, the L Aeq(Night) was assessed as 56 db (with façade correction), which exceeds the target by 6 db. By comparison, if the mean L Amax was used (with façade correction) this would exceed the design goal of 75 db L Amax by 8 db. Whilst this means the L Amax is still more critical than the L Aeq(Night), the difference between the two assessment methods is now not as large as assumed in the earlier analysis of results, which took the L Amax value to be the mean plus one standard deviation of the measured values. Hence, using the mean L Amax would tend to soften the findings of this report. Conversely, if the highest L Amax was used (i.e. 96 db L Amax with facade correction), the L Amax level would exceed the criteria by 21 db rather than the 15 db determined by this report. 8.2 Maximum Noise Criteria The analysis has used a target criteria of 75 db L Amax outside, deemed to be equivalent to 60 db L Amax inside. The basis of this is from the draft 2005 version of SPP 5.4 where the 75 db L Amax criteria was proposed. Two other relevant criteria are the World Health Organisation (WHO) and Australian Standard Acoustics Aircraft Noise Intrusion Building Siting and Construction (AS 2021). Reference: Final.docx Page 19

22 The WHO Guidelines recommend that, for the prevention of sleep disturbance, internal L Amax noise levels should not exceed 45 db more than 10 times per night. For this particular study, there would then be no L Amax criteria applicable to freight train activity, since, even assuming a worst-case of 1 train movement per hour, there would only be 8 movements at night. It could be argued that L Amax impacts from freight rail activity are not too dissimilar to the disturbance experienced from aircraft. Appendix E of AS 2021, provides the following sliding scale in terms of daily aircraft maximum noise levels: 70 db L Amax outside is acceptable where there are more than 30 events per day; 80 db L Amax outside is acceptable where there are between 15 and 30 events per day; and 90 db L Amax outside is acceptable where there are less than 15 movements per day. An important point to note is that any movement at night (7pm to 7am) counts as being the equivalent of 4 events. This would mean that, from the monitoring data, where there are typically 4 movements between 7am and 7pm and 4 movements between 7pm and 7am, this would be assessed as a total of 20 events. The forecast is to increase movements to 12 per day, which, if you assumed an even split of 6 during the day and 6 at night, would equate to 30 events. In both cases, 80 db L Amax would be deemed to be an acceptable external noise level, if the AS2021 approach was applied. Using 80 db L Amax as the criteria would mean that the calculated L Amax of 87 db would exceed the criteria by 7 db, compared to the L Aeq(Night) level exceeding the SPP 5.4 target by 6 db. (Note that AS 2021 requires measurements to be at least 3.5 metres away from reflecting surfaces so the 2.5 db façade correction has not been applied to the calculated L Amax.) Using the AS 2021 criteria, in this instance, shows that the L Amax and L Aeq(Night) are reasonably similar in impact. If the approach of assuming at least 1 movement per hour was considered, this would equate to 60 events under AS 2021 and therefore 70 db L Amax would become the applicable criteria. In which case, 87 db L Amax would exceed the criteria by 17 db. So, in this case, an L Amax assessment becomes much more critical than an L Aeq(Night) assessment. AS 2021 also sets indoor design sound levels of 50 db L Amax for sleeping areas and 55 db L Amax for other habitable spaces, which is more stringent than the internal design goal of 60 db L Amax used as the basis for this assessment of freight rail noise. 8.3 Glazing to Floor Ratio The calculations of Section 7 assume a typical glazing ratio to floor area of 27% to bedrooms and 40% to living areas. If these were to increase to 40% for bedrooms and 60% for living areas, the internal noise levels would change, as shown by Table 8-1. For bedrooms, internal noise level increases are predicted, however these are reasonably small (by no more than 1 to 2 db). For living areas, the performance of the glazing is required to change to R w + C tr 31 in accordance with the SPP 5.4 Guidelines, whereas previously it was R w + C tr 28. This results in internal noise levels actually decreasing in living areas but only by a small margin (by no more than 1 db). Reference: Final.docx Page 20

23 Table 8-1 Calculated Internal Noise Levels Larger Glazing Room Lower Level Apartment Top Floor Apartment L Aeq, db L Amax, db L Aeq, db L Amax, db Bedroom Living Bedroom glazing comprises 40% of floor area and living glazing 60% of floor area. Since the Acceptable Treatment Packages in the SPP 5.4 Guidelines work on a sliding scale, where the acoustic requirements for windows vary according to the percentage of the floor area, changing the window size is adequately controlled within Package B, in this instance. 8.4 Upper Limit of Packages For the example in this report, it was determined that the L Aeq(Night) was 56 db and the L Amax was 90 db (refer Section 4). In applying the SPP 5.4 Guidelines, the calculated 56 db L Aeq(Night) meant that Acceptable Treatment Package B was applicable, however the same treatment package would also apply if noise levels were at the upper limit of Package B. That is, if the development was to occur at around 17 metres from the railway (instead of at 25 metres) the external noise levels would be 58 db L Aeq(Night) and 93 db L Amax. Table 8-2 shows the predicted internal noise levels in a typical apartment, assuming that the external L Aeq(Night) levels are at upper limit of Package B. Note that L Aeq levels are 2 db higher than those given in Table 7-3 and L Amax levels are 3 db higher. Table 8-2 Calculated Internal Noise Levels Typical Apartment - Upper Limit Package B Room Lower Level Apartment Top Floor Apartment L Aeq, db L Amax, db L Aeq, db L Amax, db Bedroom Living Table 8-2 indicates that, even when assessed only against the L Aeq(Night) criteria in SPP 5.4, the existing Acceptable Treatment Package B may be inadequate in addressing freight rail noise for a top floor apartment, since the acceptable indoor noise objective of 35 db L Aeq(Night) in bedrooms is exceeded. As discussed in Section 7, additional acoustic treatment to the roof/ceiling construction and to living area windows would be required to properly attenuate L Amax noise levels from freight rail. With these acoustic improvements in place, but with external noise levels at the upper limit of Package B, the resultant L Amax internal noise levels are predicted to be 62 db L Amax in bedrooms and 63 db L Amax in living areas. Hence the acoustic performance of the window glazing in both bedrooms and living areas would need to be further improved by 3 db to achieve an internal noise level of no more than 60 db L Amax. Reference: Final.docx Page 21

24 9 CONCLUSION The analysis presented in this report shows that the SPP 5.4 assessment method is generally not adequate in addressing L Amax noise impacts from freight trains. It shows that, if an L Amax criteria were introduced for freight rail noise, in most situations it would dictate compliance when compared to the existing L Aeq assessment approach specified by SPP 5.4. This report also demonstrates that the architectural treatment packages provided by the SPP 5.4 Guidelines would be inadequate in achieving reasonable internal L Amax noise levels. Recommended architectural treatment packages for residential development have been developed specifically to address L Amax noise impacts from freight trains and are provided in Table 9-1. These packages aim to achieve an indoor design sound level of 60 db L Amax in both bedrooms and living areas. To aid understanding, an example form of construction is provided in Table 9-2 to show how the recommended architectural treatment packages for freight rail might be applied in practice. Tables 9-1 and 9-2 should be read in conjunction with Figure 9-1, which provides recommended land use planning guidance to assist in applying these packages. The recommended architectural treatment packages and the associated land use planning guidance is not intended to replace any requirements or criteria in SPP 5.4. Rather, their purpose is to help fully inform land use planning along rail freight lines. It is anticipated that they could be used by local government and developers to maintain an appropriate level of amenity within residential developments alongside freight rail lines. Reference: Final.docx Page 22

25 Figure 9-1 Recommended Land Use Planning Guidance for Freight Rail Noise Reference: Final.docx Page 23

26 Table 9-1 Recommended Acceptable Treatment Packages for Freight Rail Noise Area All Habitable Rooms (Including Kitchens) Orientation to Road or Rail Corridor Facing Freight Rail Package CF (up to 92 db L Amax ) Freight Rail Package BF (up to 88 db L Amax ) Freight Rail Package AF (up to 80 db L Amax ) Walls to R w + C tr 50. Windows and external door systems: Minimum R w + C tr 34 total glazing up to 40% of room floor area. R w + C tr 37 if 60%. Roof and ceiling to achieve minimum transmission loss of 22dB at 63Hz and overall R w + C tr 35 (e.g. clay roof tiles). Mechanical ventilation. Walls to R w + C tr 45. Windows and external door systems: Minimum R w + C tr 30 total glazing up to 40% of room floor area. R w + C tr 33 if 60%. Roof and ceiling to achieve minimum transmission loss of 22dB at 63Hz and overall R w + C tr 35 (e.g. clay roof tiles). Mechanical ventilation. Side As above. As above. As above. Opposite As above, except glazing may be 3 db less, or % increased by 20% (i.e. R w + C tr 34 for 60%). As above, except glazing may be 3 db less, or % increased by 20% (i.e. R w + C tr 29 for 60%). Walls to R w + C tr 45. Windows and external door systems: Minimum R w + C tr 28 total glazing up to 40% of room floor area. R w + C tr 31 if 60%. Roof and ceiling to R w + C tr 35. Mechanical ventilation. As above, except glazing may be 3 db less, or % increased by 20% (i.e. R w + C tr 28 for 60% or R w + C tr 31 for 80%). Reference: Final.docx Page 24

27 Area All Habitable Rooms (Including Kitchens) Outdoor Living Area Orientation to Road or Rail Corridor Facing Table 9-2 Example Construction for Freight Rail Noise Freight Rail Package CF (up to 92 db L Amax ) Freight Rail Package BF (up to 88 db L Amax ) Freight Rail Package AF (up to 80 db L Amax ) Walls: 2 x 110mm double brick wall with 50mm cavity and 50mm fibreglass insulation within the cavity. Window: 10.5mm VLam Hush awning windows (up to 40% of room floor area). External Doors: 10mm fully glazed hinged door (up to 20% of room floor area). External doors to bedrooms are not recommended. Roof and ceiling: Clay roof tiles with sarking and 10mm plasterboard ceiling, or, Colorbond roof sheeting with sarking, 4mm fibre cement sheeting fixed to the roof purlins and 2 x 10mm plasterboard ceiling. Mechanical ventilation. Walls: 2 x 90mm double brick wall with 20mm cavity. Windows: 6mm awning windows (up to 40% of room floor area); or, 10mm awning windows (up to 60% of room floor area). External Doors: 10mm sliding glass doors (up to 20% of room floor area). External doors to bedrooms are not recommended. Roof and ceiling: Clay roof tiles with sarking and 10mm plasterboard ceiling, or, Colorbond roof sheeting with sarking, 4mm fibre cement sheeting fixed to the roof purlins and 2 x 10mm plasterboard ceiling. Mechanical ventilation. Side As above. As above. As above. Opposite As above, except - Windows: 6mm awning windows (up to 40% of room floor area); or, 10mm awning windows (up to 60% of room floor area). External Doors: 6mm fully glazed hinged door (up to 20% of room floor area). Where practicable, locate an outdoor living area on the opposite side of the rail corridor or in an alcove on the side of the house. As above, except - Windows: 6mm awning or 10mm sliding windows (up to 40% of room floor area); or, 6mm awning windows (up to 60% of room floor area). External Doors: 6mm sliding glass doors (up to 20% of room floor area). Where practicable, locate an outdoor living area on the opposite side of the rail corridor or in an alcove on the side of the house. Walls: 2 x 90mm double brick wall with 20mm cavity. Windows: 6mm awning or 10mm sliding windows (up to 40% of room floor area); or, 6mm awning windows (up to 60% of room floor area). External Doors: 6mm sliding glass doors (up to 20% of room floor area). Roof and ceiling: Colorbond roof sheeting with 10mm plasterboard ceiling. Mechanical ventilation. As above, except - Windows: 4mm awning or 6mm sliding windows (up to 40% of room floor area); or, 6mm awning or 10mm sliding windows (up to 60% of room floor area). Where practicable, locate an outdoor living area on the opposite side of the rail corridor or in an alcove on the side of the house. Reference: Final.docx Page 25

28 Appendix A ACCEPTABLE TREATMENT PACKAGES

29 The packages and information provided on the following pages are taken from Implementation Guidelines for State Planning Policy 5.4 Road and Rail Transport Noise and freight Considerations in Land Use Planning; December Where outdoor noise levels are above the target level, excluding the effect of any boundary fences, the Guidelines propose acceptable treatment packages that may be implemented without requiring detailed review. The packages are also intended for residential development only. At higher noise levels or for other building usages, specialist acoustic advice will be needed. The acceptable treatment packages are intended to simplify compliance with the noise criteria, and the relevant package should be required as a condition of development in lieu of a detailed assessment. Transition between each package should be made on the basis of the highest incident L Aeq(Day) or L Aeq(Night) value to the nearest whole number determined for the building development under assessment. Any departures from the acceptable treatment specifications need to be supported by professional advice from a competent person that the proposal will achieve the requirements of the Policy. With regards to the packages, the following definitions are provided: Facing the transport corridor: Any part of a building façade is facing the transport corridor if any straight line drawn perpendicular to its nearest road lane or railway line intersects that part of the façade without obstruction (ignoring any fence). Side-on to transport corridor: Any part of a building façade that is not facing is side-on to the transport corridor if any straight line can be drawn from it to intersect the nearest road lane or railway line without obstruction (ignoring any fence). Opposite to transport corridor: Neither side on nor facing, as defined above.

30

31 Appendix B Terminology

32 The following is an explanation of the terminology used throughout this report. Decibel (db) The decibel is the unit that describes the sound pressure and sound power levels of a noise source. It is a logarithmic scale referenced to the threshold of hearing. A-Weighting An A-weighted noise level has been filtered in such a way as to represent the way in which the human ear perceives sound. This weighting reflects the fact that the human ear is not as sensitive to lower frequencies as it is to higher frequencies. An A-weighted sound level is described as L A db. L 1 An L 1 level is the noise level which is exceeded for 1 per cent of the measurement period and is considered to represent the average of the maximum noise levels measured. L 10 An L 10 level is the noise level which is exceeded for 10 per cent of the measurement period and is considered to represent the intrusive noise level. L 90 An L 90 level is the noise level which is exceeded for 90 per cent of the measurement period and is considered to represent the background noise level. L eq The L eq level represents the average noise energy during a measurement period. L A10,18hour The L A10,18 hour level is the arithmetic average of the hourly L A10 levels between 6.00 am and midnight. The CoRTN algorithms were developed to calculate this parameter. L Aeq,24hour The L Aeq,24 hour level is the logarithmic average of the hourly L Aeq levels for a full day (from midnight to midnight). L Aeq,8hour / L Aeq (Night) The L Aeq (Night) level is the logarithmic average of the hourly L Aeq levels from pm to 6.00 am on the same day. L Aeq,16hour / L Aeq (Day) The L Aeq (Day) level is the logarithmic average of the hourly L Aeq levels from 6.00 am to pm on the same day. This value is typically 1-3 db less than the L A10,18hour. R w This is the weighted sound reduction index and is similar to the previously used STC (Sound Transmission Class) value. It is a single number rating determined by moving a grading curve in integral steps against the laboratory measured transmission loss until the sum of the deficiencies at each one-third-octave band, between 100 Hz and 3.15 khz, does not exceed 32 db. The higher the R w value, the better the acoustic performance. Reference: Final.docx Page B1

33 Ll C tr This is a spectrum adaptation term for airborne noise and provides a correction to the R w value to suit source sounds with significant low frequency content such as road traffic or home theatre systems. A wall that provides a relatively high level of low frequency attenuation (i.e. masonry) may have a value in the order of 4 db, whilst a wall with relatively poor attenuation at low frequencies (i.e. stud wall) may have a value in the order of -14 db. Satisfactory Design Sound Level The level of noise that has been found to be acceptable by most people for the environment in question and also to be not intrusive. Maximum Design Sound Level The level of noise above which most people occupying the space start to become dissatisfied with the level of noise. Chart of Noise Level Descriptors Austroads Vehicle Class Reference: Final.docx Page B2

34 Ll Typical Noise Levels Reference: Final.docx Page B3