An Approach to Designing and Specifying Audibility Requirements of Train Activated Warning Systems.

Christabel Goode 1 AECOM Ltd 1 Tanfield, Edinburgh, EH3 5DA Seckin Basturk 2 AECOM Ltd 12 Regan Way, Chilwell, Nottingham, NG9 6RZ, UK James Block 3 AECOM Ltd 12 Regan Way, Chilwell, Nottingham, NG9 6RZ, UK Alex Southern 4 AECOM Ltd 120 Bothwell Street, Glasgow, G2 7JS, UK

ABSTRACT

Train Activated Warning System (TAWS) is a railway safety system which uses an audible warning to alert track workers of an oncoming train. When the system is active, a safe-tone sound is emitted to indicate the system is active. Activated by the train signalling system, the safe tone switches to a warning-tone when a train approaches, in enough time to allow rail workers to reach a position of safety. There is currently no specific methodology for designing a new TAWS to achieve the re- quired audibility over a defined area of track. This paper discusses the approach to the design and specification of the TAWS audibility over lengths of track where different prevailing acoustic envi- ronment conditions need to be considered. The challenges encountered and the limitations of the resulting approach are also discussed.

1. INTRODUCTION

Situations where railway staff are working on tracks open to traffic for purposes of inspecting, maintaining, enhancing or renewing the railway, have been in practice for many years in the UK. As this is considered riskier than other methods of working (e.g. stopping the trains), it is avoided

1 christabel.goode@aecom.com

2 seckin.basturk@aecom.com

3 james.block@aecom.com

4 alex.southern@aecom.com

worm 2022

where practicable. However, in recent years, owing to the increasing customer and stakeholder de- mands, working on tracks open to traffic is increasing. Therefore, managing the risks posed by working alongside lines open to traffic is one of the highest priorities of railway authorities [1]. De- spite having a number of safety methods in place, workforce casualties reported in recent years stresses the need for better safety procedures [2][3].

In the UK, as part of rail worker safety improvements, the use of human lookouts reliant on vis- ual identification of approaching trains is being phased out. Where possible, Network Rail are re- placing human lookouts with protection and warning technology to alert groups of workers to ap- proaching vehicles on the track [4]. There are a number of different technologies that are currently used and under development to help with the trackside safety objectives in the industry. One of these is the Train Activated Warning System (TAWS).

The TAWS emits a warning signal when a train is approaching the works area and a safe signal when there is no train. The systems are activated by the rail workers when working or access is re- quired within the TAWS signal reception area (SRA). The SRA includes all areas that are consid- ered to be on or near the line.

There is very limited guidance for designing a new TAWS to achieve the required audibility over a defined area of track however objective audibility specifications exist that can be used to de- termine if a given signal is audible. There is also a lack of specific guidance on how to balance the need for audibility in the SRA with minimising the potential environmental noise impact on nearby noise sensitive receptors (NSRs).

This paper presents a complete approach to the design and specification of a TAWS that is audi- ble over lengths of track where different prevailing acoustic environment conditions need to be con- sidered and where the potential impact on noise sensitive neighbours also needs to be considered. 2. RELEVANT STANDARDS AND GUIDANCE

While there is no specific TAWS guidance outlining a design methodology to balance audibility needs with environmental noise impact, there are existing relevant standards. The following can be used as a reference for TAWS installations in the UK:

• Network Rail (NR) Standard “Train Activated Warning Systems” (NR/L2/SIG/30009/Z115) (the NR Standard) [5] specifies the design rules for the implemen- tation of a TAWS: This is the primary reference for designing a TAWS to be compliant for use on the NR infrastructure. In addition, NR requires any audible warning device used to be certified by them. • International Standard ISO 7731:2003 “Ergonomics – Danger signals for public and work areas– Auditory danger signals” [6]: This provides design guidelines for auditory danger sig- nals to be used for public and work areas. It defines audibility limits based on the importance of the signal, namely whether it is a “safe” signal or a “warning” signal.

• British Standard EN 16704-2-1:2016 “Railway applications. Track. Safety protection on the track during work. Common solutions and technologies. Technical requirements for Track Warning Systems (TWS)” [7]: This covers different types of warning systems, including look- out operated warning systems which are human led and technology led systems, automatic track

worm 2022

warning systems or signal-controlled warning systems. Annex A of this standard provides a spec- ification for acoustic warning signals. However, the NR standard takes precedence over this Brit- ish Standard and some of the differences will be highlighted in this paper.

• British Standard 4142:2014+A1:2019 “Methods for rating and assessing industrial and commercial sound” [8]: Whilst not directly applicable for assessing audibility, this standard is used to assess the impact of industrial and commercial sound on people at nearby receptors by comparing the difference between the background sound level and the rating level of the sound under assessment. The rating level is an adjusted sound level which includes penalties based on the acoustic character of the sound such as tonality, intermittency, and impulsivity. In this paper this standard is used to assess the impact from the TAWS operation on nearby NSRs. The standard states that “typically, the greater difference (between background and rating sound levels), the greater the magnitude of impact”. 3. DESIGN APPROACH

The basic premise of a TAWS is that when activated by workers entering the signal reception area, all connected loudspeakers (Audible Warning Devices - AWD) mounted on posts beside the track will emit regular short tone bursts of sound, known as the safe signal. The safe signal can pro- vide audible reassurance to the rail workers in the SRA that it is safe to continue to work on or near the line within the established TAWS limits without the risk of the hazard posed by oncoming trains.

When a train is detected approaching the TAWS area by the signalling system, the TAWS sound automatically changes in both level and character to a continuous alternating tone. This alerts the rail workers and gives them time to move to a Position of Safety (PoS) until the safe signal re- sumes.

The ability of rail workers to hear and react to the change from safe to warning signal is critical to their safety. It is important to understand that the purpose of an acoustic TAWS design is not to guarantee a TAWS will be audible at a given location in all working conditions. Instead it is the re- sponsibility of the rail operative to confirm that the safe signal is audible for the types of work be- ing undertaken and the conditions encountered while on site. If at any time during the work tasks, the safe tone cannot be heard, the rail workers must move to a PoS. The warning signal is a louder and continuously sounding signal that is easily distinguishable from the safe signal, a design choice that means the warning signal will be inherently audible if the railway operative can hear the safe signal.

In this work we propose to specify a system that will provide a safe signal sound level that is au- dible against the typical ambient acoustic environment. This means that activities being undertaken by rail workers will tend not to be temporarily interrupted due to the safe signal being inaudible as a consequence of other sound sources in the area. This includes noise generated by the work being undertaken, which may also require hearing protection. Sources such as road vehicles, industrial or commercial premises that are louder on track compared to the AWDs could mean work needs to halt if the safe-tone cannot be heard reliably and the work re-planned. In some circumstances it may be necessary to implement an alternative safe system of work to TAWS if the safe tone cannot be heard reliably. Acoustic planning should be a consideration when defining the safe work plan such that work in noisier environments is minimised at the planning stage. For example BS EN 16704-1 “Railway applications. Track. Safety protection on the track during work Railway risks and com- mon principles for protection of fixed and mobile work sites.”

worm 2022

The ambient sound level within an SRA will fluctuate in time. Therefore, an adaptive system which monitors the ambient sound levels within defined working areas during the works and adjusts the output level of the AWDs would be the ideal. However, due to the properties of the certified AWD devices, the current infrastructure and NR guidance, currently in the UK a TAWS design is limited to fixed sound emission levels.

Therefore, in this paper, the proposed TAWS acoustic design is concerned with first understand- ing the prevailing ambient sound climate and then selecting a suitable number and placement of AWDs to provide a system that can be used robustly in typical ambient sound conditions for a given site. Concurrently, TAWS acoustic design should also consider the potential for causing disturbance to nearby NSRs.

The following steps outline the approach used for designing a TAWS that covers a large area, i.e. >> 100m, or run through a town or city, or rural community: • Undertake a desk-based investigation to plan the baseline sound measurement locations and de- termine the NSRs plus note any other local acoustic features (e.g. other sound sources); • Undertake baseline sound monitoring to determine the ambient sound environment experienced by staff on track across the TAWS site and at positions representative of nearest NSRs; • Establish a 3D acoustic propagation model of the TAWS and surrounding areas, including ground topography and building heights; • Determine the acoustic performance of the warning device based on manufacturer’s data or la- boratory measurements • Assess the audibility of the warnings – using the acoustic performance of the audible warning and typical highest ambient noise level; • Assess the environmental noise impact of the audible warnings on the nearest NSRs; • Provide a document of the TAWS area detailing the quantity, spacing, orientation, alarm sound levels of the AWD, and acoustic screening if needed; and • Undertake post installation monitoring to test compliance of each AWD. 4. PROPOSED TAWS ACOUSTIC DESIGN

4.1. Determine Prevailing Ambient Sound Levels on Track and at Sensitive Receptors The desk-based investigation is used to identify NSRs and determine locations for the installa- tion of unattended sound level monitoring equipment. The duration of the unattended monitoring should be as long as is practical, and generally at least one full week is required to obtain an under- standing of the prevailing ambient sound levels that could typically cause masking of the TAWS signals. Periods in which atypical conditions are present, such as periods with temporary construc- tion works in the area or public/school holidays should be avoided. One extreme can lead TAWS to be designed louder than required for the area and cause unnecessary disturbance to sensitive neigh- bours while the other can result in a TAWS design where rail workers might find the safe-tone audi- bility intermittent resulting in the interruption of works.

Measurements clearly influenced by inclement weather and noise emissions from train pass-bys are generally excluded from the ambient sound level analysis as the TAWS is not intended to be ef- fective during a train pass-by but on its approach to the work area. Similarly inclement weather can contribute a variable increase the ambient sound levels but is excluded from the ambient sound level analysis due to the variability. Instead it is left to site worker to ensure they can hear the safe-

worm 2022

tone in the prevailing conditions determined by weather and working conditions (machinery noise) while wearing whatever personal protective equipment (PPE such as helmets, caps, ear defenders) and weather protection (e.g coats with hoods) is required.

In addition, extraneous sound sources that cause infrequent short duration peaks in the recorded ambient sound levels, such as sounds of wildlife near the monitoring equipment, can also be consid- ered atypical. The short duration of such one-off sound events makes it unlikely that they would mask the AWD signals for a significant period of time.

Considering the above, the typical ambient sound level to inform the trackside audibility assess- ment, and therefore the TAWS design, is proposed to be the “Typical L Smax at 2.5kHz”. That is the maximum noise level ( L Smax in the 2.5 kHz one-third octave frequency band within which the AWD predominantly operates) exceeded for 10% of the relevant measurement period. Examples of the measured time-history graphs of measured ambient sound levels and the identified typical L Smax at 2.5 kHz for the ambient sound are provided in Figure 1 and Figure 2. In the first figure the high lev- els are generated by train pass-bys and are removed from the audibility analysis.

Where significant differences in baseline conditions (e.g. existing sound level and character) are found within a TAWS site, it is subdivided and multiple measurements made to determine the am- bient sound levels for the sub-areas.

— smactninQoatband) AWA

Figure 1: Measured ambient sound levels – typical day

worm 2022

Figure 2: Measured ambient sound levels – typical 1 hour 4.2. TAWS Audibility Assessment Based on the identified typical highest ambient sound levels across the SRA, the Effective Masked Threshold (EMT) is determined, following “Method c)” 1 of ISO 7731. Method c is the preferred method for determining the audibility because it accounts for the spectral detail of the masking sound. Relevant EMTs are determined for each of sub areas if significant differences in ambient sound are encountered.

“Te oh)

The 3D acoustic model of the area is used to investigate iterations of AWD layouts and orienta- tions to optimise the design of the TAWS to provide sound levels equal or higher than the audibility criterion provided in ISO 7731, i.e. signal level ≥ EMT+13. The 1/3 octave band considered in this work is the 2.5 kHz frequency band because the AWD was most dominant in this band. During this iteration phase, the audibility of the signal is tested on a 1x1m grid across the SRA. As a result of this process, considering both the track audibility and the environmental noise impacts at nearby NSRs (see Section 4.3 for details), the number of AWDs required, the optimum positioning and ori- entation of each AWD, and acoustic screening requirements are determined. An example specifica- tion table is given in Table 1, and corresponding TAWS signal audibility results are shown in Fig- ure 3, for the same example site.

1 Where ambient noise levels are established as one-third octave frequency band sound pressure levels (measurement method c) in ISO 7731), in order to achieve audibility, the sound pressure level of the signal in one or more one-third octave frequency bands shall exceed the effective masked threshold by 13 dB in the one-third octave frequency band under consideration.

worm 2022

Table 1: TAWS AWD design specifications

AWD

AWD units to be

Screening requirement

AWD sound emission level

Post

installed and

ID

orientation

(in 2.5 kHz)*

1 2 units per post

50 cm long 30 cm high screen on one side of the AWD post. AWDs to be located at surface (centre) of this facing the rail line. The screen is curled at top and bottom to form

Safe Signal: 91 dB L Smax

(positioned back-to-back)

2

Warning Signal:

AWD axis to

3

intersect the track at 75 o

10 cm wide screens.

96 dB L Smax

4

* Measured at 1m in front of each AWD unit (on-axis).

Figure 3: TAWS trackside audibility assessment results 4.2.1. Alternative method of establishing TAWS audibility (BS EN 16704-2-1:2016)

In BS EN 16704-2-1:2016 the audibility of the warning signal is determined when the system confirms the signal is 3dB or more over the maximum sound pressure level at the ear of the worker. This is supposed to be achieved by the sound power level of the AWD being automatically adapted based on the changing ambient level at the AWD.

The maximum sound pressure at the ear of a worker could be excessive depending on what they are doing and what is happening on site. While this should confirm the warning is audible it also has the potential to startle or momentarily panic workers as the level could increase suddenly. This could be a safety risk in itself if it happens as a train is approaching. In addition, it makes the sys- tem much more likely to be an annoyance to NSRs due to higher levels or even due to the tempo- rally varying warning sound level itself.

worm 2022

4.3. Determining TAWS Noise Impact at Nearby Sensitive Receptors The environmental noise impact of the TAWS is assessed following the principles set out in BS 4142. Where NSRs are relatively close to the railway, as in the example given in Figure 3 a BS 4142 assessment is expected to indicate significant adverse impacts when the warning signal is sounding, as the difference between rating and background levels is likely to be up to 20 dB.

The discussion of the context is considered important when utilising the methodology in BS 4142. For the TAWS assessment, it is considered important to note that the system is anticipated to operate infrequently; receptors close to the railway will already be likely to expect to hear railway operational sounds; and that TAWS are being introduced to improve safety on the railway. They can allow works that would otherwise typically need to be undertaken during a night-time closure of the track to be carried safely out during the day while the railway is operating. In practice, TAWS will operate only when track workers require access to or passage through a section of track during the daytime and therefore impacts are likely to be localised, temporary and of short duration when they occur.

Nonetheless, to minimise the noise impact, AWD acoustic screening (shielding/baffle) options can be developed and incorporated into the TAWS design. The incorporation of these can help in directing the sound energy towards the railway and reduce the propagation outside the works area. Such screens can reduce the BS 4142 rating level at the receptor by up to 7 dB. An example screen design is shown in Figure 4.

Figure 4: AWD Screening design (also described in Table 1) 5. CONCLUSION This paper has presented and discussed an approach to the design and specification of TAWS in conjunction with assessing the impact on nearby noise sensitive receptors. The motivation for this work has been on the basis that a prior methodology for this purpose is not available.

The proposed TAWS acoustic design method can be used to achieve the required audibility on the track under typical conditions, based on ISO 7731 criteria. However, under the following cir- cumstances the audibility of TAWS signals can be reduced:

worm 2022

• Atypical loud events that last more than a few seconds that can mask the safe or warning signal; • Noise from equipment used by the workers may be high enough to mask the safe or warning signal for the operator of the equipment and others in the vicinity; or • Hearing impaired workers and workers using hearing protection or wearing hoods may not be able to hear the safe or warning signal. Under the above circumstances the workers would stop work as soon as they are not able to hear the safe signal. Therefore, the safety system would still work as intended. However, if the above conditions occur frequently and/or for long durations, it can cause a significant disruption to the planned works. This can be avoided by: • An adaptive TAWS that can adjust the output sound levels based on real-time monitoring of ambient sound levels near the workers; and/or • The use of other warning systems, e.g. visual, in combination with TAWS to reduce the likelihood of work interruptions and to provide an additional layer of safety. In the UK, the use of adaptive TAWS is possible only if suitable equipment (AWDs, real-time monitoring equipment, etc.) is certified by the Network Rail. However, it is also important to note that such adaptive systems’ installation and maintenance costs would be significantly higher than the TAWS with fixed output sound levels and that more complex systems have more points at which they could potentially fail.

This paper has introduced and discussed the design and specification approach to making a TAWS audible over lengths of track where the acoustic environment needs to be considered both to achieve signal audibility and minimise environmental noise impact.

6. REFERENCES

1. Network Rail. Any Line Open . Available from: https://safety.networkrail.co.uk/safety/adjacent-

line-open/ 2. Saraogi, V. Getting on track with safety for railway workers . 2019. Available from:

https://www.railway-technology.com/analysis/getting-on-track-with-safety-for-railway-work- ers/ 3. Office of Rail and Road. Annual Report of Health and Safety on Britain’s Railways. 2021. 4. Network Rail. Keeping our workforce safe . Available from: https://www.networkrail.co.uk/run-

ning-the-railway/keeping-our-workforce-safe/ 5. Network Rail. Train Activated Warning Systems (TAWS). NR/L2/SIG/30009/Z115. 2011. 6. International Organization for Standardization. ISO 7731:2003 Ergonomics – Danger signals

for public and work areas – Auditory danger signals. 2003. 7. BSI Group. British Standard EN 16704-2-1:2016 Railway applications. Track. Safety protec-

tion on the track during work. Common solutions and technologies. Technical requirements for Track Warning Systems (TWS) . 2016. 8. BSI Group. British Standard 4142:2014 + A1:2019 Methods for rating and assessing industrial

and commercial sound. 2019.

worm 2022