Practical Considerations for
designing Road Tunnel Public
address Systems
Peter J Patrick
Genesis
Governments world wide have legislated intelligibility into the
requirements for audio announcement systems for emergency /
evacuation control.
This applies to all situations and locations - including public road
tunnels, bus-way tunnels and associated egress tunnels
Design Environment
• In spite of the critical nature of the announcement system
in tunnels the cost of the system is small in overall terms.
• Project managers in the past have not understood the level
of expertise required.
• Designs have been provided by manufacturers distributors
and others but the outcomes have not been meritorious.
• In fact few if any Australian Tunnels appear to have been
equipped with announcement systems which meet the
requirements of AS 1670.4
The Tunnel itself
Cable Tray
Painted smooth concrete
Jet fan
Jet fan
Vitreous Enamelled Steel
“Architectural Panels”
Roadway
Pre Cast smooth
natural Concrete
What’s the required Outcome ?
1.0
“Excellent”
0.75
“Good”
0.6
“Fair”
0.5
0.45
“Poor”
0.3
“Bad”
Design Procedure in common use
EASE & EASERA were used in
this investigation - CATT and
Odeon should give the same or
similar results
Including Noise levels
Alternative method
Pink Noise
gen
Graphic
Equaliser
Audio Signal
Mixer
Factors affecting the outcome
• Reverberation
• Noise
• Early / Late energy ratio
• Loudspeaker arrivals
• Reverberant decay
• Echoes
• Masking
• Fidelity
• Distortion
The STIPa test for STI
Figure 2. Easera SysTune Display - Spectra produced by STIpa test signal generator (NTi Audio MR-PRO)
STI Rating
Masking with increasing SPL
Figure 3. SPL vs STI reading according to NTI Audio, XL2
Creating an Accurate Model
Reverberation Time
The influence of open tunnel
ends
Figure 4. Graph showing proportion of total tunnel surface occupied by ends vs. length in metres for three tunnel sizes.
Putting that another way
Avge α vs Length for wall alpha of 0.01
Tunnel cross section 10m (H) * 20 m (W)
0.14
0.12
0.1
0.08
Avge α
0.06
0.04
0.02
1350
1250
1150
1050
950
850
750
650
550
450
350
250
150
50
0
The influence of surface material
choice
Figure 5. Reverberation Time vs. wall absorption coefficient
Reverberation Time in Seconds
in increments of 0.005 from .005 to 0.2
Combining the effects
30.00
Road Tunnel α = .015
Road Tunnel α = .02
Road Tunnel α = .025
Bus Tunnel α = .015
Bus Tunnel α = .02
Bus Tunnel α = .025
20.00
Egress Tunnel α = .015
Egress Tunnel α = .02
Egress Tunnel α = .025
15.00
Egress Tunnel α = .05
10.00
5.00
Tunnel Length in Metres
20,000
16,000
12,500
10,000
8,000
6,300
5,000
4,000
3,150
2,500
2,000
1,600
1,250
1,000
800
630
500
400
315
250
200
160
125
100
80
63
0.00
50
Reverberation Time in Seconds
25.00
Translating Absorption coefficient
to Direct/Reverberant outcomes
D-R Ratio vs Abs. Coefficient for 2 Km Tunnel w source in centre
40
D-R ratio in dB, RT60 in seconds
30
20
RT60 in
Seconds
10
D/R 1m from
source
D/R 10m from
source
0
0.6
.55
0.5
.45
0.4
.35
0.3
.25
0.2
.15
-10
-20
Absorption Coefficient in .01 steps
0.1
0.05
D/R 50m from
source
And the effect on STI Outcomes
But at this end of the scale an
error of 0.1 in á value produces
an STI error of 0.08
At this end of the scale an
error of 0.1 in á value
produces an STI error of 0.2
100
95
90
85
80
75
70
65
60
55
50
45
40
35
30
25
20
15
10
5
0
STI *100, 10m
from source
STI *100, 50m
from source
Absorption Coefficient in .01 increments
0.05
0.1
.15
0.2
.25
0.3
.35
0.4
.45
.5
.55
RT60 in
Seconds *3
0.6
STI * 100, RT60 * 3
STI Vs Abs Coefficient for 2Km Tunnel w source in centre
A closer look at the relationship
between RT60 and STI
STI vs RT60
0.8
0.75
0.7
0.6
0.55
STI
0.5
0.45
0.4
0.35
Reverberation Time in seconds
6.66
6.33
6
5.66
5.33
5
4.66
4.33
4
3.66
3.33
3
2.66
2.33
2
1.66
1.33
1
0.3
0.66
STI reading
0.65
Noise Data
Octave
In Shed
Corrected for
Free Field
63 Hz
80.1
78.4
125 Hz
81.1
81
250 Hz
85.6
86.6
500 Hz
77.4
78.8
1,000 Hz
79.5
78.2
2,000 Hz
78.9
78.5
4,000 Hz
75
75.2
8,000 Hz
70.1
71.8
Table 1. Noise data provided by client showing octave band noise levels for a typical axial fan
System Topography
Location naming convention
L1
1/2 L1
Seat 1
1/2 L1
Seat 2
Approximate Centre of
Tunnel under test
Testing anechoic models
Isotropic Loudspeaker spacing vs
STI
STI foR seat directly below centre loudspeaker (S1)
0.96
STI for seat betw een central loudspeakers (S2)
0.94
0.9
0.88
Figure 8. Loudspeaker spacing vs. STI in
anechoic environment - isotropic radiators
0.86
Loudspeaker separation in m etres
10m
9.5m
9m
8.5m
8m
7.5m
7m
6.5m
6.0m
5.5m
5.25m
5.0m
4.5m
4.0m
3.5m
0.84
3.0m
STI reading
0.92
STI vs spacing for horn speakers
0.91
STI for seat directly below
centre loudspeaker (S1)
0.905
STI for seat between central
loudspeakers (S2)
0.895
0.89
0.885
0.88
0.875
0.87
0.865
Loudspeaker separation in metres
20.0m
19.0m
18.0m
17.0m
16.0m
15.0m
14.0m
13.0m
12.0m
11.0m
10.0m
9.0m
8.0m
7.0m
6.0m
5.0m
0.86
4.0m
Calculated STI reading
0.9
Echo Criteria
Echo Criteria for increasing loudspeaker spacing - single spaced
spheres and double spaced, back to back horns
1.6
1.4
1.2
Echo Criteria directly
below sphere (S1)
1
Echo Criteria between
spheres (S2)
0.8
0.6
Echo Criteria directly
below Horn (S1)
0.4
Echo Criteria between
Horns (S2)
0.2
10.0 / 20m
9.5 / 19m
9 / 18m
8.5 / 17m
8 / 16m
7.5 / 15m
7 / 14m
6.5 / 13m
6.0 / 12m
5.5 / 11m
5.0 / 10m
4.5 / 9.0m
4.0 / 8.0m
3.5 / 7.0m
3.0 / 6.0m
0
Level where
untrained listeners
perceive an echo
Level where trained
listeners perceive an
echo
Echoic Egress Tunnel tests
Echoic Egress Tunnel tests
Loudspeakers in 40m
spaced line 118ms delay
Loudspeakers back to back
in 40m spaced line no delay
Loudspeakers in 20m
spaced line no delay
Anechoic Road Tunnel System Tests
Spheres
Small
Horns
Medium
Horns
Large
Horns
Premium
Infinite
Horns
Boundary
Echo Criteria Seat 1
1.15
1.58
1.33
1.36
1.16
0.92
Echo Criteria Seat 2
1.32
1.7
1.33
1.34
1.23
1.15
STI Seat 1
0.663
0.608
0.587
0.584
0.576
0.75
STI Seat 2
0.735
0.64
0.608
0.593
0.572
0.696
Centre Time Seat 1
52ms
85ms
113ms
116ms
118ms
31ms
Centre Time Seat 2
37ms
60ms
89ms
93ms
94ms
47ms
Spheres
Small
Horns
Medium
Horns
Large
Horns
Echo Criteria Seat 1
2.14
2.75
2.52
2.55
2.07
1.65
Echo Criteria Seat 2
2.36
3.41
2.56
2.52
2.26
2.37
STI Seat 1
0.862
0.737
0.71
0.69
0.646
0.872
STI Seat 2
0.77
0.739
0.685
0.686
0.705
0.755
Centre Time Seat 1
17ms
72ms
125ms
137ms
144ms
18ms
Centre Time Seat 2
40ms
45ms
55ms
55ms
59ms
44ms
Spheres
Small
Horns
Medium
Horns
Large
Horns
Echo Criteria Seat 1
2.92
3.83
2.94
2.76
2.45
1.14
Echo Criteria Seat 2
2.97
4.27
3.43
3.17
2.81
0.55
STI Seat 1
0.887
0.873
0.928
0.93
0.918
0.963
STI Seat 2
0.576
0.813
0.878
0.898
0.923
0.998
Centre Time Seat 1
44ms
35ms
22ms
20ms
17m s
27ms
Centre Time Seat 2
108ms
45ms
19ms
16ms
13m s
11ms
Premium
Infinite
Horns
Boundary
Premium
Infinite
Horns
Boundary
Effects of Added Noise I
20 m
Back to
118ms Sequential delay
spaced back 40m
Medium Medium Medium Large Premium
Horns
Horns
Horns Horns
Horns
STI Seat 1 0.505
0.617
0.736
0.732
0.751
STI Seat 2 0.533
0.595
0.686
0.698
0.758
Table 3. STI from anechoic tests with octave band noise
Comparison of test methods
20 m spaced medium horns
No Noise XL2
With Noise
No Noise
Meas
Easera Calc
Easera calc
STI Seat 1
0.587
0.57
0.505
STI Seat 2
0.608
0.57
0.533
40 m spaced Back to back medium horns
With Noise
XL2 Meas
.5 *
.51 *
No Noise XL2
With Noise
With Noise
No Noise
Meas
Easera Calc
XL2 Meas
Easera calc
STI Seat 1
0.71
0.64
0.617
.57 *
STI Seat 2
0.685
0.65
0.595
.61 *
40 m spaced sequential delay premium horns
No Noise XL2
With Noise
With Noise
No Noise
Meas
Easera Calc
XL2 Meas
Easera calc
STI Seat 1
0.918
0.93
0.751
0.76 *
STI Seat 2
0.923
0.94
0.758
.86 *
* - readings obtained by adding signal and noise and adjusting total
level to ~ 70 dB(A)
Table 3. STI from anechoic tests with octave band noise
Echoic Road Tunnel
Model is :• 20m (W) * 10m (H) * 1050m (L)
• Walls, Floor & Ceiling à = 0.05, ends absorbers
Road Tunnel Echoic test results same loudspeakers - three
topographies
STI
EkGrad
20m horns
Seat 1
Seat 2
0.23
0.24
1.43
1.1
40m B-B
Seat 1
Seat 2
0.31
0.27
1.21
0.98
40m sequential delay
Seat 1
Seat 2
0.62
0.61
0.68
0.67
Each Echoic test took ~ 7 days in a standard ray trace
routine. That’s three weeks testing for six listener
seats and three system designs.
The effect of Noise Sources
Seat 1
Seat 2
Distance v s SPL for direct & rev erberant sound
90
85
Direct
SPL
Reverbera
nt SPL
75
70
65
60
Distance in m
22
19
16
13
10
7
4
55
1
dB SPL
80
Straight Tunnel Time Alignment
110ms
0ms
110ms
125ms
250ms
220ms
30ms between
first and last
arrival
Curved Tunnel Time Alignment
50m radius curve
110ms
0ms
125ms
110ms
250ms
203ms
47ms between
first and last
arrival
Loudspeaker Fidelity
Figure 13. Easera Display of Frequency response of common use horn
speaker from Ease/ Easera IR Transfer
Speaker #1 @ 4, 8, 12 & 20m
Speaker #2 @ 4, 8, 12 & 20m
Speaker #3 @ 4, 8, 12 & 20m
• The native behavior of any sound system topography should be first proven in an anechoic environment
before implementing in a tunnel environment.
• Each large, fixed noise source, should be complemented with a nearby companion loudspeaker. to maximise
signal to noise ratio. The distance between these companion loudspeakers should then form the basis for the
rest of the design so that the string of intermediate loudspeakers is set at equidistant intervals between fans.
• Whilst the down-tilt of the loudspeakers was treated arbitrarily in this document it is nonetheless a critical
feature to be optimised in any design to suit the height of the loudspeaker and geometry of the tunnel
• Any model of a tunnel should include the full dimensions, particularly tunnel length, wherever possible. The
reliability of calculations made relate to the proportion of tunnel length modeled as shown in figures 4 & 6.
Significantly truncated tunnels will produce significantly optimistic calculated outcomes.
• It is unlikely that highly reliable calculations can be made in the presence of the hostile acoustic
environment found in long tunnels as currently built. Calculations based on structures composed of material
data sets of insufficient accuracy as described in figure 5 and associated text, are likely to render outcomes at
substantial variance with the final result.
• Computer resource restrictions remain a serious obstacle to the derivation of detailed design work. The
statistical analysis calculation engines deliver reasonable outcomes in a short space of time for plain
distributed systems but can not accommodate a sequential delay system. Detailed analysis of sequential delay
systems may take months to conclude using common ray trace technology. Computer cloud systems where a
subscriber uploads a model to a large networked computer system may be available in the near future.
• Time alignment of sequential delay systems must be critically adjusted where road curvature is
encountered.
• Loudspeaker selection should include examination of frequency response to reconcile equalisation needs
with system dynamics and STI requirements. Equalisation must be done by measuring at several locations.
Finally
In general it is unlikely that ‘good’ levels of intelligibility will
ever be delivered in a road tunnel audio system until some
measure of control over reverberation time is available. The use
of sound absorbing concrete, unpainted blockwork or some
similar product with absorption coefficients of the order of 0.1
would add a significant measure of sabins to the quota presently
found, substantially improve the outcome, and improve the
reliability of the modeling process.