Rick Morrison, principal lighting designer for Aecom in Australia gives this insight on the reflective requirements of a tunnel lining.
Lighting design for transport tunnels is defined in Australian standard AS1158.5 (2007) 'Lighting for Roads and Public Spaces Part Five: Tunnels and Underpasses', and also in European standard CIE 088:2004 'Guide for the Lighting of Road Tunnels and Underpasses'. In these standards the nature and quality of the interior wall linings is described. In AS1158.5, section E4.4 dictates that surfaces other than the carriageway in the tunnel lighting model shall be perfectly diffusing (i.e. lambertian), while the European standard CIE88 does not specify a surface reflective quality.
It can be seen from the case history tabled herein, that the nature of tunnel wall linings is not accurately understood and that the potential for reflected glare from vehicle headlamps is affected by the overall performance of the interior reflectance inside the tunnel.
The lighting installation for the Eastern Busway Two Buranda to East Brisbane, Logan Tunnel was commissioned on 13 July 2011. Like all modern transport tunnels, the Logan Tunnel was clad with interior wall panels, the characteristics of which were carefully specified in the design documents. The specification called for a minimum of 60 per cent reflectance factor and for the surface to be diffusing (lambertian) in nature.
The application of the reflective panels is intended to provide a higher degree of ambience in the tunnel, improving the visual performance of drivers. With reflective walls the comfort level of the driver is improved, as the dark tunnelling effect is reduced. Comments were received from several parties involved this project as well as the Wellington tunnels alliance that indicated a level of concern regarding the potential for reflected glare from the linings. To better understand the reflective performance of the panels, a series of test measurements were conceived. A series of luminance measurements were proposed to measure the brightness resulting from headlamp reflectances in the tunnel. This paper tables the results of this case study in the Logan Tunnel.
Method and equipment
The measurement tool used would be the same luminance meter that was being used to test and commission the carriageway luminance values. The luminance meter was a Konica - Minolta LS-110.
The meter was mounted on to a staff to provide an observer eye position of 1.5m above the carriageway. The interior lining of transport tunnels is designed to add to the reflection of lighting within the space of the tunnel, and is specifically required to have a reflectance value of at least 60 per cent, and in general practice are often supplied in white or light colours. They have an inherent 'gloss factor' - which often provides the added bonus of reflecting the interior of the tunnel ahead of the driver in a way that allows a predictive cognition. Using a pole mounted luminance meter at 1,350mm height above the carriageway, measurements were set up as follows.
A fixed observation point of 60m inside the tunnel at the drivers' observation position of the lane facing the oncoming traffic lane was adopted. A Mazda Bravo Cab Chassis Ute - 2004 with Halogen headlamps was used in the test. The vehicle was positioned at the start of the portal entrance.
Measurements were taken for luminance on barrier surface adjacent to vehicle, carriageway in front of vehicle, bright spots on wall adjacent to vehicle and the headlamps in low and high beam. The vehicle was then moved 10m closer to observer position and the measurements taken again. Successive 10m moves were taken with measurements until the vehicle was at the 50m point.
The switched level for the tunnel lighting system was Night Switch.
Wall Linings - lambertian or semi-specula
A film was also taken with the vehicle driving from one end of the tunnel to the other in order to capture the essence of the visual scene within. The photographs of the interior scene during the measurements also contribute to the understanding of the tunnel space.
The footage and photos show that the interior lining in the tunnel is not lambertian in reflectance, but of a semi-specula nature.
The published information from Alpolic Linings indicates the reflectance is split as 30 per cent specula (glossy), and 45 per cent diffuse.
This situation is not mentioned in AS1158.5 (2007) where reflectance is described as diffusing, or in CIE88 (2004), which does not mention reflectance, however it is considered in CIE189 (2010) - where they discuss the inter-reflections in the tunnel.
The author has discussed the possibility of having the wall lining material tested for an 'R' table, with several photometric laboratories and manufacturers with photometric facilities. R tables for wall linings similar to the R tables for road carriageway luminance calculations might be able to be used in radiosity software to improve the accuracy of the design process. Whilst the R tables for wall materials can be produced, the software and lighting product companies have expressed doubts as to the usefulness of this due to the narrow field of view (alpha = 1° ? R-Table), wherein the R table is expressed in a one degree field of view as from an observer position - and this would not assist in understanding the fullness of inter-reflected components in the tunnel. So the situation remains that we are estimating the lighting design based on a perfect diffusion pattern to the wall - and in reality are installing lighting with a specula wall surface. There are pros and cons to all situations. The benefits or pros to having a semi-specula surface in the tunnel is demonstrated, where there is a clearly visible image of the tunnel around the corner for the driver, and this will assist in predictive driving.
Even though the essential point of the wall linings is to improve general ambient visibility and to provide a contrast background for seeing the other vehicles in silhouette, anecdotally the surface of the semi-specula material slowly evolves with continued grime build up, and with cleaning cycles and eventually loses some of the specularity. This however does not appear to have been tested in the field or lab. So once again an intuitive or educated guessing process is used in design. If 'R' tables are too narrow or road observer based to be of use, then perhaps the next best thing would be to have a lab test of wall lining materials produced in various ages of wear in order to allow for a more accurate modelling of the space. Alternatively, if the strictures of the (alpha = 1°? R-Table) don't allow for a proper study of inter-reflectance, then perhaps the software suppliers can provide revised programs which can take into account the dynamics of specula reflection.
One other perhaps scary consideration is the results of the 'Jacket Frith' report, in which the various carriageway surfaces across New Zealand were tested for luminance performance and were found to be in error.
Using the 3D lighting model of the Logan Tunnel in AGi32 software, the lighting was first calculated with full radiosity (gathering all the inter-reflected light), and then redone with radiosity switched off.
The results are a comparison of direct light only to full inter-reflected lighting. The results show a contribution from interreflected lighting of five to seven per cent. This is a significant value. Any improvement in the design and modelling process would therefore enable a leaner design with lower energy consumption.
As a comparison reference point, the following table provides the calculated portal luminance values for the Logan tunnel at the Eastern Busway project in Brisbane, which is geometrically similar in size to the terrace tunnel.
The L20 is the calculated portal luminance of the tunnel as seen by the driver at the safe stopping distance. The threshold value is the amount of compensating brightness in candela per meter squared that is then provided in the portal threshold zone by the lighting system. These values are provided as a way of gauging the measurements taken in the tunnel.
It can be seen that the apparent and transient brightnesses in the reflected lighting on the wall of the tunnel are in the same scale as the portal entrance brightness values, which are witnessed by the drivers' eyes every day. As another point of reference or scale table 3 and 4 extract from AS1158.5 (2007) provides values for typical luminance of various surfaces in daylight conditions.
Referring to the Jacket Frith report
The testing of 140 sites across New Zealand in 2009 proved that the assumed reflective characteristics of road surfaces are not accurate, and that in the cases studied, the reflectance factors (Q0 and S1) were significantly lower than required.
This means that if the same was true in Australia, and there is no reason to suspect that this is not the case, then we are effectively under-designing road lighting, and perpetrating an error in design that increases the danger of driving on the road.
Mitigating this concern for the Eastern Busway project is the fact that the pavement surfaces produced there were all very carefully specified and manufactured. The reflectivity of these modern carriageways is reasonably certain.
In combination with a complete lighting system audit, the wall linings reflective characteristics were combined with vehicle headlamps and were measured and filmed. The conclusion supported by the tabled measurements is that the brightest things in the tunnel are the vehicle headlamps.
Use of wall linings of a semi specula nature may actually improve the drivers performance in the tunnel, however there is no real understanding of how these surfaces perform in the design, and assumptions are still being used in the design process.
Results of the tunnel lighting audit have demonstrated accuracy between the design and the installation of less that five per cent margin.
This suggests that the design method used in the Eastern Busway project is intrinsically accurate and capable of handling the many significant variables in the process.