Revision of Threshold Luminance Levels in Tunnels Aiming to Minimize Energy Consumption at No Cost: Methodology and Case Studies †

: Because of the absence of lighting calculation tools at the initial stage of tunnel design, the lighting systems are usually over-dimensioned, leading to over illumination and increased energy consumption. For this reason, a ﬁne-tuning method for switching lighting stages according to the tra ﬃ c weighted L20 luminance is proposed at no additional cost. The method was applied in a real –case scenario, where L20 luminance of the access zone at eleven (11) existing tunnels was calculated. The tra ﬃ c weighted method of CR14380 was used in order to calculate the actual luminance levels for the entrance zone. The new transition zone, which decreases luminance curves, was produced and compared with the existing ones. Thus, a new switching control was proposed and programed for the Supervisory Control and Data Acquisition (SCADA) system of the tunnel. The signals of the corresponding eleven L20 meters for a period of eight days were used and the corresponding annual energy consumptions were calculated using the proposed switching program for each tunnel. The results were compared with a number of scenarios in which the existing lighting system was retroﬁtted with Lighting Emitting Diodes (LED) luminaires. In these scenarios, the new luminaire arrangement was based not only on the existing luminance demand value for the threshold zone, but also on the newly proposed one with two di ﬀ erent control techniques (continuous dimming and 10% step dimming). The ﬁne-tuning method for switching resulted in energy savings between 11% and 54% depending on the tunnel when the scenario of the existing installation at no extra cost was used. Energy savings, when LED luminaires were installed, varied between 57% (for the scenario with existing luminance demand value for the threshold zone and 10% step dimming) and 85% (for the scenario with the new calculated luminance demand and continuous dimming).

In tunnels, the use of the lighting system should guarantee a safe pass through it, not only during the night but in daytime as well. The drivers should be able to discern the presence of other vehicles and possible obstacles in the road [31,32]. The effect of the "black" hole at the entrance of the tunnel during the day must be avoided. Thus, higher luminance values are required in order to enhance the visual adaptation of the incoming drivers. The higher luminance values needed at the entrance zone are defined by standards. The required luminance values are dependent on the incident daylight on the surrounding surfaces adjacent to the entrance. As mentioned above, visual adaptation demands increased illuminance not only at the tunnel entrance but also for a considerably long distance inside the tunnel all the way up to the interior zone. Unlike buildings, where daylight minimizes lighting needs [33][34][35], in tunnels it results in an increased number of installed luminaires as well as in increased power consumption for each luminaire [31,32]. This design approach increases energy consumption during the day, since the threshold luminance (Lth) is directly linked to the access zone luminance, which is represented by L20. The latter is defined as the luminance of the tunnel entrance surrounding areas within a conical field of view of 20 • , within stopping distance of design speed. It is evident that the variation of daylight throughout the day affects L20 and, thus, the required luminance values inside the threshold zone. Consequently, the control of the active lighting stages of the tunnel is crucial for minimizing energy consumption during the day.
In an effort to minimize the initial costs, the decision-making process of designing a tunnel takes into consideration only the construction costs. The life cycle cost analysis, which also takes into account the maintenance and lighting operational costs, is ignored. Moretti, Cantisani, and Di Mascio compared the expected costs for pavement construction, maintenance, and road lighting of a highway tunnel in Rome [36]. A lighting system was tested inside a tunnel with a concrete pavement and the energy consumption was 29% lower than in a tunnel with an asphalt pavement [36]. Furthermore, Moretti, Cantisani, Mascio, and Caro investigated the life cycle costs of two different road tunnel pavements and their corresponding lighting systems [37]. López, Grindlay, and Peña-García [38] suggested a sustainability vector for the initial design of a tunnel. The vector presents the degree of sustainability and highlights the necessity for corrective actions when necessary, combining three parameters (a) energy consumption, (b) landscape integration, and (c) construction cost. An installation of semi-transparent tension structures at the entrance portal can lead to significant energy savings [39]. Another way to reduce the luminance requirements and, thus, the energy consumption is the forestation of the surroundings of the portal of tunnels. Energy consumption can be reduced by up to 50%, as long as the specific species that will be used are permitted by the climatic and hydrological conditions of the zone where the tunnel is [40]. Moreover, García-Trenas, J.C. López, and A. Peña-García [41] analyzed how changes in the vegetation at the area surrounding the tunnel entrance can contribute to energy savings for a lighting installation in an Alpine environment. The required illumination levels can also decrease by using structural measures at the approaching zones or at the tunnel mouth [42]. A pre-tunnel lighting may ensure adequate, progressive, physiological adaptation of the user's eyes when approaching the entrance of the tunnel, and contain the overall costs of the artificial lighting system throughout its service life [43].
As energy consumption has become a crucial factor for tunnel design, a number of control systems based on daylight compensation have been investigated for installation in the tunnel entrances [44][45][46][47]. Gil-Martín, Peña-García, Jiménez, and Hernández-Montes used a scale model in order to test a system with light-pipes [48]. In a follow-up study, the aforementioned authors used a heliostat to guide sunlight into the light-pipes. The results showed a remarkable improvement in the efficacy of light-pipes, in electrical energy consumption and in the number of luminaries used [49]. A semi-transparent tension structure of a polyester set was used just before the entrance to the tunnel. Hence, the threshold zone was extended towards the outside of the tunnel, in order to minimize lighting demands through the utilization of sunlight [50]. In addition, Peña-García and Gomez-Lorente investigated the installation of solar panels in the areas surrounding tunnel portals [51] while Peña-García and Gil-Martín investigated the use of pergolas for energy savings [52]. Unfortunately, the requirement of road surface uniformity was not fulfilled because the lighting levels were extremely low in the shadowed zones as compared to the sunlit zones [52]. Using a diffuser material in the spaces between the pergola beams improved the homogeneity of sunlight and, thus, energy savings [53]. Salam and Mezher [54] calculated 50% saving in the lighting electrical load with the use of shading structures in existing tunnels. However, energy savings must not be the only parameter to consider during lighting design. In very long tunnels, people's safety may depend on their reactions to the claustrophobic conditions of tunnels, which could range from stress and anxiety to distraction or fear [55].
Except for the initial design of a tunnel and the methods for reducing energy savings, the renewal, measurements [56,57], and maintenance procedures are also crucial elements for the operation of an existing tunnel. These require, among others, the redetermination of the L20 luminance. This can be realized by taking photographs of the entrance of a tunnel from a fixed point at the center of the motorway exactly from the stopping distance, a method that would require stopping or diverting the traffic completely. Lopez and Pena-Garcia [58] proposed a methodology that uses vehicle-based images and trigonometric considerations and does not affect the traffic. Shuguang [59] presented a tunnel lighting optimal control model taking into account both traffic safety and energy-saving issues. His control model takes the demand on brightness, the total average brightness, and the minimum dimming ratio of the luminaires as parameters. The role of dimming [60,61] and light control [62][63][64] is significant for the selection of luminaires [65][66][67]. Pachamanov and Pachamanova [68] presented models for the optimization of the lighting distribution of luminaries for tunnels, which allows the incorporation of the characteristics of the reflective properties of the surface of the road in order to obtain energy-efficient light distributions. Salata et al. [69] optimized energy savings considering the lighting system (High Pressure Sodium lamps (HPS) or Lighting Emitting Diodes (LED)) and the type of asphalt (traditional or special asphalt). Furthermore, Salata et al. [70] investigated whether it is possible to minimize energy demands through the usage of an automatic new control system regulating the luminous fluxes of artificial sources with respect to the variation of daylight, which is characteristic of the outdoor environment.
In general, the reduction of tunnel lighting consumption can be realized through proper optimization of the pavement or by retrofitting the lighting system with cost effective LED luminaires. However, energy savings can also be achieved a) with proper control of a tunnel's lighting system, since this is quite commonly organized in a number of active stages, and b) by reevaluating the corresponding luminance values in the threshold zone (Lth) using the L20 values. The scope of this paper is to propose a control strategy according to the new luminance level requirements based on the traffic weighted L20 method (CR14380, [31]) in existing tunnels. The early results of this method were presented in the 2019 IEEE (Institute of Electrical and Electronics Engineers) International Conference on Environment and Electrical Engineering and the 2019 IEEE Industrial and Commercial Power Systems Europe (EEEIC/I&CPS Europe), in Genova, Italy, June 11th-14th 2019 [71]. In this paper, the Over Lit Triggering Percentage (OLTP) of various circuits was defined. Eleven tunnels were examined and considerable amounts of energy savings and CO 2 emission reductions were achieved.

Materials and Methods
Most of the tunnels were constructed before the establishment of the European standards and prior to the advent of LEDs. Comparing the luminance requirements of the existing tunnels with the new weighted L20 method, over illumination is evident. Nowadays, that LED technology is mature, retrofitting the existing lighting system has become of particular importance. This paper presents a method that can take place before the renovation of the tunnels. This method results in significant energy savings and lower CO 2 emissions at no additional cost. In short, the actions involved are the following: (a) a new calculation method of the stopping distance, (b) the estimation of the corresponding L20 value, and (c) the programming of the Supervisory Control and Data Acquisition (SCADA) system of the tunnel. The proposed method is presented more analytically in Figure 1. The traffic weighted L20 method was used. The main influencing factors for the examined cases were medium for traffic flow (500-1500 vehicles per hour per lane for one-way traffic), and motorized traffic only. According to these two factors, the tunnel class was defined and then, as a next step, the new corresponding threshold zone luminance (Lth') was calculated. The new Lth' value, was used for each tunnel, in order to define the new triggering points of the lighting stages using the SCADA control system. Because of the new lower Lth' values in comparison to the initial Lth values, the triggering points correspond to lower lighting levels and thus to lower amounts of associated energy consumption.
Energies 2020, 13, x FOR PEER REVIEW 4 of 23 method that can take place before the renovation of the tunnels. This method results in significant energy savings and lower CO2 emissions at no additional cost. In short, the actions involved are the following: (a) a new calculation method of the stopping distance, (b) the estimation of the corresponding L20 value, and (c) the programming of the Supervisory Control and Data Acquisition (SCADA) system of the tunnel. The proposed method is presented more analytically in Figure 1. The traffic weighted L20 method was used. The main influencing factors for the examined cases were medium for traffic flow (500-1500 vehicles per hour per lane for one-way traffic), and motorized traffic only. According to these two factors, the tunnel class was defined and then, as a next step, the new corresponding threshold zone luminance (Lth') was calculated. The new Lth' value, was used for each tunnel, in order to define the new triggering points of the lighting stages using the SCADA control system. Because of the new lower Lth' values in comparison to the initial Lth values, the triggering points correspond to lower lighting levels and thus to lower amounts of associated energy consumption.
The following scenarios were examined: • Scenario A: Switching control with existing Lth values.
• Scenario B: Switching control with newly calculated Lth' values.

Existing Lighting Infrastructure
The examined case studies are part of the national motorway Patra-ATHens-Efzoni, (PATHE) in Greece. This national motorway PATHE is one of the 2 motorways connecting Athens to the rest of Greece with an approximate length of 172.5 km. The motorway starts at Metamorfossi (an area in the Prefecture of Attika) and ends at Skarfia, (Prefecture of Fthiotida), after Kamena Vourla. It is a modern motorway using international standards. This PATHE section crosses two regions and three counties and its technical features include among others, 8 bridges, 30 interchanges, 11 tunnels ( Figure 2), 1 short tunnel, and 84 underpasses and overpasses. The following scenarios were examined: •

Existing Lighting Infrastructure
The examined case studies are part of the national motorway Patra-ATHens-Efzoni, (PATHE) in Greece. This national motorway PATHE is one of the 2 motorways connecting Athens to the rest of Greece with an approximate length of 172.5 km. The motorway starts at Metamorfossi (an area in the Prefecture of Attika) and ends at Skarfia, (Prefecture of Fthiotida), after Kamena Vourla. It is a modern motorway using international standards. This PATHE section crosses two regions and three counties and its technical features include among others, 8 bridges, 30 interchanges, 11 tunnels ( Figure 2), 1 short tunnel, and 84 underpasses and overpasses. Most of the lighting fixtures for the main road are installed at the intersections. The lighting installation of the road includes 6565 luminaires (main road, intersections, toll areas, parking sites) while the lighting fixtures for the tunnels, without taking into account the underpasses, are 5344 along 8.5 km. The basic characteristics of the eleven tunnels, which were examined together with their lighting system installed power including the power losses from the electromagnetic ballasts, are presented in Tables 1-3. The road has 2 lanes with a total lane width of 7.5 m and a speed limit of 80 km/h in the tunnels.  Most of the lighting fixtures for the main road are installed at the intersections. The lighting installation of the road includes 6565 luminaires (main road, intersections, toll areas, parking sites) while the lighting fixtures for the tunnels, without taking into account the underpasses, are 5344 along 8.5 km. The basic characteristics of the eleven tunnels, which were examined together with their lighting system installed power including the power losses from the electromagnetic ballasts, are presented in Tables 1-3. The road has 2 lanes with a total lane width of 7.5 m and a speed limit of 80 km/h in the tunnels.  Table 3. Installed power and the corresponding energy indicators for the interior zone and nighttime stage of the examined tunnels (the luminaires of the interior zone were installed along the full length of the tunnel).

Luminance Calculations (L20)
As the proposed methodology compares the existing threshold luminance Lth, with the new Lth', the luminance L20 at the access zone has to be calculated. The L20 value can be obtained either from estimation [31] or by using a combination of a photo of the tunnel entrance and corresponding calculations according to the standards [31]. More specifically, the photograph should be taken from a point at a distance equal to the stopping distance from the tunnel portal in the middle of the specific Energies 2020, 13, 1707 7 of 23 motorway or traffic lane with the road closed off [58]. The evaluation of L20 was obtained using the photographs, one for each of the eleven tunnels, presented in Figure 3 with the aid of the equation (1):   Where: L20 is the access zone luminance, L C is the sky luminance, γ= (%) of sky, L R is the road luminance, ρ = (%) of road, L E is the surrounding luminance and ε = (%) of surroundings.
The parameters used for the calculation of L20 are presented in Table 4 while Table 5 shows the luminance requirement for the new threshold zone Lth'. The k factor was calculated from Table 6 using interpolation. All the tunnels were classified as class 2 for motorized traffic only and medium  Table 5. The corresponding Lth' values varied from 26% to 60%, which was lower than the corresponding existing one Lth, a fact meaning that the lighting systems are over-dimensioned for all the tunnels.

Defining Switching Control
While the incident daylight at the tunnel portal is not stable during the day, there is a need for a switching control based on the daylight levels. Depending on the control of the lighting stage, the energy consumption for the same tunnel can be considerably modified.

Existing Switching (Scenario A)
The existing switching program of the examined tunnel is shown in Table 7. It is based on the signal generated by the L20 luminance meter, which activates the luminaires through the SCADA system. For the existing control switching (Scenario A, Lth) when daylight increases and the L20 value sent to SCADA is larger than the values in Table 7, Stage 1 (S1) is switched on (full light output). If the L20 signal is lower than the corresponding values, then the next lighting stage 2 (S2) is engaged. For example, for Tunnel 1a (Table 7), when L20 value is larger than 1530 cd/m 2 , Stage 1 is switched on (full light output, 305 luminaires, 123.8 kW). If the L20 signal is less than 1530 cd/m 2 , then the corresponding Stage 2 (S2) is engaged (182 luminaires, 67.3 kW).

Proposed Switching (Scenario B)
For the proposed switching, the threshold luminance (Lth') is different than the threshold luminance Lth of the initial design. The new CIE (Commission Internationale de l'Eclairage or International Commission on Illumination) curves regarding the new Lth' values are presented in Figure 4, and it is evident that the existing stage 2 can satisfy the maximum lighting needs in many cases. This results in greater energy savings, since stage 1 will be set permanently to inactive (Tunnels 2a and 2b, Table 8). In addition, for the next lighting level with the new lower than the existing configuration luminance requirements, S4 is engaged instead of S3. In order to determine the new triggering of the lighting system and the associated L20, a new parameter called Over Lit Triggering Percentage of various circuits (OLTP) is proposed. This percentage is defined as follows: Where: Lth (SN) is the Lth of the initial design of the existing tunnel for SN light stage, Lth'(S1) is the luminance requirement for the new threshold zone for S1 stage and SN is the corresponding lighting stage (S1 for N = 1, S2 for N = 2, etc.). Thereafter OLTP SN represents (in percentage) the proposed triggering of each of the existing lighting circuit for SN light stage. Values above 100% meaning that the corresponding switching stage is inactive. This percentage is necessary for the specification of the new triggering of the existed lighting circuits based on the new lower lighting requirements. It is evident that the proposed triggering will now depend on the Lth' and since the existing lighting achieves specific lighting levels due to the existing lighting circuits, their triggering has to be redefined. Hence, the proposed switching control, Scenario B (Table 8) enables all stages at higher values of L20 when compared to the existing configuration. This means that the use of lighting control stages with fewer luminaires and less installed power instead of the existing ones, for the same incident daylight at the portal of the tunnel, can lead to a larger amount of energy savings and lower amounts of CO 2 emissions.   In addition, four scenarios were examined where the existing lighting system for each tunnel was retrofitted with LED luminaires (C: 10% step control dimming and D: continuous dimming) using both the existing lighting requirements Lth and the new calculated Lth' (C1 and D1 for Lth and C2 and D2 for Lth'). Tables 9 and 10 present the number of LED luminaires needed and the corresponding installed power for all scenarios. The data were extracted with the use of the Relux Tunnel light simulation tool [72]. Furthermore, the power density indicator for the entrance zone of the tunnel was calculated as the ratio of its installed power to the area that is defined by the length of the entrance zone of each tunnel and the width of both lanes of the road (7.5 m).

Discussion and Conclusions
The total annual electrical and primary energy consumptions of the existing installation (11 tunnels) were 2715 MWh and 7874 MWh correspondingly (Figure 7). The energy consumption from the examined period (Table 12) was normalized for a year, while the Primary Energy Numeric Indicator (kWh p = 2.9 × kWh e [75]) from Greece was used in order to convert the electrical to primary energy.

Discussion and Conclusions
The total annual electrical and primary energy consumptions of the existing installation (11 tunnels) were 2715 MWh and 7874 MWh correspondingly (Figure 7). The energy consumption from the examined period (Table 12) was normalized for a year, while the Primary Energy Numeric Indicator (kWh p = 2.9 × kWh e [75]) from Greece was used in order to convert the electrical to primary energy. Using the proposed methodology, the primary energy consumption can be reduced to 5396 MWh (Figure 7) while the corresponding annual CO2 emission reduction is 904.6 tn. Thus, the energy savings can reach a figure of 31% using the new switching control strategy, according to the calculated Lth'. If combined with the retrofitting of existing luminaires with LED technology, the Using the proposed methodology, the primary energy consumption can be reduced to 5396 MWh (Figure 7) while the corresponding annual CO 2 emission reduction is 904.6 tn. Thus, the energy savings can reach a figure of 31% using the new switching control strategy, according to the calculated Lth'. If combined with the retrofitting of existing luminaires with LED technology, the energy savings can increase and reach 62%. The corresponding difference in energy savings between Scenario B and C2 (31%) could not be viable as the cost of a LED tunnel luminaire, including the labor work for the new installation, is still high. However, using the new Lth' values (Scenario D, Figure 8) energy savings are 23%, while by retrofitting the existing luminaires with LEDs, additional energy savings of up to 62% can be achieved. Figure 9 presents the annual energy costs together with the initial costs of the LED luminaires versus the annual primary energy consumption per square meter of the entrance zone of the examined tunnels. Each dot represents a case examined while the cases are grouped (different color) according to the scenarios. A price of 1200 euros per luminaire was taken as the initial cost of the LED luminaire, the cost of energy was calculated at 0.15 euros per kWh while the Primary Energy Factor was considered equal to 2.9 (Greece, [75]). The lower primary energy consumption and the cost are, (lower left part of the diagram in Figure 9) the more the beneficial is the action of the examined scenario. It seems that Scenario B (orange dots), represents the most beneficial one. energy savings can increase and reach 62%. The corresponding difference in energy savings between Scenario B and C2 (31%) could not be viable as the cost of a LED tunnel luminaire, including the labor work for the new installation, is still high. However, using the new Lth' values (Scenario D, Figure 8) energy savings are 23%, while by retrofitting the existing luminaires with LEDs, additional energy savings of up to 62% can be achieved. Figure 9 presents the annual energy costs together with the initial costs of the LED luminaires versus the annual primary energy consumption per square meter of the entrance zone of the examined tunnels. Each dot represents a case examined while the cases are grouped (different color) according to the scenarios. A price of 1200 euros per luminaire was taken as the initial cost of the LED luminaire, the cost of energy was calculated at 0.15 euros per kWh while the Primary Energy Factor was considered equal to 2.9 (Greece, [75]). The lower primary energy consumption and the cost are, (lower left part of the diagram in Figure 9) the more the beneficial is the action of the examined scenario. It seems that Scenario B (orange dots), represents the most beneficial one.  energy savings can increase and reach 62%. The corresponding difference in energy savings between Scenario B and C2 (31%) could not be viable as the cost of a LED tunnel luminaire, including the labor work for the new installation, is still high. However, using the new Lth' values (Scenario D, Figure 8) energy savings are 23%, while by retrofitting the existing luminaires with LEDs, additional energy savings of up to 62% can be achieved. Figure 9 presents the annual energy costs together with the initial costs of the LED luminaires versus the annual primary energy consumption per square meter of the entrance zone of the examined tunnels. Each dot represents a case examined while the cases are grouped (different color) according to the scenarios. A price of 1200 euros per luminaire was taken as the initial cost of the LED luminaire, the cost of energy was calculated at 0.15 euros per kWh while the Primary Energy Factor was considered equal to 2.9 (Greece, [75]). The lower primary energy consumption and the cost are, (lower left part of the diagram in Figure 9) the more the beneficial is the action of the examined scenario. It seems that Scenario B (orange dots), represents the most beneficial one.  From the aforementioned results, it is evident that over-illumination and oldness of the existing tunnels result in increased and unwanted energy consumption especially in the daytime. A technical committee from the International Commission of Illumination (CIE 4-53 Tunnel Lighting Evolution) for tunnel upgrading has been formed [76] in an effort to minimize energy consumption. In addition, many lighting experts propose various actions such as using daylight control systems and different types of pavement in a similar attempt to reduce energy consumption. The proposed methodology, although simple, is not fully integrated into current energy saving policies. The revision of the Lth values should be a step taken prior to the action of replacing the existing lighting system with LED luminaires. The paper proposes a switching control strategy, which can be a useful tool for lighting designers, road authorities, and lighting experts. This switching control combined with the traffic weighted L20 method as described in CR14380 (Scenario B), can result in significant energy savings at no extra cost. Calculations were performed and energy savings was, on average, 31% varying from 11% to 54% depending on the tunnel. By replacing existing luminaires with LEDs with the existing threshold luminance Lth (Scenarios C1 and C2), energy savings can reach 62% while with the new threshold luminance Lth' (Scenarios D1 and D2), the corresponding values can reach a figure of 85%. Even with the replacement of the existing lighting systems with LEDs, the effect of determining the new threshold luminance Lth' can result in 23% more energy savings (comparing C and D scenarios). Thus, the proposed methodology is suitable for being considered in retrofit actions with LED luminaires. However, this increase in energy savings is accompanied by the additional cost of the 2018 new LED luminaires (scenario C) or of the 1227 luminaires for scenario D, together with a new Supervisory Control and Data Acquisition system, as well as extra installation costs, such as wiring and the corresponding labor cost. In many cases, the extra cost for a new lighting installation compared with the no-cost switching strategy could make the renewal of the installation unsustainable if the cost of the luminaires is high. In addition, the proposed methodology a) is easy to apply with immediate results, b) the calculation of the new L20 values could be necessary in order to evaluate the initial design due to safety reasons, and c) no tender is required for its realization.
For future research, the proposed method could also be combined with traffic detection sensors, as the traffic volume can determine the tunnel class and thus the necessary lighting needs. For example, a tunnel class 3 with high traffic flow, could result in class 2 with medium traffic flow for a time period. As factor k will be defined by lower values, the new Lth' values should determine a new control switching. In general, the energy savings using traffic intensity detector parameters could end up to 50% [77][78][79][80][81][82][83]. Furthermore, frequent luminance measurements could enhance the energy savings. Monitoring the real situation of the lighting system, the switching system of SCADA can be fine-tuned, taking into account the lumen maintenance control strategy technique and the actual lighting levels. For this procedure, there are several novel methods for road luminance measurements, where luminance measurements are combined into mobile mapping systems and three-dimensional (3D) measuring [84][85][86][87][88][89].

Factor k
Threshold zone luminance ratio (k) at a point: the ratio between the threshold zone luminance Lth and the access zone luminance L20. Typical values are given by [31] Lth Threshold zone luminance, the average road surface luminance of a transverse strip at a given location in the threshold zone of the tunnel (as a function of the measurement grid).
Lth (S1) Lth (SN) is the Lth of the initial design of the existing tunnel for SN light stage, S N is the corresponding lighting stage (S1, S2, etc.) Lth' (S N ) Lth' (S1) is the luminance requirement for the new threshold zone for S1 stage L20 Average luminance contained in a conical field of view, subtending an angle of 20 • with the apex at the position of the eye of an approaching driver and aimed at the left of the tunnel mouth. L20 is assessed from a point at a distance equal to the stopping distance from the tunnel portal at the middle of the relevant carriage-way or traffic lane. L20 formula: Lc Typical values of sky luminance depending the driving direction given by [31] L20 formula: L R Typical values of road luminance depending the driving direction given by [31] L20 formula: L E Typical values of surrounding luminance depending the driving direction given by [31] L20 formula: γ Percentage of the area of the sky covering the area contributing to the L20 value at the tunnel entrance L20 formula: ρ Percentage of the area of the road covering the area contributing to the L20 value at the tunnel entrance L20 formula: ε Percentage of the area of the surrounding covering the area contributing to the L20 value at the tunnel entrance Over Lit Triggering Percentage (OLTP) Lth (SN)/Lth' (S1) Where Lth (SN) is the Lth of the initial design of the existing tunnel for SN light stage, Lth' (S1) is the luminance requirement for the new threshold zone for S1 stage and SN is the corresponding lighting stage (S1 for N=1, S2 for N=2, etc.). Thereafter OLTPR SN represents (in percentage) the new triggering of each of the existing lighting circuit for SN light stage. This percentage is necessary for the specification of the new triggering of the existed lighting circuits based on the new lower lighting requirements. It is evident that the proposed triggering will now depend on the Lth' and since the existing lighting achieves specific lighting levels due to the existing lighting circuits, their triggering has to be redefined.