Integrated Modelling and Experimental Verification of Energy Consumption and Performance of the Lighting Systems of Tunnels P. Varilone, P. Verde
F. Caporaso (°), E. Cesolini (*), S. Drusin (*)
Dipartimento di Ingegneria elettrica e dell’informazione Università di Cassino e del Lazio Meridionale
(°) ANAS, SpA, Compartimento per la Viabilità della Basilicata (*) ANAS SpA, Direzione Centrale Ricerca e Nuove Tecnologie
Cassino, Italy verde@unicas.it
Italy css.cesano@stradeanas.it
Abstract— This paper presents the results of the work conducted in cooperation with ANAS, which is in charge of the Italian road and highway network of national interest. The cooperation also is framed into the Italian Technical Committee on Strategic Theme 4 – Infrastructure for the specific topic of “Balancing of environmental and engineering aspects in management of road networks.” The paper presents an improved model, the EST (Energy Screening of Tunnels) model, to account for both the consumption of electricity and the performance of tunnel lighting systems. The models were developed so that each tunnel was represented on an adequate (x, y) plane that was defined by two indices. The main improvements of the model include the introduction of boundary values of the indices derived from International Standards and from resolutions of the Italian Authority of Electrical Energy and Gas. Keywords: performance
lighting
I.
systems,
energy
efficiency,
light
INTRODUCTION
It is a great challenge for the management of significant infrastructures, such as primary roads and highways, to reduce operating costs while ensuring adequate levels of safety and quality of service to users. One important contributor to the operating costs is the energy consumption of the electric systems associated with the roads, such as the air treatment and lighting systems inside tunnels, at roundabouts, and at exit zones. Among those, the lighting systems of tunnels (LSTs) are emblematic since they consume significant quantities of energy and are essential for the safety of the drivers who use the tunnels. Energy savings of the LSTs is dealt with in the specialised literature [1, 3] with different proposals mainly referred to the design of new systems to be integrated into existing LSTs or of new LSTs. In [1], the installation of advanced controlled systems is proposed to take advantage of periods of low traffic volume when the LSTs can furnish reduced lighting levels and, consequently, use less energy. In [2], appropriate transparent tension structures are proposed to extend the threshold zone of the tunnel and to make it possible to use sunlight to light the tunnels to the extent possible, thereby reducing energy consumption at the entrance zone of the tunnel. In [3], a tool
based on a genetic algorithm was presented for automating the design of new LSTs; the tool allows optimization processes to be applied in all the tunnel zones to ensure adequate visual perception and safe driving. This paper presents the results of an ongoing study of the energy efficiency of LSTs that is being conducted in cooperation with ANAS, the company that is in charge of Italian roads and highway networks of national interest. The study also is integrated with the work that is being conducted by the Italian Technical Committee of the World Road Association, i.e., the Association Internationale Permanente des Congrès de la Route/Permanent International Association of Road Congress (AIPCR/PIARC). This Committee is a component of the Working Group on Strategic Theme 4– Infrastructure for the specific topic of “Balancing of environmental and engineering aspects in the management of road networks.” The principal objectives of the study are to investigate the electric energy consumption of existing LSTs and analyse the lighting performance of installed luminaries. Based on the results the analysis, possible interventions will be proposed that reduce the consumption of electric energy and improve the lighting performance. To that end, a model entitled the ‘Energy Screening of Tunnel’ (EST) model has been developed. The key idea of EST is to characterize simultaneously every tunnel in the Cartesian plane based on electricity consumption and performance, by introducing appropriate indices. The first version of EST model was presented in [4], and this version showed the results obtained for three different inservice tunnels. In this paper, we have improved the model by including the proper limits on the lighting levels in tunnels. In the following, the main features of an LST are described with respect to the actual standards, and, then, the EST model is presented to show the advancement of the model. Finally, some actual cases are provided to demonstrate the value of the EST model as a tool for deciding possible interventions for improving the service conditions of LSTs. Finally, we present the ongoing experimental activities that are being performed by ANAS’s advanced system “Tiresia.”
II.
MAIN FEATURES OF LIGTHING SYSTEMS OF TUNNELS AS THEY RELATE TO THE ACTUAL STANDARDS
To ensure driving safety, LSTs must provide good visibility both day and night, with sufficient vertical illumination to allow drivers to detect the presence of obstacles and to ensure the absence of direct or reflected shadows. They must also provide sufficient lighting to ensure that drivers can perceive the change in driving conditions as they approach the entrance of the tunnel without developing a sense of uncertainty or, even worse, fear [5]. The actual Standards [6, 7] indicate that the luminance profile must provide the aforementioned features. The luminance profile is the required values of the luminance along the longitudinal axis of the tunnel versus the travel time along the tunnel at the reference speed; the reference speed is equal to the speed limit in the tunnel, and it is provided by the operator of the road that the tunnel serves. The values of luminance are expressed in percentages of the luminance at the entrance to the tunnel. (Fig. 1 also reports an additional x-axis that indicates the distance travelled at the speed of 60 km/h.) Fig.1 shows that the luminance at the entrance of the tunnel is dependent on the stopping distance, i.e., ‘da,’ which, in turn, depends mainly on the speed limit, the road gradient, and the hydro-hygrometric state of the road’s surface; Annex A of the standard UNI 11095 provides appropriate tables that can be used to calculate this distance. Also, the luminance at the entrance of the tunnel must be less than a prescribed maximum value to avoid glare.
Fig. 1: Plot of the required luminance along the longitudinal axis of the tunnel
Fig. 2: Plot of the minimum (curve 2) and maximum (curve 1) luminance along the tunnel
The UNI 11095 also addresses the energy savings of an LST. It clearly states that after the designer of an LST has ensured that all safety requirements have been implemented, he or she must verify, in a second phase, that the energy consumption is not excessive. With this goal in mind, the UNI 11095 establishes the luminance limits for each zone of the tunnel. Fig. 2 shows the maximum values of the luminance along the tunnel superimposed on the minimum luminance values required. III.
EST MODEL: ENERGY SAVING OF TUNNEL MODEL
The key idea of the EST model is to characterize simultaneously every tunnel in the Cartesian plane from the standpoints of electricity consumption and performance by introducing appropriate indices. It is important to understand that the EST model is not a tool to be used in the design of LSTs; rather it is a tool that is used to ascertain the lighting performance and electric energy consumption of LSTs. The consumption of electric power is quantified through an index of annual consumption per kilometre of tunnel, i.e., the Electric Power Consumption Index (ECPI); the ECPI allows the comparison of the energy consumptions of LSTs that have different lengths. The EPCI can be computed in different way, based on the data that are available, including measurements that are made on-site, data from billings, and estimations derived from installed luminaries. The lighting performance is quantified through an index of the average lighting performance for each square meter of the tunnel, i.e., the Lighting Performance Index (LPI). In accordance with the Standards, LPI must be linked to the luminance levels along the tunnel. The only way to obtain the values of the luminance levels inside the tunnel is to measure them by using a luminance meter. If such measurements are not available, as is often the case, LPI can be related to the illuminance levels, which, in turn, must be linked to the luminance values. In the absence of measured luminance levels, the LPI uses the values of illuminance that are properly correspondent to the values of luminance. This correspondence is derived from a resolution of the Italian Authority for Electricity, Gas and Water supply system (AEEGSI) on the subject of energy efficiency for LSTs [8]. In this resolution, in fact, the AEEGSI defines a baseline case of an LST for different types of roads (e.g., different luminaire layouts, dual lanes and single lane) and recognizes economic incentives only for efficiency interventions that reduce the consumption of electric power below that of the baseline LST. With reference to this aim, [8] also gives an important indication concerning the correspondence between the levels of luminance and the illuminance of the baseline LST1. In particular, it stated the illuminance levels that guarantee the assigned levels of luminance for given types and conditions of tunnels. First, this correspondence was verified by means of simulation, and, then, it was used for the LPI. 1
The resolution of AEEGSI [8] refers to the tunnels of highways and suburban primary roads.
Fig. 3: Cartesian plan (electricity consumption, performance) of three generic LSTs
With the preceding choices, every LST can be represented by means of the two indices, i.e., EPCI and LPI, as a point on the plane {electricity consumption, performance}, which allows comparisons among the LSTs. For EPCI and LPI, two reference values have been introduced, i.e., EPCI* and LPI*, respectively, that correspond to 1) the baseline case of the LST related to the type of road where the tunnel is located and 2) to the typology of the LST with respect to the same resolution of AEEGSI [8] mentioned previously. For EPCI, it is possible to quantify the electric power consumption of the LST as the difference between EPCI and EPCI* (ΔEPCI). If ΔEPCI is greater than zero, the consumption must be reduced; otherwise, the consumption is acceptable. In a similar way, the corresponding index of the lighting performance is the difference between LPI and LPI* (ΔLPI). If ΔLPI is less than zero, the lighting performance must be improved. If ΔLPI is not less than zero, the lighting performance is acceptable. Fig. 3 shows the results of EST on the Cartesian plane in which three generic LSTs are represented as points. The Cartesian coordinates of each point are the value of ΔEPCI for the X-axis and the value of ΔLPI for the Y-axis; Fig. 3 also shows four zones, each of which is characterised by different values of lighting performance and different consumptions of electrical power, but only one zone defines good performance for both of the aspects (second quadrant of the Cartesian plane). The EST model, the results of which are represented on the Cartesian plane of Fig. 3, was improved to account for the maximum acceptable value of the lighting levels inside a tunnel. As shown in Fig. 2, the Standards [5] introduce the curve of maximum luminance along the tunnel to avoid excessive energy consumption. The values of the curve (1) of Fig. 2 are expressed as percentages of the luminance at the entrance of the tunnel. Three main steps were needed. First, the tunnel in the study and the tunnel in the baseline case were assumed to operate in the worst conditions in order to calculate the maximum acceptable value of luminance at the entrance.
Fig. 4: Cartesian plane {electricity consumption, performance} of LSTs with the limit on ΔLPI (red line).
Second, using the correspondence derived from [8], the values of maximum luminance of curve (1) of Fig. 2 were matched with the corresponding values of maximum illuminance along the tunnel. The maximum value of ΔLPI was computed as the difference between LPIMAX and LPI*MAX, which are the average of the maximum illuminances along the tunnel in the study and along the baseline tunnel, respectively. With the introduction of the limit on ΔLPI, i.e., ΔLPI,MAX, the Cartesian plane was divided in six sub-sectors, as shown in Fig. 4; only the sector n.3 represents the region of good service for both energy consumption and lighting performance. First, the EST model was used to characterize the actual service of an LST in the plane of Fig. 4. If the position of the considered LST was not in the good service zone (sector 3 of Fig. 4), some interventions can be planned. IV.
ACTUAL CASES OF TUNNELS IN SERVICE
The application of the EST model allowed the characterization of several LSTs that were actually managed by ANAS in the Basilicata region. In the following, the analysis of five LSTs is presented as an example of the results that can be obtained; the LSTs were installed in the tunnels with the characteristics specified below: • tunnel “A”: length of 446 m on a two-way traffic road with a dual lane; • tunnel “B”: length of 334 m on a two-way traffic road with a dual lane; • two tunnels “C”: (right and left) with each having a length of about 650 m, on two one-way traffic roads with a dual lane. Fig. 5 shows the characterization of the actual operating conditions of the LSTs derived from the electric energy bills. It is evident that: − the LSTs of the two tunnels of “C” have very similar characteristics; − all of the LSTs had unacceptable lighting performances; − all of the LST had reduced electric energy consumption. The characteristics of Fig. 5, which were determined by the EST model, refer to the real operating conditions that were strongly influenced by maintenance operations on the roads where the tunnels were located. To determine what the rated operating conditions would be, i.e., in the absence of any
Fig. 5: Cartesian plan {electricity consumption, performance} of the LSTs that were analysed for real operating conditions
maintenance, the EST model was applied to the same LSTs using the data recorded by the management. This implies that the electric energy consumption and the lighting performance of the LSTs were estimated by the EST model from the rated data rather than being derived from the bills; the results are shown in Fig. 6. It is very interesting to note that the operating conditions that were estimated using the rated data resulted in unacceptable electric energy consumption. To determine whether the lighting performances were within the maximum acceptable limits, the corresponding baseline case for each LST was defined and compared with ΔLPI,MAX. Then, some interventions were planned to improve the lighting performance and reduce the electric energy consumption. The possible interventions for each LST were: a) installing flux regulators on the existing High Pressure Sodium (HPS) lamp luminaires ; b) substituting higher efficiency HPS lamps (New HPS lamps) for the existing luminaires; c) installing new HPS lamps with flux regulators; d) substituting LEDs for the existing luminaires .
Fig. 7: Cartesian plane {electricity consumption, performance} of the LST of the tunnel “A”
Fig. 8: Cartesian plane {electricity consumption, performance} of the LST of tunnel ”B”
Figs.7, 8, and 9 show the results of the EST model applied to the LSTs of tunnels “A”, “B”, and “C right”. In the same figures, both the results of the actual conditions (indicated in the figures as “Actual condition”) and those derived from the recorded data by the management (indicated in the figures as Fig. 9: Cartesian plane {electricity consumption, performance} of the LST of the tunnel “C right”
Fig.6: Cartesian plan {electricity consumption, performance} of a the analysed LSTs in the estimated operating conditions
“Design condition”) are shown. From these figures, it is evident that: − for the design condition, the LSTs of tunnels “A” and “B” provide lighting performance that is below the maximum acceptable limit, but the LSTs of tunnel “C right” exceeded the maximum acceptable limit; − installing flux regulators on the existing luminaires does not provide significant improvements in any LST;
− replacing existing luminaires with higher efficiency HPS lamps and installing new HPS lamps with flux regulators and substituting LEDs for the existing luminaires result in all of the LSTs in sector 3 having good performance; − substituting LEDs for the existing luminaires gives the maximum savings of electric energy consumption for all of the LSTs. To choose the best interventions, an economic comparison of the previous alternatives is needed; several financial methods can be used for this purpose, such as net present value, payback time (PBT), break-even analysis, and costbenefit analysis. To give a quantitative example, we estimated the PBT. Assuming as a reference the intervention of New HPS lamps, the PBTs of the other interventions as percentages of the PBT of the New HPS lamps are reported in Table I. The financial analysis for LEDs was conducted by assuming two possible scenarios, i.e., 1) assuming that the price of the LEDs was constant and 2) assuming that the price of the LEDs decreases.
Fig. 10: Tiresia system on the road
Table II: Measured values of illuminance Illuminance Tunnels A lane 1 A lane 2 B lane 1 B lane 2 C right C left
Table I: Pay Back Time (PBT) of possible interventions PBT [%] Intervention Flux regulator New HPS + Flux regulator New HPS LED at constant price LED at decreasing price
V.
Tunnel A 41 82 100 88 88
Tunnel B 35 77 100 97 88
Tunnel C right 48 88 100 96 96
EXPERIMENTAL ACTIVITIES
The high efficiency equipment named “Tiresia” was developed in collaboration with the National Electro-technical Institute "Galileo Ferraris" INRIM (Istituto Nazionale di Ricerca Metrologica) and the Experimental Centre Road of ANAS. Tiresia allows the photometric characterization of the lighting in the tunnels and, in general, of all of the street lighting in accordance with actual Standards [6, 9, 10]. The system of measurements and sensors is mounted in a vehicle (Fig. 10) and, in particular, four photocells arranged in pairs were mounted in the front and rear of the vehicle, allowing the measurement of illuminance, free from any shielding caused by the vehicle itself. Tiresias also is capable of measuring luminance by means of a camera equipped with a charge coupled device (CCD) sensor that has a resolution of 512 x 512 pixels that can operate as a multi-directional luminance meter. A digital panoramic camera can record images during the measurement activity, allowing accurate interpretation of the results in the post-processing phase; an odometer was used to measure distances. The odometer and the photocells must be calibrated properly before each measurement campaign. We performed the first measurement campaign on tunnels A, B, and C. The illuminance was measured during the month of July. As an example, Table II shows the average value and the standard deviation of the measured illuminance; for tunnels A and B, the values were referred to each lane.
Average [lm/m2] 64 61 55 52 73 70
VI.
Standard deviation [%] 58 60 50 42 53 44
CONCLUSIONS
In this paper, we presented advancements in the modelling of lighting systems used in tunnels that can simultaneously account for the consumption of electricity and the performance of the lighting systems. The model was developed so that each tunnel is represented on an adequate (X, Y) plane defined by two indices. The Cartesian plane was divided into six subsectors based on the introduction of a proper limit for the lighting performance; only the sector n.3 represents the region of good service for both the energy consumption and the lighting performance. The representation on such a plane gives an immediate indication of the service conditions of the lighting system of the tunnel, and it also allows comparisons with other systems. The paper shows the analyses that were conducted on some real cases of tunnels, starting from the existing operations of the tunnels and from the design conditions. Various possible investments were assumed for the new positioning in the plan of the lighting systems of the tunnels. The results allow comparisons of the alternatives for each tunnel and also guide the decision maker for sorting the choices with due consideration for the economic issue.
REFERENCES [1]
[2]
S. Nagai, S. Ishida, M. Shinji, K. Nakagawa: Energy-saving lighting system for road tunnel, Underground Space Use: Analysis of the Past and Lessons for the Future – Erdem & Solak, 2005 Taylor & Francis Group, London, ISBN 04 1537 452 9. A. Peña García, L.M. Gil-Martín, A. Espín Estrella and F. Aznar Dols: Energy saving in road tunnels by means of transparent tension structures, International Conference on Renewable Energies and Power Quality (ICREPQ’10) Granada (Spain), 23rd to the 25th of March 2010.
[3]
S. Leitao, E. J. Solteiro Pires, P. B. de Moura Oliveira: Road Tunnels Lighting using Genetic Algorithms, 15th International Conference on Intelligent System Applications to Power Systems, 2009. ISAP '09, 2009 [4] F. Caporaso, M. Montecuollo, P. Verde , P. Varilone, Actual Cases of Energy Savings and Performance Modeling of Tunnel Lighting Systems for Investment Decision Making, accepted for publication in the journal “Routes/Roads.” [5] J. Buraczynski, T. K. Li, C. Kwong, P. J. Lutkevich, Tunnel Lighting Systems, Fourth International Symposium on Tunnel Safety and Security, Frankfurt on Main, Germany, March 17-19, 2010. [6] Standard UNI 11095, “Illuminazione delle gallerie stradali,” Novembre 2011. [7] CIE88:2004, “Guide for the Lighting of Road Tunnels and Underpasses,” ISBN 3 901 906 31 2. [8] AEEG, Resolution EEN 4/11, 5 May 2011, http://www.autorita.energia.it/allegati/docs/11/004-11een.pdf. [9] Standard UNI EN 13201-3:2004, “Illuminazione stradale - Parte 3: Calcolo delle prestazioni,” September 2004. [10] Standard UNI 11248:2012, “Illuminazione stradale - Selezione delle categorie illuminotecniche,” October 2012.
ACKNOWLEDGMENTS The work was developed inside the Italian Technical Committee of the World Road Association (PIARC) on Strategic Theme 4 – Infrastructure for the specific topic “Balancing of environmental and engineering aspects in management of road networks.” The authors thank the members of the Committee for the interesting, stimulating, and beneficial discussions during this work. Also, the authors thank Dr. Alessandro Turchetta for his vital contributions in the numerical applications.