An analysis of the main elements is necessary to understand how each component affects the final energy consumption. These elements are divided as follows: lamps, ballast, street lamp globes, hours of operation, lighting level control devices and renewable energies. To be sure about their involvement in the final energy consumption, each component was studied individually. Then, after studying how each component influences the final energy, different criteria to save energy were established. According to Boyce [
19], there are four options to save energy: changes in technology, in patterns of use, standards and basis of design, but from our point of view, changes to the basis of design require a careful reconsideration of what such lighting is for and how it might best be achieved, so each element was studied excluding the fourth option.
3.1. Lamps
There is no doubt that lamps are the most representative component of street lighting. There are several types of lamps on the market which can be used on this kind of installation, including, among others, high pressure mercury (HPM), high pressure sodium (HPS), low pressure sodium (LPS), metal halide (MH) and light-emitting diode (LED) lamps. At present, in street lighting applications, HPS and MH lamps are the most widely used light sources. LEDs are fast developing light sources and are considered a promising light source for general lighting, although this kind of source on the market is not that cheap yet. Currently, HPS lamps are the dominant light source used in road lighting because of the long lamp lifetime and high luminous efficacy. MH lamps offer high luminous efficacy and good color rendering properties [
20].
There are two different options to save energy in the case of lamps: changing the standards and changing the technology; for example, the British Standard BS 5489 [
21] allows reducing the required lighting class when the color rendering index (CRI) of the lamp is higher than 60 (white light) [
22]. On the other hand, the Hong Kong regulations only allow reducing it if the lamp has a CRI equal to or greater than 80 [
17]. This reduction is only permitted on subsidiary roads. If the current Spanish standard took into account this reduction, the illuminance level would be reduced by at least 25%.
Table 2 shows the reduction of illuminance level.
Analysing
Table 2, we notice that there are two more lighting classes in the British standard than in the Spanish standard. These lighting classes are S5 and S6. In this respect, we must agree with the Royal Decree because 40% of night-time street crime occurs when lighting levels are at 5 lux or below [
23]. Before recommending the incorporation of this advantage into the Spanish standard, it is necessary to be sure that this change does not decrease the quality of the installations. There are several researches that confirm the benefits of white light. One of them is the study conducted by Godfrey [
24], who concluded that driver reactions with cool white light are more efficient than with “warm” coloured light.
Table 2.
Illuminance savings by reducing lighting class.
Table 2.
Illuminance savings by reducing lighting class.
Lighting class (Spain) | Lighting class (BS) | Illuminance level (lux) | New class | New illuminance level (lux) | Illuminance savings (%) |
---|
S1 | S1 | 15 | S2 | 10 | 33% |
S2 | S2 | 10 | S3 | 7.5 | 25% |
S3 | S3 | 7.5 | S4 | 5 | 33% |
S4 | S4 | 5 | S5 | 3 | 40% |
- | S5 | 3 | S6 | 2 | 33% |
- | S6 | 2 | - | - | - |
Lewis [
25] also reported the results of reaction time tests where detection of a pedestrian was conducted using MH, HPM, HPS and LPS. He found an approximately 50% increase in reaction time for sodium sources
versus MH, at a luminance level of 0.1 cd/sq.m. At a relatively high lighting level of 1 cd/sq.m, he reported an increase in reaction time of approximately 15% of HPS
versus MH, and 25% for LPS
versus MH. In our opinion, there are several evidences that prove the benefits of the white light yet the current Spanish Standard does not include it. This should be incorporated in order to improve the energy efficiency.
Related to changes in the technology, it is necessary to guarantee that these changes do not decrease the amount of light output. The best parameter to compare two kinds of lamps without decreasing the luminous flux is the luminous efficacy of the lamp. This parameter is the quotient luminous flux emitted by the power consumed by the source, unit lumen per Watt [
26].
Table 3 shows the main features of the different kinds of lamps.
Table 3.
Main features of the different kind of lamps.
Table 3.
Main features of the different kind of lamps.
Lamp type | Luminous efficacy (lum/W) | Colour rendering index (CRI) |
---|
HPM | 50 | 15 |
HPS | 130 | 25 |
MH | 80–108 | 75–90 |
LED | 90–130 | >80 |
Luminous efficacy is also used by the Spanish regulations [
5], where the minimum values specified is 65 lum/W. As it can be seen in
Table 3, HPM lamps do not comply with the requirements, so it does not make sense for this kind of lamp to appear in the Spanish standard. In the case that lighting designers wanted to change the kind of lamp, they may follow this criterion because it is possible to find lamps with the same or higher luminous flux and less power consumption. For example, by simply replacing common bulbs with energy-saving LED lamps one can reduce energy consumption by up to 80% [
27].
3.2. Ballast
All kinds of lamps require a ballast to operate correctly. For this reason, the presence of this device in street lighting systems is indispensable to ignite the discharge and control the lamp. Ballast devices can be divided mainly into two types: electromagnetic and electronic. Electronic ballasts are considered more energy efficient than electromagnetic ballasts, and for this reason they have been promoted as replacements the latter, to the point that some countries have changed their regulations to encourage their use. Other advantages are that electronic ballasts produce no flicker effects and provide an instantaneous startup [
28]. Due to the fact that electromagnetic ballasts have high power loss from the iron and copper losses in the magnetic choke, they are 10%–15% less efficient than electronic ballasts [
29,
30]. To verify that the power of electronic ballasts is lower than that of electromagnetic ballasts, different ballasts were studied of Philips [
31].
Table 4 shows the power savings for two different LPS lamp powers.
Table 4.
Power savings using electronic ballasts with LPS lamps.
Table 4.
Power savings using electronic ballasts with LPS lamps.
Lamp type & power (W) | Electromagnetic ballast power (W) | Electronic ballast power (W) | Power saving |
---|
1 × SOX 35 W | 11.7 | 3.7 | 8 |
1 × SOX 55 W | 19.5 | 5.5 | 14 |
These power savings are under nominal conditions and although they might be considered insignificant, they should be taken into account because the power saving percentage in the case of a 55 W SOX is 18%. To analyze the benefit of this replacement under normal conditions, the research carried out by Omar [
32] was studied. They examined the energy consumption of 277 units of 250 W HPS for a month. The energy consumption with electromagnetic ballasts was 30,913.2 kWh and the energy used with electronic ballasts was 20,172.7 kWh. Therefore in this case the energy saving was 34.74%. Besides, there are other researchers that have studied the benefits regarding the supply voltage. A good example is the research done by Dolora [
33], who studied the savings for HPS 150 W lamps. This research concluded that the supply voltage bears on in the final energy consumption.
Table 5 shows the power variation regarding the supply voltage.
Table 5.
Power variation between electronic and electromagnetic ballasts [
33].
Table 5.
Power variation between electronic and electromagnetic ballasts [33].
Supply voltage (Vac) | Electromagnetic ballast | Electronic ballast | Power variation (%) |
---|
Power (W) | Illuminance (lx) | Power (W) | Illuminance (lx) |
---|
200 | 122.7 | 161 | 154.2 | 256 | 25.6% |
210 | 136.7 | 190 | 156.8 | 256 | 14.7% |
220 | 152.3 | 220 | 157.1 | 256 | 3.1% |
230 | 168.3 | 255 | 156.0 | 256 | −7.3% |
240 | 187.1 | 291 | 154.8 | 256 | −17.2% |
250 | 204.2 | 327 | 154.3 | 256 | −24.4% |
260 | 224.6 | 367 | 154.2 | 256 | −31.3% |
As it can be seen, when the supply voltage is 250 V, the percentage of power variation is 24.4%, this means that the luminaire power can vary by up to 49.9 W. The problem with the Royal Decree [
5] is that it only specifies the maximum power per luminaire, when in our opinion the maximum ballast power should be specified because designers sometimes are not aware if the kind of ballast that satisfies the requirements.
Table 6 shows a good example.
Table 6.
Luminaire power for different kind of ballasts.
Table 6.
Luminaire power for different kind of ballasts.
Nominal lamp (W) | Maximum lamp power allowed (R.D 1890) [5] | Lamp power plus electromagnetic ballast (W) | Lamp power plus electronic ballast power (W) |
---|
1 × SOX 35 W | 42 W | 46.7 W | 38.7 W |
1 × SOX 55 W | 65 W | 74.5 W | 60.5 W |
As it can be appreciated, lighting designers must pay attention when choosing the ballast because although the maximum power is defined, the luminaire power must be checked because in the analyzed case the installation of electromagnetic ballasts would not satisfy the minimum requirements.
3.3. Street Lamp Globes
Although people believe that street lamp globes do not influence energy consumption, the choice of this part is very important because it influences the upward reflected light and thereby light pollution. Light pollution is not simply any astronomical or ecological light pollution, because enormous amounts of energy are wasted with light pollution. For example, at the end of the 1990s the amount of sky glow over Sapporo, Japan was equivalent to 15 million kWh of energy, 29 million kWh over London, UK and 38 million kWh over Paris, France [
34]. The total amount used for public outdoor lighting in Helsinki, Finland is roughly 170 million kWh, meaning that all Helsinki could be illuminated with just five days of the “waste light” of Paris. The light sent upward is thus estimated to produce economic losses worth billions of euros every year [
35]. The best option to save energy regarding the light pollution is by changing standards. The current Croatian regulation establishes lower levels than the Spanish regulations.
Table 7 compares the maximum upward light ratio of the installation (ULR) for Croatia [
36] and Spain [
5].
Table 7.
Maximum percentage of ULR for Croatia and Spain.
Table 7.
Maximum percentage of ULR for Croatia and Spain.
Croatia standard | Spanish standard |
---|
Classification zone | Maximum ULR (%) | Classification zone | Maximum ULR (%) |
---|
EO | 0% | Not exist | 0% |
E1 | 0% | E1 | 1% |
E2 | 2.5% | E2 | 5% |
E3 | 5% | E3 | 15% |
E4 | 15% | E4 | 25% |
As it can be appreciated, the maximum percentage of ULR in Spain is higher than in Croatia. Although Croatia is not the country with the strictest regulations, in our opinion the Spanish regulation should incorporate at least the minimum level established in the Croatian rules.
The Chilean D.S.N° 686/98 regulation [
37] defines that a lamp with a luminous flux equal to or less than 15,000 Lm cannot emit more than 0.8% of its nominal flux above horizontal level when installed in a luminaire. Lamps with a luminous flux of more than 15,000 Lm should not emit more than 1.8% of their nominal flux above horizontal level when installed in a luminaire.
In 2007 Slovenia adopted a law (Official Gazette of the Republic of Slovenia, No. 81/2007) aimed at tackling light pollution. The law requires that 0% of the output of a luminaire should shine above the horizon (90°) [
38].
To analyze how ULR influences this kind of installation, several simulations were done with the DIALux software. The analysis consisted in studying what happens if the luminaire has the same kind of lamp and the street lighting globes are different. The model of the studied luminaire was the CitySpirit Modern (Philips, Amsterdam, The Netherlands), the street lighting globes were four and the kind of lamp was LED.
Figure 1 shows the average illuminance regarding the ULR for 22 X XR-E-PE/WW, 22 X XR-E-Q3/NW and 22 X XR-E-Q5/CW lamps.
Figure 1.
Illuminance regarding the ULR.
Figure 1.
Illuminance regarding the ULR.
As it can be seen for this luminaire model, if the ULR increases, the average illuminance decreases, but the lamp power and the lit-up surface were the same for the three simulations, therefore the energy efficiency bears upon the ULR. Another model of lamp analyzed was the Urbana (Philips) and again ULR was studied and the same performance can be appreciated, the street lighting globes were two in this case and the lamp was an HPL-N80W.
Table 8 shows the results.
Table 8.
ULR regarding the street lamp globes.
Table 8.
ULR regarding the street lamp globes.
Globes | Total lamp flux | System flux | Luminaire power | Lit-Up surface | E average (Lux) | Overall uniformity | ULR (%) |
---|
| 3600 | 684 | 90 | 80 | 5.02 | 0.2 | 4.5 |
| 3600 | 792 | 90 | 80 | 4.74 | 0.273 | 16 |
It is possible to think that as the system flux is higher in the second option than in the first option, the average illuminance would be higher than the first one, but the reality is that as this sort of street lighting globes does not have any device to avoid the light pollution, and thus the average illuminance is lower than in the first case. From our point of view, ULR magnitude should be taken into account for the energy label, because with the current systems only assess the illuminance on the lit-up surface.
3.4. Hours of Operation
The current Spanish standard [
5] includes three possible devices for that purpose: astronomic time switches, twilight switches and remote management systems for electrical boards. Astronomic time switches turns lights on and off with a fixed time offset from sunrise and sunset. To estimate the daily hours of sunrise and sunset the latitude and longitude are needed because of the movement of the Sun, as it can be seen on the sunrise sunset calculator program tool [
39].
Twilight switches measure the amount of natural light available to turn on and off the lamps regarding this level. As happens with astronomic time switches, it is possible to establish an approximation of the number of burning hours using the latitude and the level of natural light required to turn the system on or off [
40]. An option to decrease the hours of operation and therefore to save energy with this kind of device is by changing the use pattern. Angus Council (U.K.) [
41] studied the trimming of photocells; the factory setting of the switch on/switch off levels are 70 lux on and 35 lux off (70/35). By reducing the switch ratio to 35/18 they could typically save 92 burning hours per year per luminaire. The Institution of Lighting Professionals (ILP) [
42] estimated that if the switching levels were reduced 35/16, a saving of 1%–2% per luminaire could be achieved. This regulation is not recommended for older lamp types such as LPS and HPM operating on conventional ballasts. Such installations should be operated at 70 lux on and 35 lux off as a minimum to allow the lamps to fully run up by the time the lighting is required. The only drawback of the previous studies is that they did not specify the latitude. This lack of information was solved in the study carried out by American Electric Lighting (AEL) [
43] because the latitude was taken into account in the results.
Table 9 shows the hours of operation at latitude 35° (Los Angeles, California) for various photocell settings.
Table 9.
Hours of operation regarding the twilight settings [
39].
Table 9.
Hours of operation regarding the twilight settings [39].
On (lux) | Off (lux) | Hours of operation |
---|
8.6 | 10.7 | 4113 |
10.7 | 12.9 | 4130 |
32.2 | 19.3 | 4187 |
16.1 | 24.7 | 4167 |
27.9 | 33.3 | 4204 |
10.7 | 32.2 | 4265 |
21.5 | 107.6 | 4340 |
Remote management systems are composed by a server-client architecture system for monitoring, detecting, controlling and communicating problems instantly to a central control room or directly to maintenance technicians [
44]. Telemanagement integration in street lighting networks of small cities has hardly been developed both in a conceptual and applicative way, especially due to limited economical resources of local communities which have become responsible for too many new tasks, public illumination being one of them [
45].
The hours of operation depend on these devices which also consume energy. Analyzing the data of the manufacturer ORBIS [
46], it can be appreciated that the power consumptions are very similar independently of the kind of device.
Table 10 shows the power consumption.
Table 10.
Power consumption of street lighting control systems.
Table 10.
Power consumption of street lighting control systems.
Type | Model | Self-Consumption (VA) |
---|
Astronomic | DATA ASTRO | 5 |
ASTRO NOVA CITY | 6 |
ASTRO UNO | 6 |
Twilight | ORBIFOT | 8 |
VEGA | 8 |
ORBILUX | 3.4 |
Remote Management | XEO LUM | 4.8 |
As each device uses different technology and criteria to turn on and off, the hours of operation established for each device will be different. We have measured the natural light level during different days with the purpose of understanding the operation of each device.
Figure 2a shows the natural light level several days at sunrise and
Figure 2b shows the natural light level of several days at sunset, where the data of both figures were measured in Madrid (Spain) in September 2014. A PCE-174 (ORBIS) digital illuminance meter was used to obtain the data.
Figure 2.
Natural light level during the sunrise (a); and during the sunset (b) in Madrid.
Figure 2.
Natural light level during the sunrise (a); and during the sunset (b) in Madrid.
As it can be seen, the tendency is different for each day because of the weather conditions are different and therefore climate bears upon the natural light levels. In that aspect we agree with Howell [
40] that the weather conditions are even more significant than latitude in determining days. Hence the main drawback of astronomic time switches is that they do not take into account the real level of natural light. Besides, analyzing in detail the data of the previous trimming,
Table 11 shows the time when the natural light reached a certain value. It can be seen, trimming the photocells allows decreased the hours of operation, while on the other hand natural light level reached 35 lux twice on 21 September. This issue is the main problem of photocells because undulations in light level can cause erratic operation, but this can be solved with the controller.
Table 11.
Time when the natural light reached a certain value.
Table 11.
Time when the natural light reached a certain value.
Day | Sunrise | Sunset |
---|
18 (lux) | 35 (lux) | Savings Minutes | 70 (lux) | 35 (lux) | Savings Minutes |
---|
19 September 2014 | 7:46 | 7:48 | 2 min | 20:29 | 20:32 | 3 min |
20 September 2014 | 7:46 | 7:47 | 1 min | 20:28 | 20:31 | 3 min |
21 September 2014 | 7:46 | 7:48 | 2 min | 20:21 | 20:24 and 20:27 | ------- |
22 September 2014 | 7:47 | 7:48 | 1 min | 20:25 | 20:28 | 3 min |
23 September 2014 | 7:41 | 7:45 | 4 min | 20:24 | 20:27 | 3 min |
As it can be seen, photocell trimming could save approximately 4 min per day. This means that the amount of burning hours may reduce by 24 per year.
3.5. Lighting Level Control Devices
There are three different types of level control devices contemplated in the Spanish standard [
5]: series inductive type ballasts for dual power level, power controlled electronic ballasts and regulators and stabilizers in the head of the line.
The main problem of using ballasts for dual power levels is that these systems act locally, requiring an adjustment device attached to each of the individual charges and also a general control system to control all of them [
47]. Regulators and stabilizers are able to control the voltage according to different parameters such as number of vehicles per hour [
48], weather conditions or the presence of pedestrians [
49]. Their operation consists of hanging the input mains voltage to a variable voltage within the range from 220 to 170 V [
50]. Those changes are accompanied by variations of illuminance and lamp power.
Figure 3 shows the working of these sort of systems, where it can be seen their potential on energy savings.
The main advantage of stabilizer lighting systems is that they are able to avoid overvoltage situations. The research carried out in China [
51] showed how, despite the fact the nominal voltage is established at 230 V like in Spain, it reached values as high as 246 V. This overvoltage situation is the main reason for the shortened lifetime of lamps.
Figure 3.
Regulator and stabilizer devices.
Figure 3.
Regulator and stabilizer devices.
Taking into account that the energy savings depend on input voltage, it is necessary to define the input voltage in order to satisfy the minimum luminous flux level allowed. According to Bacelar [
52], the minimum luminous flux level should be established at 50%, because it was shown that this dimming does not seem to have a great influence to the visibility of observers nor drivers. Furthermore this minimum level coincides with the current standard [
5]. Following the recommendations of General Electric [
53], the minimum voltage regarding the kind of lamp is shown in
Table 12.
Table 12.
Minimum voltage regarding the kind of lamp according to GE [
53].
Table 12.
Minimum voltage regarding the kind of lamp according to GE [53].
Kind of lamp | Minimum voltage (Vac) |
---|
HPM | 200 |
HPS | 180 |
LPS | 190 |
MH | 180 |
Analyzing in detail the research conducted by Yan [
50], who studied the characteristics of HPS lamps of 50, 70, 100, 150, 250 and 400 W dimming the voltage. It can be observed that the percentage of light output decreases more than 50% for 180 Vac in the case of HPS and MH.
Figure 4 shows the percentage of light output for the case of 50 and 70 W HPS lamps.
Figure 4.
Lamp power, light output and minimum voltage for HPS lamps (50 and 70 W).
Figure 4.
Lamp power, light output and minimum voltage for HPS lamps (50 and 70 W).
Therefore the minimum voltages showed in
Table 12 are not completely right because they do not satisfy the minimum requests impose by the current standard. In our opinion the minimum voltage for each kind of lamp should be the values shown in
Table 13.
Table 13.
Minimum voltage to decrease the light output 50%.
Table 13.
Minimum voltage to decrease the light output 50%.
Kind of lamp | Minimum voltage (Vac) | Decrease luminosity flux (%) |
---|
HPM | 200 | 30% |
HPS | 190 | 50% |
LPS | 190 | 10% |
MH | 190 | 50% |
From our point of view, the unique shortcoming of Spanish standard [
5] regarding lighting level control devices is that it does not specify when it can be used. If we followed the recommendations of the Dutch ministry, dimmable road lighting systems could operate at 20% when the density of traffic at night is low, at 100% when the traffic density is high and 200% when there is a combination of high traffic density and exceptional conditions such as fog. The conclusions were that 20% light level has no negative safety effects and is sufficient for low traffic density but 200% light level is not justified because the cost is high and the safety improvements are marginal at best [
54]. Another project [
55] also investigated the effect of dimming, the lighting level setting were determined as follows; 100% when there are more than 3000 vehicles per hour, 75% when the range of vehicles is 3000–1500 and 50% when the number of vehicles per hour is lower than 1500. Following both projects and observing the behavior of Spanish roads,
Figure 5 shows the number of vehicles per hour of a road in the Community of Murcia.
Figure 5.
Number of vehicles per hour in a road from the Community of Murcia [
56].
Figure 5.
Number of vehicles per hour in a road from the Community of Murcia [
56].
As it can be seen, lighting level control devices can operate perfectly from 1:00 am to 5:00 am, because the number of vehicles decreases considerably. In this aspect, the Croatian normative [
57] specifies that if the local government does not prescribe a schedule, the street lighting must be turned off or reduced by 50% at least at 1:00 am. In our opinion, it should be mandatory within the Spanish normative that lighting level control devices reduce the light levels at least from 1:00 am, because most of the time the conditions allow it.
3.6. Renewable Energies
The global necessity for energy savings requires the usage of renewable sources in many applications and outdoor lighting installations are no exception. Spain, owing to its location and climate, is one of the countries in Europe with the most abundant solar resources [
58]. Global solar irradiation on a horizontal plane is estimated as being between 1.48 and 3.56 kW/m
2 day in Spain.
The solar energy option may be the best solution in the case of an autonomous street lighting system because of the long life time, easy installation and modularity [
59]. This sort of renewable energy allows reducing the CO
2 emissions considerably and thus the energy consumption. A good example of the benefits of solar energy in street lighting is the research carried out by Nunoo [
60], who achieved energy savings per day of 603 kWh. Analysing in detail the research carried out by Constantinos [
61], who optimized a photovoltaic system for street lighting, the total autonomous days of operation may reach up to 315 per year. In other words, in this case the energy savings were about 86%.
On the other hand, maintenance of the photovoltaic panels is very important, because dust effects reduce the performance of solar panels. The research carried out by Al-Almmri [
62] shows that the losses of the output power of the fixed solar panel can reach 26% for one month. As well, their orientation can cause a considerable loss of efficiency. Likewise, the slope of the panel should be changed two to four times a year to maximize the solar absorption, since the optimum slope in the summer is not the same as the optimum one in the winter [
63]. These drawbacks can be solved with regular maintenance.
Outdoor lighting can be supplied with other kinds of renewable sources or even a combination of several types of renewable sources like the research performed by Al-Fatlawi [
64], who combined solar and wind energy. Power systems which include photovoltaic systems and wind turbines typically include energy storage devices so that loads can be operated when solar energy is not available or when wind velocities are too low to generate power [
65].
Nowadays, renewable energies are indispensable to satisfy the normative for buildings, however the Royal Decree [
5] overlooks this subject in the field of street lighting. Previous research shows that the incorporation of solar energy for street lighting is an incredible opportunity to reduce energy consumption and improve the quality.