Energy and Buildings 159 (2018) 542–557

Contents lists available at ScienceDirect

Energy and Buildings

journal homepage: www.elsevier.com/locate/enbuild

Comparative study of lighting quality and power quality for LED and
HPS luminaires in a roadway lighting system
Suntiti Yoomak a , Chaiyan Jettanasen a , Atthapol Ngaopitakkul a,∗ , Sulee Bunjongjit b ,
Monthol Leelajindakrairerk b
a
Faculty of Engineering, King Mongkut’s Institute of Technology Ladkrabang, Bangkok 10520, Thailand
b
Faculty of Engineering, Rajamangala University of Technology Rattanakosin, Phutthamonthon Sai 5 Road, Salaya Phutthamonthon, Nakhon Pathom,
73170, Thailand

a r t i c l e i n f o a b s t r a c t

Article history: Roadway lighting was one of the earliest applications for electric lighting technologies. Well-designed
Received 1 February 2017 roadway lighting offers many benefits and aids the visual tasks of drivers and pedestrians. Currently,
Received in revised form 5 October 2017 the trend of energy management for exterior general lighting has increased in several cities. Generally,
Accepted 25 November 2017
high-pressure sodium (HPS) lamps have been the standard throughout Thailand since the 1980s, but the
Available online 5 December 2017
use of HPS lamps has driven high energy consumption. To conserve energy, light emitting diode (LED)
street lighting has quickly become a new trend. The primary objective of this paper is to deliver an initial
Keywords:
and in-depth assessment of LED technology for roadway lighting applications. This analysis focuses on
LED lamp
Roadway lighting
the performance of LED luminaires compared with the existing standard of HPS cobra-head luminaires
Energy savings in terms of lighting quality, energy savings, power quality, and investment. The analysis is conducted
Illuminance via three components. First, simulation of the roadway lighting quality is performed using the DIALux
Power quality programme to evaluate and compare with the standard of Thailand. Second, an experimental setup is
used to evaluate the energy savings and power quality for cases with and without solar power. Third,
the discounted payback period (DPP) criterion and internal rate of return (IRR) are applied to evaluate
the investment scenarios. The obtained results conclude that the LED luminaire performs better than
the HPS luminaire in terms of power quality and energy savings (energy-efficient index reduction of
approximately 40% over the HPS luminaire), which indicates that the utility benefits from use of LED
luminaires. However, although the lighting quality results for LED luminaire have values fairly close to
that of the standard, the lighting quality should be improved in terms of uniformity. The results obtained
from this analysis are expected to help to improve the roadway lighting standard of Thailand in the future.
© 2017 Elsevier B.V. All rights reserved.

1. Introduction savings and CO2 emissions reduction was presented in [5]. To

address these critical aspects and improve the base case scenario,
Energy consumption in the transport sector [1–3] accounts for many energy conservation strategies have been considered. The
37% of total energy consumption, primarily due to reliance on road obtained results show that great improvements in lighting quality
transport, which requires 95% of energy consumption of all trans- are foreseeable together with large energy and economic savings.
port modes. Past literature reviews [4–20] show that road lighting The study in [6], analysed public lighting energy management in the
has significant safety benefits. Field measurements and evalua- Croatian city of Rijeka to determine the connection between energy
tion of new lighting systems were compared with the conventional market liberalization and sustainable development in urban areas.
high-pressure sodium (HPS) lighting, as proposed in [4]. The results The result indicated a significant connection between investment
showed that the difference between the rated and measured power in energy management of public lighting and its influence on lower
values could be significant. A method for assessing savings, select- emissions of carbon dioxide (CO2 ). In [7], the study assessed the dif-
ing technology options and offering proof of achievable energy ference between a passenger and driver in a night-time detection
task while driving on a circuit under road lighting. The results pre-
sented in this paper showed that driving activity, even with a low
demand on information processing and a low demand on vehicle
∗ Corresponding author.
handling, should be included in future road visibility indexes. To
E-mail address: knatthap@kmitl.ac.th (A. Ngaopitakkul).

https://doi.org/10.1016/j.enbuild.2017.11.060
0378-7788/© 2017 Elsevier B.V. All rights reserved.
 S. Yoomak et al. / Energy and Buildings 159 (2018) 542–557 543

improve night-time road safety, road lighting design should also lamp in lighting systems has a long history. The designs of the elec-
consider the various levels of environmental complexity that lead tronic ballast and EMI for the HPS lamp were presented in [21].
to various levels of driver workload. The impact of public lighting on The results showed that a single-stage electronic ballast with a full
pedestrian well-being and perceived safety was studied in [8]. The bridge inverter reduced the total harmonic distortion, and the cur-
data for each street were considered, including the average illu- rent waveform could be stabilized using an EMI filter circuit. In [22],
minance, illuminance uniformity, and colour of light. The results the development of a dimmable electronic ballast with ZigBee was
indicate that well illuminated streets with higher illuminance lev- presented for HPS lamps in flexible public lighting. The results were
els tend to make people feel safer and more comfortable. In [9], the measured in terms of real power and energy consumption using a
paper summed up as many relevant influence factors as possible to central controller to control the power consumption of the HPS
offer recommendations intended for both energy savings and opti- lamp by turning it on and off. Voltage amplitude modulation via
mal performance of street lighting. The results in [10,11] showed a an autotransformer was presented in [23]. Certain characteristics
relationship between the population and energy consumption that of HPS and high intensity discharge (HID) have been investigated
is valid for several regions. Previous work also showed that the with variation of the supply voltage network between 160 V–250 V.
acceptability of reduced street lighting policies can be increased by In [24], a device based on phase modulation of the network volt-
offering information on environmental impact of street lighting, as age was used to increase the energy efficiency by dimming the
proposed in [12]. This result suggests that in terms of acceptability, high intensity discharge (HID) lamps and magnetic ballasts. The
the difference between each lighting level decreased as the lighting results showed that this device offered simple individual control
levels increased. Therefore, information on environmental benefits for single or double lamps and easy installation but the drawbacks
congruent with the relevant biospheric values can aid in making included a low power factor and harmonic waste. Installation of
reduced street lighting levels more acceptable. A methodology for power reducer equipment for energy conservation in street lighting
linking roadway lighting characteristics to visual performance lev- via a transformer was proposed in [25]. In this work, dimmer tech-
els and to traffic crashes was presented in [13,14]. A developing nology was presented in terms of 2 types: dimming of power using
methodology and architecture for public street lighting remote a transformer to reduce the amplitude of the sinusoidal wave and
operation and a supervision system was proposed in [15]. In a pre- dimming of power using a power electronic circuit for a chopping
vious paper [20], the mounting height adjustment method and the wave shape. The result showed that the energy savings obtained
pole spacing adjustment method were varied to improve roadway were 25% and 30% using the transformer and electronic circuit,
lighting. The results suggested that adjustment of the pole spac- respectively. However, the cost of power reducer equipment tends
ing between lighting poles could significantly improve the average to increase due to costly equipment, including the metreing and
illuminance and uniformity. telecommunication system, with higher investment requirements.
As a result, from the end-user or pedestrian point of view, A comparative study of the operational characteristics of a 250 W
roadway lighting is important in increasing visibility. From the HPS lamp fed by an electromagnetic ballast versus an electronic
lighting system owner or utility point of view, roadway lighting ballast was presented in [26]. The comparison study covered both
is important because the selected system meets the photometric the warming period and steady state (or nominal power opera-
requirements of the application as effectively as possible such that tion). The results concluded that the performance of a lighting
the costs of operating the system are minimized. From a lighting system composed of an HPS lamp fed by an electronic ballast is
perspective, roadway lighting should deliver sufficient light levels, superior in almost all operational aspects to the same lamp fed
uniformity, and target contrast according to the type of road and through a conventional electromagnetic ballast even if a power fac-
accommodate visual needs under various lighting conditions. The tor correction capacitor is used. In [27], a method was presented
most common visual needs of drivers and pedestrians include iden- for reducing energy consumption from lighting with a centralized
tification of objects, obstacles and individuals in both the direct field energy-saving system and HPS lamps mounted with a magnetic
of view and in the periphery. Equally important is to minimization ballast and capacitor. Centralized systems were analysed because
of glare (direct and reflected) and light pollution/trespass to make they are the most commonly used system for energy savings in pub-
the appearances of the spaces appealing. Thus, the effectiveness lic lighting. In addition, this paper presented a comparative study of
of a given roadway lighting system should be determined by how the response of a test system supplied by rectangular or trapezoidal
well the lighting goals are met with consideration of the power and voltage rather than conventional sinusoidal voltage. The goal of this
cost needed to achieve them, rather than judging only the luminous comparative study was to identify new techniques for HPS lamp
efficacy of the light source or luminaire in the absence of context. dimming using a centralized energy-saving system for streetlights.
For public and roadway lighting in Thailand, the responsible As a result, the installation of HPS lamps has driven high energy
organizations that have authority over public and roadway lighting consumption. To conserve additional energy; light emitting diode
are the Provincial Electricity Authority (PEA), Metropolitan Elec- (LED) street lighting has quickly become a trend.
tricity Authority (MEA), and Expressway & Rapid Transit Authority LED lighting represents the next generation of evolution in
of Thailand (ETA). PEA and MEA are the only major organizations energy-efficient lighting solutions. LED lighting is also referred to
of free electricity supply for streetlights and public lights, as pre- as the fourth generation of lighting and has already been applied
sented in Table 1 [3]. According to the data in Table 1, the energy in illumination systems. Several papers [30–57] have studied and
consumption for public lighting in each year is quite high, especially developed LED lighting systems for either interior or exterior
the streetlights, and is likely to immensely due to the government’s lighting. Additionally, research has been reviewed in terms of appli-
policy on urbanization. cations, advantages and emerging issues of LED luminaires [30–35],
The lamps that are primarily used in public and roadway lighting and comparative studies have been conducted on the performance
in Thailand are fluorescent, mercury vapour, metal halide (MH) and of various types of luminaires, including LEDs [30–32]. The result of
HPS versions. The HPS lamp is one of the most commercially popu- this study shows a reduction of power consumption compared with
lar lamps used in roadway and highway lighting systems since 1968 the traditional luminaire. However, power quality, lighting quality
because the HPS lamp is superior in terms of efficiency, lifetime, and comfort in occupancy must be considered before implement-
and luminosity. The HPS lamp has become the standard through- ing LED into the lighting system. In [33], energy-saving calculations
out Thailand since the 1980s. However, energy efficiency plays an were performed for LED technology only. One of the most interest-
important role in energy consumption and service costs. Accord- ing and quickly developing lighting technologies used to replace
ing to past literature reviews [21–29], the development of the HPS old technologies is solid-state lighting, especially LED. Even if the
544 S. Yoomak et al. / Energy and Buildings 159 (2018) 542–557

Table 1
Classification of Electrical energy for Thailand (kW h) [3].

Classification MEA PEA

2012 2011 2010 2009 2012 2011 2010 2009

Roadway and public lighting sector 358.83 349.58 346.09 271.96 29,046.9 27,364.9 27,290.2 24,862.5
Business and industry sector 47,885.17 43,847.75 44,714.07 41,461.34 80,845.5 73,761.1 73,459.9 65,282.4
Total 48,244.0 44,197.33 45,060.16 41,733.30 109,892.4 101,126.0 100,750.1 90,144.9
Index (based on 2009 = 100%) 115.60 105.90 107.97 100.0 121.9 112.2 111.8 100.0

development of LED leads to increasing consumption of light, the 75 W) was used in the experimental tests. A range of passive LED
possible energy savings are immense compared with the current drivers with high efficiency, high reliability, high luminous perfor-
technology. Previous results demonstrated the potential of using mance and low component counts were introduced and compared
LEDs to reduce household lighting energy use and the correspond- in [51] including comparison of electrical performance in terms of
ing CO2 emissions in Finland. However, the savings estimation in energy efficiency, input harmonics, loss analysis, and distribution. A
this paper was based on the current level of household lighting novel single-stage, high-power-factor LED driver for street-lighting
energy use. In [34], the study focused on the change in the lighting applications was presented in [53]. A 144 W prototype LED driver
set in a house, particularly LEDs. The economic cost for LED lumi- with a 110 V input utility line voltage was successfully built and
naire replacement has also been calculated [35]. Another important tested. In addition, the measured current harmonics of the proto-
factor that must be addressed is the heat generated from LED light- type LED driver were compared with the IEC 61000-3-2 Class C
ing and its effect on energy savings in a building [36,37]. Another standards and experimental results to verify the feasibility of the
advantage of the LED luminaire is that the technology can be fur- proposed LED circuit.
ther developed by an applied control strategy used to dim the light As a results, in the literature for LED lighting systems [30–57],
output during daylight and harvest natural light instead of artificial researchers have mostly focused on interior lighting with LED
light from the luminaire, thus further reducing energy consumption lamps, but these research papers have not mentioned exterior
in the lighting system. lighting with LED lamps. Street lighting (or roadway lighting) is
The work in [39] reports the advantages and disadvantages of one of the fastest growing application sectors for LED technology
demonstrative LED street lighting in Taiwan from 2009 to 2011 because the current trend of using LED lamps for exterior gen-
with the aim of determining whether these projects can meet the eral lighting has increased in several large and small cities. LEDs
government’s expectations. The study showed that the replace- in exterior general lighting lamp applications (streets, parking lots,
ment of mercury vapour lights with LED lights could save up to roadway tunnels, bridges, landscaping, pools/fountains, buildings,
50% of electricity and that the acceptance level of the general pub- and architectural and other general lighting) represent nearly 93%
lic reached 80% such that LED lights should be promoted for wide of worldwide consumption in 2010. Up to 2019, the relative market
adoption. In [40], research on energy efficient greenhouse light- shares of component-level LEDs used in exterior lamps is fore-
ing with LEDs was reviewed. In [41], the effect of light distribution casted to decrease significantly to nearly 59% but will increase
on perceived personal safety at night was explained according to in value to $5.4 billion. In China, the street lamp market size is
changes in people’s perceptions of option, escape, and conceal- approximately 28–30 million units and nearly 1.5–2 million new
ment. Using LED-based luminaires, this paper manipulated light street lamps are installed every year, mainly mercury and sodium
distribution independently of other physical street characteristics lighting sources. China began using LED street lamps in 2007, later
that might affect perceived personal safety. In addition, this paper than in the USA and Japan. After the Chinese government replaced
presented data from two experiments with stationary and walk- conventional street lamps with LED street lamps, many negative
ing participants. The results demonstrated that people preferred to opinions of the product characteristics arose, such as the high fail-
light in their own immediate surroundings rather than on the road ure rate, shorter-than-expected lifespan, and glare problems. These
ahead. The harmonics and the effect of voltage sags on the LED results suggest that significant opportunity for improvement in the
lamp were investigated in [42]. Twelve samples of LED lamps with product characteristics of the LED street lamps exists because the
different power ratings from various manufacturers were tested current product is not completely satisfactory.
to analyse the characteristics of the LED lamp. The experimental For LED lighting in Thailand [58], an international exhibition
results showed that the LED lamps produced a considerable amount in electric lamps known as the “LED Expo Thailand 2013” was
of current harmonics, but this harmonic distortion could be reduced organized in May 2013 by the Electricity Generating Authority of
by combining various types of LED lamps. Voltage sag tests showed Thailand (EGAT). Technologies and innovations in LED products
that all of the tested lamps were sensitive to sag depth and sag from manufacturers in different countries were presented to offer
duration. knowledge of high-performance lighting products and to promote
To offer a solution for use of LED to replace the traditional light- their wide use, contributing to the energy savings of the country.
ing device, a comparison of HPS lamps and LED lamps was proposed A study in a project promoting the use of LED lamps found that
in [44] based on energy efficiency and energy conservation, and the LED lamps could reduce the use of electricity by approximately
photometric quantities of LEDs were compared with those of the 60%–70%. EGAT has recently promoted the use of LED light bulbs in
traditional light source for road lighting. The optimal choice of LEDs many campaigns, including distribution of LED bulbs to the lead-
and thermal management solutions for a street-lighting applica- ing outlets country-wide. In addition, EGAT hosted the LED Expo
tion was studied in [47]. However, due to the low luminous flux Thailand in 2015. All of the campaigns aimed to create a power
of one single LED compared with a single HPS lamp, it is neces- conservation model. A pilot project was conducted to replace 8500
sary to integrate additional LED devices on a luminaire to reach the LED street lamps at EGAT’s 8 dams, power plants, and head office
equivalent luminous flux. An optical design method was proposed and to collect data on energy use and savings. The project can be
to solve the problem of illuminance distribution for LED street used as public example in the use of LED lamps for lighting pur-
lighting luminaires in [49]. This method selected LED, lenses, and poses. EGAT also planned to replace and install more than 21,000
driver for a full luminaire design. Moreover, a luminaire prototype LED light bulbs in 2015.
with 30 LEDs (total luminaire power consumption of approximately
 S. Yoomak et al. / Energy and Buildings 159 (2018) 542–557 545

Fig. 1. Luminous intensity distribution curve or polar curve.

In addition, PEA launched a management project to reduce (or cobra-head style lamp). The results reflected in this work are
energy consumption and greenhouse gas emissions as a result of generally representative of the operating conditions. However, the
PEA’s direct and indirect day-to-day activities. Following a thor- results in any other situation might vary depending upon particu-
ough study of the Corporate Social Responsibility (CSR) strategy lar operating conditions. The remainder of this paper is organized
[1], PEA’s Social and Environmental Master Plan 2011–2014 was as follows. Section 2 presents simulation and modelling of a typ-
developed with the aim of improving the social and environmen- ical street lighting scenario using lighting simulation software to
tal operations to comply with a new strategic plan that featured examine photometric performance. The simulation evaluations
three foremost strategic groups: Fair Electricity, Sufficient Electric- are compared in terms of average illuminance, average lumi-
ity, and Renewable Electricity. A conceptual project on the smart nance, disability glare, and uniformity of luminance. Based upon
use of public electricity by adjustment of electric power and bright- the simulation results, laboratory analyses were conducted on an
ness to suit the street types and operating periods was developed in experimental setup and the results are presented in Section 3. The
compliance with the Highway Department regulations. Previously, performance evaluation in the laboratory (experimental setup) is
PEA launched a pilot project by installing energy-saving devices aimed at determining the photometric performance and energy-
for LED street lighting in countrywide public lights. In addition, the savings potential. Power consumption and power quality measures
government has established a policy for subsidizing the use of LED such as voltage, current, real power, frequency, power factor, and
lighting for road illumination to promote energy-saving measures. current THD are measured and monitored in this section. The break-
Currently, more than 20,000 LED lights are already in service, and even point is calculated to compare the LED luminaires with the
additional lights will be replaced and put into service on a yearly HPS luminaires, and the results are presented in Section 4. Finally,
basis [58]. In terms of market application of roadside LEDs a suffi- selected conclusions are summarized in Section 4.
cient amount of positive results and benefits should be gathered,
and the advantages and disadvantages of the use of LED should be
scrutinized to avoid the negative results from LED products with 2. Simulation of lighting quality
inferior quality that could adversely affected the market.
As a result, according to the abovementioned papers [21–57], Many LED luminaires for roadway lighting applications are
LED technology is already available, mostly for small roads in urban available in the market with additional luminaires entering the
and pedestrian areas. Although the use of LEDs for street lighting is market almost every day. In a previous paper [20], LED luminaires
increasing, a limitations exist in lamp replacement. The replace- with three types of luminous intensity distribution curves (or polar
ment of existing street lamps with LED lamps is limited by the curves) were compared on the surface road in terms of average hor-
high initial cost and the limited (but increasing) number of manu- izontal illuminance (or average illuminance) and uniformity to find
facturers that offer LED modules. Estimation of the energy-saving the best polar curve. The results showed that the type of polar curve
potential of LED technology for street lighting strongly depends on could also significantly improve the average illuminance and uni-
the existing system technology in an individual community or in a formity. Generally, the average illuminance on the surface road was
city. Streetlights in Europe can be equipped with HPS lamps, mer- calculated in terms of pedestrian safety in visualizing and the iden-
cury vapour lamps, or compact fluorescent lamps. Currently, LED tifying objects and individuals in the direct field of view, and the
street lights are not suitable for large streets because the current average luminance and disability glare were calculated for drivers
characteristics lead to unpleasant blinding effects that influence in terms of aiding the visual tasks of drivers. In all of the above
road safety. The goal of this paper is to offer an in-depth anal- papers [20,30–42], researchers primarily focused on the obtained
ysis of the feasibility of LED technology that could replace HPS average illuminance from LED luminaires, but the average lumi-
lamps in roadway lighting applications. In addition, this analy- nance and disability glare were neglected. In this section, the 250 W
sis aids in identifying the performance factors, maximum energy HPS lamp is compared with the 115 W LED lamp using the DIALux
efficiency indicators, energy savings, and maintenance cost reduc- programme for all arrangements of lighting pole. The luminous
tion potential of LED luminaires compared with HPS luminaires intensity for the HPS and LED lamps is plotted on a polar curve,
546 S. Yoomak et al. / Energy and Buildings 159 (2018) 542–557

Fig. 2. Layout of roadway for simulation.

as shown in Fig. 1. In addition, the luminous flux of the HPS lamp than that of the HPS luminaire case, with values close to that of the
is 28,000 lumens and that of the LED lamp is 17,400 lumens. How- base case (roadway lighting standard of Thailand). For the next case
ever, the total lumen output for the LED and HPS luminaires cannot study, after changing the arrangement of roadway to opposite, the
be compared directly due to how the light is distributed by the average illuminance (Eav ) of both luminaires is greater than the base
luminaire. The simulation evaluations were conducted in terms of case, which indicates that the arrangement of roadway luminaires
average illuminance, average luminance, disability glare, unifor- has an impact on average illuminance (Eav ), as shown in Table 2. To
mity of luminance, and maximum energy efficiency indicator. support this assertion, the staggered roadway arrangement shows
The roadway for the case study is simulated using the DIALux that the obtained average illuminance (Eav ) of the HPS and LED
programme, as depicted in Fig. 2. According to Fig. 2, two arrange- luminaire cases are still greater than that of the base case.
ments of roadway lighting were investigated: without isle (single The uniformity ratio of illuminance (Emin /Emax ) shows that
side, staggered, and opposite) and with isle (opposite on the car- when a single-side arrangement is considered as an example, the
riage way, twin central, and staggered and opposite). To determine uniformity ratio of illuminance obtained from the LED luminaire is
the illuminance for a roadway application, the task plane is defined less than that of the base case, and the obtained uniformity ratio of
as the area on the roadway prior to simulation. The X-dimension the HPS luminaire is greater than that of the base case. However,
of the task plane is the spacing between luminaires on each side of if the arrangement of the roadway is changed, both cases (HPS and
the roadway and the Y-dimension of the task plane is the direction LED luminaires) are greater than the base case.
of travel encompassing the travel lanes on the roadway. As shown Considering the average luminance (Lav ), it can be observed that
in Fig. 2(a), the width of the road (Y) is 7 m and that of the street when the single-side arrangement is considered as an example, the
isle (Z) is 0 m (without street isle). Similarly, in Fig. 2(b), the width average luminance obtained from the LED luminaire is less than
of the road is also 7 m but the street isle width is 3 m. that of the base case, whereas the average illuminance in the case
Considering the simulation, the mounting height, overhang, and of the HPS luminaire is greater than that of the base case, which
spacing between luminaires are determined to be 9, 1, and 36 m, is the same behaviour observed for the average illuminance. If the
respectively. In addition, the boom angle is 15◦ . Once the task arrangement of the roadway is chaged, both cases (HPS and LED
plane dimensions were determined with the illuminance, a calcu- luminaires) are greater than the base case.
lation grid with 1 m by 1 m cells is created, and the corresponding For the uniformity ratio of luminance (uo and ul ), when the
illuminance values are calculated. Thus, the grid data between single-side arrangement is considered as an example, the over-
luminaire-to-luminaire account for 252 data points per width of all uniformity ratio of luminance (uo ) in the case of the LED
road. Although, many standards exist for roadway lighting, the luminaire has a value less than that of the base case, but in the
Department of Highway of Thailand standard was considered as case of the HPS luminaire, this value is greater than that of the
the base case for evaluation of the HPS and LED luminaires. After base case. Similarly, the longitudinal uniformity ratio of lumi-
applying the various arrangements and lamps, the results obtained nance (ul ) obtained from the cases (HPS and LED luminaires)
from the DIALux programme for each case study are illustrated in is less than the base case, which indicates that the single-side
Tables 2 and 3. arrangement with the LED luminaire should also be improved. If
The results for road arrangements without the street isle are the arrangement of the roadway is changed, the overall unifor-
shown in Table 2. In Table 2, the HPS and LED luminaires of the mity ratio of luminance (uo ) for both luminaires (HPS and LED
single-side arrangement are considered as an example. The average luminaires) is greater than that of the base case for roadway
illuminance (Eav ) in the case of the LED luminaire has a value less arrangements that are opposite and staggered. In addition, the lon-
 S. Yoomak et al. / Energy and Buildings 159 (2018) 542–557 547

Table 2
Comparison between illuminance of HPS and LED lamps for various arrangements in a road lighting system without street isle (surfaces of road = Portland cement concrete).

Data Standard HPS lamp LED lamp

Single side Opposite Staggered Single side Opposite Staggered

Eav (lux) ≥21.5 36 73 73 21 42 42

Emin (lux) – 17 39 54 4.64 17 31
Emax (lux) – 69 133 92 44 86 55
Emin /Eav ≥0.4 0.472 0.534 0.74 0.22 0.40 0.74
Emin /Emax ≥0.167 0.249 0.29 0.589 0.11 0.20 0.56
Lav (cd/m2 ) ≥2 3.53 7.09 7.09 1.84 3.69 3.69
uo ≥0.4 0.7 0.72 0.75 0.36 0.59 0.76
ul ≥0.7 0.58 0.68 0.78 0.42 0.50 0.84
TI (%) ≤10 11 / / 8 7 5
SR 0.5 0.53 0.53 0.53 0.53 0.53 0.53
LPI (W/m) ≤7.1 15.19 30.45 22.82 6.37 12.81 9.59
LPI (W/m2 ) – 2.17 4.35 3.26 0.91 1.83 1.37
SL ≤0.674 0.616 0.613 0.46 0.500 0.495 0.371

Table 3
Comparison between illuminance of HPS and LED lamps for various arrangements in a road lighting system with a 3 m-wide of street isle (surfaces of road = Portland cement
concrete).

Data Standard HPS lamp LED lamp

Opposite on dual Staggered and Twin central Opposite on dual Staggered and Twin central
carriage way opposite carriage way opposite

Eav (lux) ≥21.5 43 43 46 32 32 24

Emin (lux) – 20 22 22 12 12 6.9
Emax (lux) – 79 75 82 54 53 48
Emin /Eav ≥0.4 0.465 0.512 0.478 0.38 0.38 0.29
Emin /Emax ≥0.167 0.258 0.293 0.271 0.22 0.13 0.14
Lav (cd/m2 ) ≥2 4.24 4.24 4.41 2.89 2.89 2.05
u0 ≥0.4 0.68 0.69 0.59 0.51 0.46 0.40
ul ≥0.7 0.61 0.69 0.62 0.60 0.54 0.41
TI (%) ≤10 10 10 11 9 6 7
SR 0.5 0.75 0.75 0.67 0.71 0.71 0.69
LPI (W/m) ≤12.3 12.53 9.38 12.53 5.25 3.92 5.25
LPI (W/m2 ) – 1.79 1.34 1.79 0.75 0.56 0.75
SL ≤0.674 0.422 0.317 0.406 0.260 0.195 0.367

gitudinal uniformity ratio of luminance (ul ) obtained from both Similar to the analysis above, the results of further analysis in
luminaires is less than that of the base case, except for the staggered the case of the street isle are shown in Table 3. The results shown
roadway arrangement. This result indicates that the longitudinal in Table 3 are presented for road surfaces of Portland cement con-
uniformity ratio of luminance (ul ) should be improved because uni- crete. According to the data presented in Table 3, considering the
formity of luminance is required to create visual comfort for the overall results of the LED luminaire, the average illuminance (Eav ),
driver. average luminance (Lav ), overall uniformity ratio of luminance (uo ),
Considering the threshold increment (TI), it can be observed that and the threshold increment (TI) are greater than those of the base
when the single-side arrangement is considered as an example, the case, and the energy efficiency index (LPI and SL) is significantly
threshold increment (TI) in the case of LED luminaires is less than decreased. However, considering the uniformity ratio of illumi-
the base case, whereas in the case of the HPS luminaire, this value nance, the longitudinal uniformity ratio of luminance (ul ) should
is greater than that of the base case, indicating that the HPS lumi- be improved.
naire should also be improved in terms of the threshold increment As a result, the LED luminaire plays an important role in energy
(TI). If the arrangement of the roadway is changed, the threshold conservation for roadway lighting, but the lighting quality should
increment (TI) of the LED luminaires is still less than the base case be improved. Although the results of the LED luminaire show values
for both roadway arrangement, indicating that installation of LED similar to those of the standard, this indicates a significant mis-
luminaires is beneficial to the threshold increment (TI). match between the lighting quality of the LED luminaire and the
For the lighting performance index (LPI), if all roadway arrange- standard roadway lighting. Generally, the data for standard road-
ments are considered, the LPI (W/m) in the case of the LED luminaire way lighting in Thailand were designed for conventional lamps.
is less than that of the HPS luminaire but has value greater than that However, when the LED lamp is used, the luminous flux of the LED
of the base case (roadway lighting standard of Thailand), except lamp is less than that of the conventional lamp, and thus the LED
for the single-side roadway arrangement, due to the impact of the lamp is not efficient for lighting quality.
number of luminaires. Considering the area of power consumption
(W/m2 ), the LPI in the case of LED luminaires is less than that in the 3. Experimental setup and results
case of the HPS luminaires by a factor of approximately 2.5 which
indicates that the utility benefits from using LED luminaires. As previously mentioned, the objective of this paper is to exam-
Examining the maximum energy efficiency indicator in terms of ine on the performance of LED roadway lighting luminaires. In this
luminance (SL), it can be observed that if all roadway arrangements section, electrical parameters and power quality measures such as
are considered, the SL of both luminaires is less than that of the base voltage, current, real power, frequency, power factor, percentage
case. However, the obtained SL for the LED luminaire is significantly of total harmonic distortion voltage (%THDv ) and current (%THDi )
less than that of the HPS luminaire. are measured and monitored using the experimental setup and
548 S. Yoomak et al. / Energy and Buildings 159 (2018) 542–557

Fig. 3. Experimental setup for HPS luminaire and LED luminaire.

electrical measuring instrument. The single-line diagram of the obtained from this point include current, voltage, power and power
experimental setup is shown in Fig. 3. The experimental setup is quality of both the LED and HPS luminaires. In the case of road-
used to evaluate the performance of two types of luminaire (LED way lighting with solar power shown in Fig. 3(b), the measurement
and HPS) in cases with and without solar power. points are described as followes. Measurement point 1 is located
The experimental setup in the case of a roadway lighting system between the photovoltaic simulator and solar charge controller to
without solar power is shown in Fig. 3(a). The system consists of a measure the generated power from the photovoltaic system. Mea-
power source from the laboratory (represented by number 1) with surement point 2 is located between the solar charge controller
a circuit breaker and fuse box (represented by number 2) before and battery to obtain data from the charging operation. Measure-
the connection to the roadway luminaires with LED (represented ment point 3 is located between the solar charge controller and
by number 3) and HPS (represented by number 4). Measurements inverter to evaluate the performance of the charge controller, and
for both case studies (HPS and LED cases) are collected between measurement point 4 is located between the inverter and the LED
the driver and power source. In terms of electrical parameters, and HPS luminaires. In addition, the first three measurement points
%THDi , and %THDv are measured with a Fluke 435 power quality (points 1–3) consist of direct current, and the last measurement
analyser (represented by number 5). Readings are logged every 5 s point (point 4) consists of alternating current.
and manually monitored every 5 min for stability during controlled The electrical parameters obtained from experimental setup in
environment room testing. In addition, the waveforms for current the case of HPS and LED without solar power are summarized in
and voltage are obtained using Wave Surfer 3000 oscilloscopes Table 4. In Table 4, the power consumption of the HPS luminaire
(represented by number 6). is 287 W with a 0.81 power factor, whereas the LED luminaire has
The experimental setup for the roadway lighting system with a lower power requirement with 121 W and a higher power factor
solar power and energy storage is shown in Fig. 3(b). The experi- at 0.97. The power consumption measured from the experimental
mental setup equipment is described as follows; 3-phase AC source setup is greater than the specification of the lamp due to the power
from the laboratory (represented by number 1) supplies power to loss of the ballast in case of HPS or the switching power supply in
the photovoltaic simulator through a protection device consisting the case of LED. The results show that the power quality of the LED
of a 3-phase 30 A circuit breaker and 3-phase 30 A fuse box (rep- luminaire in terms of the %THDv and %THDi are better than those of
resented by number 2). A photovoltaic simulator (represented by the HPS luminaire, indicating that with the LED luminaire, electrical
number 7) generates voltage and current, as modelled after selected energy can be conserved and fewer harmonics can be generated.
photovoltaic panel V–I characteristics from the simulation pro- The current and voltage waveforms obtained from the exper-
gramme (represented by number 8). Power generated from the imental setup using the Wave Surfer 3000 oscilloscopes, shown
photovoltaic simulator represents direct current that uses a solar in Fig. 3(a), are shown in Fig. 4 to evaluate the characteristics of
charger controller to control the need to charge power into the power quality in terms of waveform distortion. It can be clearly
energy storage system with a 3-phase 20 A fuse box (represented observed that the voltage waveforms of both the HPS and LED lumi-
by number 9) for the protection inverter (represented by number naires show little change, whereas the current waveform of the HPS
10). In the daytime, power produced by the photovoltaic panel is luminaire is significantly distorted from the sinusoidal beseline and
charged into a 12 Ah 24 V solar battery (represented by number 11) heavily distorted compared with the LED luminaire. This trend is
via a solar charge controller (represented by number 12) for storage also aligned with the percentage total harmonic distortion value
purposes. During the nighttime, the solar battery is used as a power presented in Table 4.
source to deliver power to the roadway luminaire. Direct current Electrical parameters such as voltage, current, and real power
from the battery is converted to alternating current by the inverter obtained over the night-time period (from sunset to sunrise or
(represented by number 8) for both the LED and HPS (represented 12 h) were measured using the Fluke 435 power quality analyser, as
by number 3 and 4, respectively). Moreover, for the LED luminaire, shown in Fig. 5. The experimental time used in this study has been
an A- to-DC converter is installed inside the luminaire because the accelerated to 20 times normal speed, thus reducing the night-time
LED module is driven by direct current. period to a 36-min period. From Fig. 5, it can be observed that volt-
Considering the measurement point of the experimental setup, age, current, and real power increase instantaneously in the case
in the case of a roadway lighting system without solar power shown of the LED luminaire when operating the experimental setup, and
in Fig. 3(a), the measurement point is located after the main cir- the value is constant across the time period. For the HPS luminaire,
cuit breaker (represented by the measurement point). The data the voltage level is steady across the time period, but current first
 S. Yoomak et al. / Energy and Buildings 159 (2018) 542–557 549

Table 4
Summary results from experimental setup in the case without solar power.

Measurement point Frequency (Hz) %THDv %THDi Vrms (V) Irms (A) Apparent power (VA) Real power (W) Reactive power (Var) Power factor

HPS lamp 50 1.8 16.4 225 1.6 351 287 197 0.81
LED lamp 50 1.6 13.7 223 0.56 125 121 −22 0.97

Fig. 4. Waveforms from experimental setup for LED lamp and HPS luminaire without solar power.

increases and subsequently decreases steadily to a curtain level. and power charging to the battery. At measurement point 1, the
The real power obtained from the HPS luminaire indicates that voltage generated from the photovoltaic simulator shows a spike
in the starting operation, the real power is lower and increases for a short period of time and a reduction to operation level across
steadily due to the ignition process of the HPS luminaire. A short the time period. In addition, the current and real power increase
period of time is required for the HPS luminaire to operate at full instantaneously, and the voltage and current fluctuate in a narrow
output. During that time period, reactive power is increased, result- range. For the period of sunset (at 28 min in Fig. 6 and Table 5)
ing in a decrease of the power factor value followed by a steady and solar irradiance reduction to below operating level, the volt-
regain as the HPS luminaire operates at full output. age is increased, but the current and power are decreased before the
Examining the measurement points of the experimental setup, photovoltaic panel stops generating power. Examining the electri-
in the case of the roadway lighting system with solar power shown cal parameter at measurement point 2, the battery (represented by
Fig. 3(b), the power quality measurement is divided into two peri- number 11 in Fig. 3(b)) is charged with constant voltage and steady
ods. During the daytime (from sunrise to sunset or 12 h), the decreases as the battery reaches full capacity. The charging current
electrical parameters were measured at point 1 and 2, which rep- depends on the power generated from the photovoltaic panel and
resents the power quality from charging with solar power (PV steadily decreases, similar to measurement point 1. This result indi-
simulator) and during the nighttimes (from sunset to sunrise or cates that the voltage and current can be controlled using the solar
12 h), the electrical parameters and power quality were measured charge controller (represented by number 10 in Fig. 3(b)) during
at point 1, 2, and 3, which represents the power quality comparison the charging operation.
using HPS and LED luminaires. Summary results for both cases of When the HPS luminaire is carefully considered based on
roadway lighting with solar power are shown in Tables 5 and 6. Table 6, it can be observed that the voltage from battery (measure-
When the electrical parameters during the daytime (charging ment at point 2) is approximately 23 VDC , the current is 14.38 A,
with solar power) are considered, as shown in Table 5, the sys- and the power is 339 W. In terms of voltage at each measure-
tem undergoes energy loss, as evidenced by comparing the real ment point, the voltages at point 2 and point 3 show little change,
power at point 1 and point 2. The energy loss usually occurs due but the voltage at point 4 increases significantly to approximately
to charger controller operation. Additionally, this system is imple- 220 VAC , which is the required voltage for the HPS lamp. Based
mented at 24 V with the maximum power point tracking (MPPT) on further analysis of Table 6 in terms of current, the current at
algorithm to control the voltage at which the system operates at point 3 is slightly decreased compared with point 2 and this change
the maximum power point on the power voltage curve, as shown has a value of approximately 0.8 A. However, the current at point
in Fig. 6. Fig. 6 shows that the time used in the experimental setup 4 decreases significantly to 1.61 A because the voltage at point 4
has been accelerated to 20 times normal speed, thus reducing a one- increases and the real power of the HPS lamp at point 4 decreases.
day period to a 36-min period. In Fig. 6, electrical parameters such Considering the real power in Table 6, the real power generated
as voltage, current, and real power with a period of one day (from from the battery through the solar charge controller (measurement
measurement point 1 and point 2 of the experimental setup) are at point 3) is 321 W, which drops approximately 18 W compared
used to evaluate the power generated from the photovoltaic panel with the real power at point 2 due to efficiency and power loss
550 S. Yoomak et al. / Energy and Buildings 159 (2018) 542–557

Fig. 5. Electrical parameters from experimental setup in the case of LED and HPS luminaires without solar power.

Table 5
Summary results from experimental setup in the case with solar power during the daytime.

Time (mins) Electrical parameters at Point 1 Electrical parameters at Point 2

Voltage (V) Current (A) Real Power (W) Voltage (V) Current (A) Real Power (W)

0 0.38 0.1 0 24.5 0 0

4 35.6 7.5 266 27.1 9.3 253
8 36.5 7.2 264 27.2 9.2 250
12 35.7 7.5 266 27.4 9.2 253
16 36.0 7.4 266 27.6 9.1 252
20 36.0 7.4 266 27.8 9.1 252
24 36.4 7.3 265 28.1 8.9 251
28 37.8 6.7 255 28.6 8.4 241
32 41.6 3.7 154 28.4 5.0 141
36 42.7 2.3 96 28.6 2.9 83

Table 6
Electrical parameters and power quality in the case roadway lighting with solar power during the night-time.

DC systems AC systems

Power quality at point 2 Power quality at point 4

parameter HPS luminaire LED luminaire Parameter HPS luminaire LED luminaire
Vrms (V) 22.98 23.7 Vrms (V) 222.1 224.2
Irms (A) 14.83 6.3 Irms (A) 1.61 0.548
Real power (W) 339 149 Real power (W) 282 121
Power quality at point 3 Reactive power (Var) 220 −29.15
Parameter HPS luminaire LED luminaire Apparent power (VA) 358 124.4
Vrms (V) 23.02 23.6 Power factor 0.788 0.972
Irms (A) 14.06 5.8 %THDv 0.9 1.045
Real power (W) 321 138 %THDi 21.4 17.5
 S. Yoomak et al. / Energy and Buildings 159 (2018) 542–557 551

Fig. 6. Electrical parameters from experimental setup in the case of solar power charging (during the daytime).

in the solar charge controller. Moreover, the real power supply to sured using the Fluke 435 power quality analyser and are shown
the HPS luminaire through the inverter (measurement at point 4) in Figs. 7–9. The current and voltage waveforms of each luminaire
is 282 W with a voltage near 220 V according to the voltage level were recorded using the Wave Surfer 3000 oscilloscopes and are
in Thailand, and its change has a value of approximately 39 W due shown in Fig. 10. The obtained electrical parameter from the output
to efficiency and power loss in the inverter. In addition, the power of the battery (measurement point 2) was considered, as shown in
factor is 0.788, indicating that the electrical parameters in the case Fig. 7. For the LED luminaire, it can be observed that voltage beings
of roadway lighting with and without solar power are minimally at a higher value during the short period of time without current
impacted but the power quality in terms of %THDi has a greater and subsequently decreases to the operating level. The current and
value of 21.4%. real power increase instantaneously to the operating level, and this
Based on further analysis for the case of the LED luminaire in value is constant over the night-time period. The HPS luminaire has
Table 6, the results show that for the electrical parameter at point 2, a trend in voltage similar to that of the LED luminaire but with a
the battery voltage is 23.7 V, the current is 6.3 A, and the real power longer time period. The current and real power increase instan-
is 149 W. In terms of voltage at each measurement point, it can be taneously and steadily increase as the HPS luminaire ignites the
observed that the voltage shows the same behaviour as in the case high-pressure sodium gases inside until the lamp is fully lit.
of the HPS luminaire. Considering another electrical parameter, the Similar to the analysis above, as shown in Fig. 8, the electrical
current has the same behaviour as in the case of the HPS luminaire parameter from the output of the solar charge controller (measure-
but the value of the current in the case of the LED luminaire is less ment point 3) is considered. In the case of the LED luminaire, a trend
than that of the HPS luminaire. In addition, the real power tends to similar to that of measurement point 2 can be observed with a con-
decrease, as does the behaviour of the current. The obtained real stant value over the period of night-time, but the current and real
power after the solar inverter shows that the real power at point power at measurement point 3 are less than that of measurement
3 is 138 W, which is reduced from the battery (point 2) by approx- point 2 due to power loss and efficiency in the charger controller
imately 11 W, and after the inverter (point 4), the real power is equipment. Additionally, in the case of the HPS luminaire, the result
decreased to 121 W. The power factor in the case of the LED lumi- of measurement point 3 is similar to that of measurement point 2.
naire shows little change, which also indicates the same trend as The current and real power steadily increase during the luminaire
the HPS luminaire. In terms of power quality, the LED luminaire ignition process, and the voltage is decreased to a certain level as
does not affect the %THDv with a value approximately of 1.045%, the HPS luminaire operates at full output.
whereas the %THDi value is 17.5%. Based on further analysis for the case of roadway lighting with
During the night-time, the electrical parameter obtained from solar power, Fig. 9 shows the electrical parameters from the output
each measurement point of the experimental setup were mea- of the inverter at measurement point 4. In the case of the LED lumi-
552 S. Yoomak et al. / Energy and Buildings 159 (2018) 542–557

Fig. 7. Electrical parameters from experimental setup at measurement point 2 in the case of power from battery.

naire, the results show that the voltage, current, and real power been distorted from the sinusoidal shape due to noise in the signal.
are constant over the period of time. In the case of the HPS lumi- The HPS luminaire with solar power causes more distortion than
naire, the result shows that the voltage is approximately constant the LED luminaire, which is also in agreement with the data from
at 220 V over the period of time, and the current is higher than the Table 6. However, for the power quality in the case of solar power,
normal condition during a short period of time during the luminaire it can be clearly observed that the obtained %THDi for each lumi-
ignition process. However, the real power is less than the normal naire has a greater value than that of the power quality in case of
condition at first but steadily increases because the reactive power no solar power, as shown in Table 6 and Fig. 10.
is increased with the ignition process resulting in a lower power
factor in that period. In addition, at full operation, the real power 3.1. Payback period
is lower than the two previous measurement points (point 2 and
point 3) due to the power loss in the ballast of the HPS luminaire HPS luminaires are relatively inexpensive because the technol-
and inverter. ogy in the market is well past maturity. However, LED luminaires
Considering the power quality in the case of roadway lighting for street lighting have only been on the market for a few years. The
with solar power, when both luminaires are carefully compared, it initial cost of LED-based general illumination sources is too high
is clear that the power consumption of the HPS luminaire shows compared with conventional lighting technologies. Prices remain
a greater change than that of the LED luminaire by a factor of high due to the smaller scale of production and due to a high
approximately 2. Therefore, the HPS luminaire requires a battery demand for LEDs in other applications, such as automotive and dis-
with larger capacity to cover the usage for an entire night. The LED plays. Because the lighting market has historically been strongly
luminaire has a better power factor because the components inside affected by first costs (lifetime benefits notwithstanding), low-cost
the LED luminaire consist mainly of power electronics. In terms of LED packaging and luminaire materials are needed as well as low-
power quality, if the %THDi and %THDv are considered, the obtained cost, high-volume, and reliable manufacturing methods. The typical
power quality from the LED luminaire is better than that of the cost for a LED street lighting luminaire was upwards of $230 USD
HPS luminaire. Fig. 10 shows the waveforms for current and volt- just a few years ago and is steadily dropping each year. However,
age obtained from an oscilloscope in the experimental setup in the the cost of a LED luminaire is up to five times higher than the cost
case of LED and HPS luminaires with solar power. It can be observed of a HPS luminaire. These costs are based on small-quantity orders,
that both types of luminaires do not have much effect on the volt- but for larger quantities, the costs may be lower. In addition, the
age waveform. The current waveform shows that the waveform has installation costs for HPS and LED luminaires are the same. How-
 S. Yoomak et al. / Energy and Buildings 159 (2018) 542–557 553

Fig. 8. Electrical parameters from experimental setup at measurement point 3.

ever, this value does not include any overhead incurred by cities In this section, the Discounted Payback Period (DPP) and Internal
and municipalities. Rate of Return (IRR) criteria are investigated. The DPP is simply the
A total cost of ownership comparison shows the cost-saving period during which the cumulative net present value of a project
potential for customer-owned equipment. When considering the is equal to zero. If DPP is less than its economic life or another
equipment and installation costs, the customer-owned costs show predetermined period, the project is acceptable [59]. The IRR is the
that the LED cost is close to matching that of HPS. This assess- discount rate at which the net present value of future cash flows
ment also assumes that the LED luminaire operates without any from an investment is equal to zero. The higher the IRR rates, the
maintenance for that same period of time. The tariff structure is higher the returns and the faster payback will occur. Thus, both
another factor to consider together with the energy costs. Although DPP and the IRR are used to assess the economic feasibility of the
energy savings of 40% are possible, the light output of the tested LED roadway lighting system.
luminaires does not meet the average illuminance offered by the Using the case studies, the DPP and IRR were calculated to eval-
HPS. If using LED luminaires with increased light output, meeting uate the investment for the LED luminaire, and the 250 W HPS
the HPS luminaires performance is possible, but the energy sav- luminaire was considered as the base case for the 120 W LED lumi-
ing is reduced and luminaire costs are increase. These factors affect naires. The periods of the project are set to ten years, and the
the overall savings analysed using the payback time. The payback discount rate is 5%. The installation costs are assessed for one kilo-
time considers all related costs required to operate and maintain metre with a single-sided arrangement (without isle). The case
a luminaire throughout a given time period. As a result, the pay- study is divided into three case studies. The first case considers the
back time is the time necessary for a LED customer to break even HPS luminaire with 25 poles (the distance between poles is 40 m).
on his/her investment in a more expensive LED bulb. Generally, The second case considers the LED luminaire with 25 poles (the dis-
high energy prices decrease the time to payback for more efficient tance between poles is 40 m) and the final case considers the LED
bulbs. Customers have a 10% probability of adopting a technology luminaire with 33 poles to improve lighting quality such that the
if the payback is two years, a 30% probability if the payback is one distance between poles is decreased to 30 m. The summary of the
year, and a 40% probability if the payback is six months. Therefore, results is presented in Tables 7–9.
the time-to-payback metric aids in predicting LED adoption with a In Table 7, the lighting installation and maintenance costs
certain level of historical and empirical accuracy. The observation were considered to estimate the installation equipment cost (not
that payback periods are shrinking together with LED retail prices including the luminaires) for the HPS luminaire case and the LED
is expected to ensure significant market penetration (50%+) in the luminaire case. The prices for each equipment case were denoted
next decade. by the Comptroller General’s Department of Thailand [60], and
554 S. Yoomak et al. / Energy and Buildings 159 (2018) 542–557

Fig. 9. Electrical parameters from experimental setup at measurement point 4.

Fig. 10. Waveforms from experimental setup for LED lamp and HPS luminaire with solar power.

each unit of the systems includes luminaires, light poles, power maintenance costs per luminaire per year and electricity price.
lines, ground rods, and other components. The maintenance and Additionally, the energy consumption cost of roadway lighting sys-
operation costs include operation cost per luminaire, estimated tems was calculated from the eergy consumption during one year at
 S. Yoomak et al. / Energy and Buildings 159 (2018) 542–557 555

Table 7 the results clearly show that the case of the LED luminaire with
Details of installation and maintenance costs for roadway lighting systems.
33 poles is more expensive than that of the LED luminaire with 25
General List Costs poles because the amount of equipment (such as wires, grounding
Roadway lighting installation [002] system, etc.) tends to increase with increasing numbers of lumi-
LED 120 W luminaire $396.86/luminaire naires and light poles. However, according to Table 8, the annual
HPS 250 W luminaire $158.10/luminaire savings or cash flow savings per year is estimated using the data
9-m light pole $288.84/luminaire from Table 7, including operation costs, maintenance costs, and
IEC10 wire 4*10 SQ mm (between pole distances) $6.38/metre
electricity price. The results show that by increasing the number
Ground rod for grounding system $12.16/luminaire
Other (e.g., fuse and relay) $111.58/luminaire of light poles and luminaires, the annual saving is decreased for the
case of the LED luminaire with 33 poles compared with the annual
Maintenance and Operation [003]
Operation costs $21.28/luminaire
savings from the case of the LED luminaires with 25 poles. This
Estimated maintenance costs HPS per luminaire $24.44/year result indicates that the number of light poles and luminaires are
Estimated maintenance costs LED per luminaire $7.78/year an important parameter for determining the payback period.
Electricity price $0.09/kW h After the installation costs and annual savings were calculated,
as previously mentioned, the Discount Payback Period (DPP) and
Table 8
the Internal Rate of Return (IRR) arer estimated to evaluate the
Total installation and annual energy-saving costs for HPS and LED luminaires along investment for roadway lighting with LED luminaires. The data
the length of the roadway (1000 m). from Table 8 were used to estimate the DPP using Eqs. (1) and (2),
Case Studies Installation Incremental Annual
and the obtained results are presented in Table 9. The data in Table 9
Costs Costs Saving (cash show that the DPP is one year when the cumulative present values
flow) of cash flow are zero such that the case of LED luminaires with 33
HPS 250 W 25 poles $ 21,179 – – poles yields a DPP of eight and a half years, whereas the case of LED
LED 120 W 25 poles $ 27,148 $5969 $1917 luminaires with 25 poles gives a DPP of three and quarter years.
LED 120 W 33 poles $ 33,794 $12,615 $1494 Thus, both cases of LED luminaires are acceptable because the DPP
is less than ten years. However, considering the IRR, it can be clearly
observed that the case of LED luminaire with 25 poles is increased
approximately 4200 h and an electric retail rate of $0.09/kW h. The to 30% after a ten-year project. As a result, the case of LED lumi-
maintenance costs were estimated based on material and reported naires with 25 poles is economically feasible after considering the
labour with reference to [61]. The obtained data show that com- DPP and the IRR. Because both the DPP and the IRR have the same
pared with the HPS luminaire, the roadway lighting system with the conclusion this case study is feasible for installation in roadway
LED luminaires has higher installation costs (approximately 2 times lighting systems to replace the inefficient conventional system. In
as large) due to the cost of the LED luminaire but the LED luminaire addition, the case of LED luminaires with 33 poles is modified by
can save maintenance costs per year by a factor of approximately decreasing the distance between light poles to improve lighting
3. quality, but the initial installation costs tend to increase such that
Based on further analysis of the total costs for the HPS and LED the DPP is greater than the other value, indicating that the case of
luminaires along the length of a roadway of 1000 m, as shown LED luminaire with 33 poles is not acceptable because the IRR is
in Table 8, the installation costs were calculated using the data approximately 3%, which is lower than the other case.
obtained from Table 7, including luminaires, light poles, wires,
grounding system, and other components but not including the Cash flow
Discounted Cash Flow = (1)
maintenance and operation costs. The results show that the HPS (1 + i)
n
luminaires have an installation cost of $21,000 USD, which is lower
than those of both LED luminaire cases ($27,000 and $34,000 A
Discounted Payback Period = T + (2)
USD for LEDs with 25 poles and LEDs with 33 poles, respectively) D
because the price of the LED luminaire is approximately 2.5 times
higher than that of the HPS luminaire. Thus, the case of the LED where
luminaire with 33 poles can improve the lighting quality, but the i – Discount rate;
installation costs increase by increasing the number of light poles n – Period of cash flow;
and luminaires. To support this assertion, considering the incre- T – Period when the cumulative discount is negative;
mental costs in Table 8, the difference between the installation A – Absolute value of the cumulative discount at the period T;
costs for HPS luminaire and LED luminaires was calculated, and D – Discounted cash flow during the period after T

Table 9
Discount Payback Period (DPP) and the Internal Rate of Return (IRR) of roadway lighting systems along the length of roadway (1000 m).

Period Present Value Cash flow ($) Present Values of Cumulative Present Values IRR
factor (i) at 5% Cash Flow ($) of Cash Flow ($)

n (year) PV = 1/(1 + i)n LED 25 poles LED 33 poles LED 25 poles LED 33 poles LED 25 poles LED 33 poles LED 25 poles LED 33 poles

0 1 −5969 −12,615 −5969 −12,615 −5969 −12,615 – –

1 0.9542 1917 1494 1829 1426 −4140 −11,189 −68% −88%
2 0.9070 1917 1494 1739 1355 −2401 −9834 −25% −59%
3 0.8638 1917 1494 1656 1291 −745 −8544 −2% −38%
4 0.8827 1917 1494 1692 1319 947 −7225 11% −24%
5 0.7835 1917 1494 1502 1171 2449 −6055 18% −15%
6 0.7462 1917 1494 1430 1115 3879 −4940 23% −9%
7 0.7107 1917 1494 1362 1062 5242 −3878 26% −4%
8 0.6768 1917 1494 1297 1011 6539 −2867 28% −1%
9 0.6446 1917 1494 1236 963 7775 −1904 29% 1%
10 0.6139 1917 1494 1177 917 8952 −987 30% 3%
556 S. Yoomak et al. / Energy and Buildings 159 (2018) 542–557

4. Conclusions [4] Yi Jiang, Shuo Li, Bowen Guan, Guangyuan Zhao, Cost effectiveness of new
roadway lighting systems, J. Traffic Transp. Eng. 2 (3) (2015) 158–166.
[5] Marco Beccali, Marina Bonomolo, Giuseppina Ciulla, Alessandra Galatioto,
The objective of this paper was to determine the feasibility of Improvement of energy efficiency and quality of street lighting in South Italy
replacing HPS cobra-head style luminaires with LED luminaires to as an action of Sustainable Energy Action Plans. The case study of Comiso
achieve lighting quality, energy savings, power quality, and invest- (RG), Energy (2015).
[6] Dusko Radulovic, Srdjan Skok, Vedran Kirincic, Energy efficiency public
ment targets. However, when replacing an existing HPS luminaire lighting management in the cities, Energy 36 (4) (2011) 1908–1915.
on a one-for-one basis, several parameters must be considered. The [7] Anais Mayeur, Roland Bremond, J.M. Christian Bastien, The effect of the
replacement luminaire should meet or exceed the performance of driving activity on target detection as a function of the visibility level:
implications for road lighting, Transp. Res. 13 (2) (2010) 115–128.
the incumbent technology, including overall light output, distri-
[8] A. Pena-Garcia, A. Hurtado, M.C. Aguilar-Luzon, Impact of public lighting on
bution and quality. The LED luminaires tested in this paper show pedestrians’ perception of safety and well-being, Saf. Sci. 78 (2015) 142–148.
a wide range of energy-savings potential, with energy consump- [9] Miomir Kostic, Lidija Djokic, Recommendations for energy efficient and
visually acceptable street lighting, Energy 34 (10) (2009) 1565–1572.
tion as a function of their performance. This paper examines the
[10] Alejandro Sanchez de Miguel, Jaime Zamorano, Jose Gomez Castano, Sergio
three major aspects more deeply with respect to the feasibility of Pascual, Evolution of the energy consumed by street lighting in Spain
luminaire replacement. The results can be summarized as follows: estimated with DMSP-OLS data, J. Quant. Spectrosc. Radiat. Transf. 139 (2014)
109–117.
[11] M. Indra al Irsyad, Rabindra Nepal, A survey based approach to estimating the
- The lighting quality and LPI were considered using a simulation
benefits of energy efficiency improvements in street lighting systems in
programme. The results indicated that for the lighting qual- Indonesia, Renew. Sustain. Energy Rev. 58 (May) (2016) 1569–1577.
ity with the LED luminaire, average illuminances were reduced, [12] Christine Boomsma, Linda Steg, The effect of information and values on
minimum illuminances were maintained, and uniformity was acceptability of reduced street lighting, J. Environ. Psychol. 39 (2014) 22–31.
[13] John D. Bullough, Eric T. Donnell, Mark S. Rea, To illuminate or not to
improved. The LED luminaires showed significant energy efficient illuminate: roadway lighting as it affects traffic safety at intersections, Accid.
index reductions of approximately 40% over the HPS luminaire Anal. Prev. 53 (2013) 65–77.
indicating that the utility benefited from using LED luminaire. [14] A. Jafari Anarkooli, M. Hadji Hosseinlou, Analysis of the injury severity of
crashes by considering different lighting conditions on two-lane rural roads, J.
- The electrical parameters and power quality were considered Saf. Res. 56 (2016) 57–65.
using an experimental setup. The experimental setup was used to [15] Cleber Costa da Fonseca, Rodrigo Palucci Pantoni, Dennis Brandao, Public
evaluate the performance of two types of luminaire LED and HPS street lighting remote operation and supervision system, Comp. Stand.
Interfaces 38 (2015) 25–34.
in the cases with and without solar power. The obtained results [16] Mohammad Javad Mirzaei, Reza Dashti, Ahad Kazemi, Mohammad Hassan
from both cases shown that the LED luminaire performs better Amirioum, An asset-management model for use in the evaluation and
than the HPS luminaire in terms of power quality and energy regulation of public-lighting systems, Util. Policy 32 (2015) 19–28.
[17] Michael Jackett, William Frith, Quantifying the impact of road lighting on
savings.
road safety—a New Zealand study, IATSS Res. 36 (2013) 139–145.
- The payback period was also considered. The initial costs of LED [18] Katja Kircher, Christer Ahlstrom, The impact of tunnel design and lighting on
luminaires are still quite high compared with HPS luminaires. the performance of attentive and visually distracted drivers, Accid. Anal. Prev.
47 (2012) 153–161.
However, the case of LED luminaires gives a DPP of three and
[19] L.M. Gil-Martin, A. Pena-Garcia, A. Jimenez, E. Hernandez-Montes, Study of
a half years, whereas the IRR is 30% for a ten-year project life. light-pipes for the use of sunlight in road tunnels: from a scale model to real
The results show that the current cost of LED luminaires is one tunnels, Tunn. Undergr. Space Technol. 41 (2014) 82–87.
of the major factors in determining total cost savings over HPS. [20] S. Yoomak, A. Ngaopitakkul, Study and analysis on lighting energy
management for highway, 7th International Exergy, Energy and Environment
The market low for larger quantities at this time results in lower Symposium (2015).
LCCs than that of HPS if compared over a 50,000-h life. However, [21] M.H. Omar, H. Abdul Rahman, M.S. Majid, M.Y. Hassan, N. Rosmin, The
the higher initial costs might still be a barrier. However, the short reduction of total harmonic distortion and electromagnetic interference in
high pressure sodium street lighting using single stage electronic ballast, IEEE
duration of the testing did not lend itself to revealing any sig- International Power Engineering and Optimization Conference (PEOCO)
nificant depreciation due to normal operation, and many other (2012) 230–235.
factors might be involved in determining the life of the product. [22] Fernando B. dos Reis, Julio Cesar Marques de Lima, Fernando S. dos Reis,
einaldo Tonkoski Jr, Development of a flexible public lighting system, 39th
Annual Conference of the Industrial Electronics Society (IECON) (2013)
Finally, the LED luminaire plays an important role in the energy 6046–6051.
savings in roadway lighting, but the lighting quality should be [23] L.F.F. Rakotomalala, Z. Randriamanantany, G. Chiriac, D.D. Lucache, HID lamps
dimming in the public lighting installation dominated by magnetic ballasts,
improved. To overcome this problem, the results obtained from
International Conference on Applied and Theoretical Electricity (ICATE)
the analysis are useful for improvement in the acceptable value of (2014) 1–5.
standard roadway lighting in Thailand in the future. Further work [24] Lovasoa F.F. Rakotomalala, R. Zely, E. Danila, D.D. Lucache, Fadel C.M. Kebe,
High energy saving for Africa street lighting using individual automatic
will consider the optimal mounting height of LED luminaires and
dimming of HID ballasts, International Conference and Exposition on
spacing between poles as a function of the width of the roadway to Electrical and Power Engineering (EPE) (2014) 683–688.
increase the average illumination and uniformity for each roadway [25] K. Hyodhyad, K. Supanaroj, Energy saving project for street lighting of
arrangement. Moreover, during the daytime, the grid-connected Provincial Electricity Authority (PEA), 2nd Joint International Conference on
Sustainable Energy and Environment (SEE2006) (2006) 1–6.
solar power to the utility will be considered in terms of power qual- [26] Ahmed A. Mansour, Osama A. Arafa, Comparative study of 250 W high
ity and investment such that the utility benefits from use of solar pressure sodium lamp operating from both conventional and electronic
power. ballast, J. Electr. Syst. Inf. Technol. 1 (2014) 234–254.
[27] A. Burgio, D. Menniti, A novel technique for energy savings by dimming high
pressure sodium lamps mounted with magnetic ballasts using a centralized
Acknowledgement system, Electr. Power Syst. Res. 96 (2013) 16–22.
[28] W. Nsibi, M. Nehdi, A.J. Chammam, A. Sellami, G. Zissis, Dimmable electronic
ballast for HPS lamp operating in LF, in: 2016 7th International Renewable
The authors wish to gratefully acknowledge financial support for Energy Congress (IREC), Hammamet, Tunisia, 2016, pp. 1–4.
this research (No. KREF045611) from the King Mongkut’s Institute [29] M.N. Nehdi, W. Nsibi, A. Chammam, A. Sellami, G. Zissis, Frequency dimmable
of Technology Ladkrabang Research fund, Thailand. electronic ballast for a 250W HPS lamp, in: 2016 7th International Renewable
Energy Congress (IREC), Hammamet, 2016, pp. 1–3.
[30] N. Khan, N. Abas, Comparative study of energy saving light sources, Renew.
References Sustain. Energy Rev. 15 (1) (2011) 296–309.
[31] W.R. Ryckaert, K.A.G. Smet, I.A.A. Roelandts, M. Van Gils, P. Hanselaer, Linear
[1] Provincial Electricity Authority, Corporate Social Responsibility 2012, 2012. LED tubes versus fluorescent lamps: an evaluation, Energy Build. 49 (2012)
[2] Metropolitan Electricity Authority, Annual Report 2012, 2012. 429–436.
[3] Energy Policy and Planning office (EPPO), Ministry of Energy, Energy Statistics [32] M.M. Aman, G.B. Jasmon, H. Mokhlis, A.H.A. Bakar, Analysis of the
of Thailand, 2014. performance of domestic lighting lamps, Energy Policy 52 (2013) 482–500.
 S. Yoomak et al. / Energy and Buildings 159 (2018) 542–557 557

[33] E. Tetri, A. Sarvaranta, S. Syri, Potential of new lighting technologies in [48] Navin Kumar, Smart and intelligent energy efficient public illumination
reducing household lighting energy use and CO2 emissions in Finland, Energy system with ubiquitous communication for smart city, International
Effic. 7 (2014). Conference on Smart Structures & Systems (ICSSS-2013) (2013) 152–157.
[34] Mats Bladh, Energy efficient lighting meets real home life, Energy Effic. 4 [49] V.C. Bender, F.B. Mendes, T. Maggi, M.A. Dalla Costa, T.B. Marchesan, Design
(2011) 235–245. methodology for street lighting luminaires based on a photometrical analysis,
[35] M.F. Braga, F.J. Nogueira, M.F.C. Campos, L.H.B. Gouveia, H.A.C. Braga, A Power Electronics Conference (COBEP) (2013) 1160–1165.
comparative study regarding linear fluorescent and LED lamps for indoor [50] Aurora. Gil-De-Castro, Math Bollen, Antonio Moreno-Munoz, Street lamps
lighting, 2014 11th IEEE/IAS International Conference on Industry aggregation analysis throught IEC 61000-3-6 approach, 22nd International
Applications (INDUSCON) (2014) 1–7. Conference on Electricity Distribution (2013) 633–637.
[36] Byung-Lip Ahn, Cheol-Yong Jang, Seung-Bok Leigh, Seunghwan Yoo, Hakgeun [51] W. Chen, S.N. Li, S.Y.R. Hui, A comparative study on the circuit topologies for
Jeong, Effect of LED lighting on the cooling and heating loads in office offline passive light-emitting diode (LED) drivers with long lifetime & high
buildings, Appl. Energy 113 (2014) 1484–1489. efficiency, IEEE Energy Conversion Congress and Exposition (ECCE) (2010)
[37] Byung-Lip Ahn, Seunghwan Yoo, Jonghun Kim, Hakgeun Jeong, Seung-Bok 724–730.
Leigh, Cheol-Yong Jang, Thermal management of LED lighting integrated with [52] Farah Ramadhani, Kamalrulnizam Abu Bakar, Muhammad Gary Shafer,
HVAC systems in office buildings, Energy Build. (2016). Optimization of standalone street light system with consideration of lighting
[38] Jingyu Liu, Wen Zhang, Xiaodong Chu, Yutian Liu, Fuzzy logic controller for control, International Conference on Technological Advances in Electrical,
energy savings in a smart LED lighting system considering lighting comfort Electronics and Computer Engineering (TAEECE) (2013) 583–588.
and daylight, Energy Build. 127 (September) (2016) 95–104. [53] Chun-An Cheng, Tsung-Yuan Chung, Fu-Li Yang, A single-stage LED driver for
[39] Su-Chin Huang, Li-Ling Lee, Ming-Shan Jeng, Yao-Ching Hsieh, Assessment of street-lighting application with high PF, IEEE International Symposium on
energy-efficient LED street lighting through large-scale, International Industrial Electronics (ISIE) (2013) 1–4.
Conference on Renewable Energy Research and Applications (ICRERA) (2012) [54] Chun-An Cheng, Chien-Hsuan Chang, Fu-Li Yang, Tsung-Yuan Chung, A novel
1–5. single-stage high-power-factor LED driver for street-lighting applications,
[40] Devesh Singh, Chandrajit Basu, Merve Meinhardt-Wollweber, Bernhard Roth, 10th IEEE International Conference on Power Electronics and Drive System
LEDs for energy efficient greenhouse lighting, Renew. Sustain. Energy Rev. 49 (PEDS) (2013) 330–333.
(2015) 139–147. [55] Hugo Andres Macias, Y. Ulianov, Yesid Ramos, Illumination Benefits using LED
[41] Antal Haans, Yvonne A.W. de Kort, Light distribution in dynamic street high brightness bulb compare to traditional illumination systems, IEEE
lighting: two experimental studies on its effects on perceived safety, International Symposium on Alternative Energies and Energy Quality (SIFAE)
prospect, concealment, and escape, J. Environ. Psychol. 32 (2012) 342–352. (2012) 1–5.
[42] Sohel Uddin, Hussain Shareef, Azah Mohamed, Power quality performance of [56] Jin Hu, Zhengyu Lu, Huipin Lin, Fengwu Zhou, A novel HPS and LED
energy-efficient low-wattage LED lamps, Measurement 46 (2013) 3783–3795. compatible driver with flexible resonnant tank, International Electronics and
[43] Chun-An Cheng, Chien-Hsuan Chang, Tsung-Yuan Chung, Fu-Li Yang, Design Application Conference and Exposition (PEAC) (2014) 892–897.
and implementation of a single-stage driver for supplying an LED [57] N. Yoshiura, Y. Fujii, N. Ohta, Smart street light system looking like usual
street-lighting module with power factor corrections, IEEE Trans. Power street lights based on sensor networks, 13th International Symposium on
Electron. 30 (2) (2015) 956–966. Communications and Information Technologies (ISCIT) (2013) 633–637.
[44] Fusheng Li, Dahua Chen, Xianjie Song, Yuming Chen, LEDs: a promising [58] Electricity Generating Authority of Thailand, Corporate Social Responsibility
energy-saving light source for road lighting, Asia-Pacific Power and Energy 2014, 2014.
Engineering Conference (APPEEC) (2009) 1–3. [59] Shyam B. Bhandari, Discounted payback period—a viable complement to net
[45] D. Camponogara, G.F. Ferreira, A. Campos, M.A. Dalla Costa, Off-line LED driver present value for projects with conventional cash flows, J. Cost Anal.
for street lighting with an optimized cascade structure, IEEE Industry (Summer) (1989) 43–53.
Applications Society Annual Meeting (IAS) (2012) 1–6. [60] The Comptroller General’s Department (CGD) (no date). Available at: http://
[46] D. Camponogara, G.F. Ferreira, A. Campos, M.A. Dalla Costa, J. Garcia, Offline www.cgd.go.th/cs/internet/internet/Contact Us.html?page locale=en US.
LED driver for street lighting with an optimized cascade structure, IEEE Trans. (Accessed 6 October 2016).
Ind. Appl. 49 (6) (2013) 2437–2443. [61] Final Report prepared in support of the U.S. DOE Solid-State Lighting
[47] Andras Poppe, Gabor Molnar, Tamas Temesvolgyi, Temparature dependent Technology Demonstration Gateway Program and PG&E Emerging
thermal resistance in power LED assemblies and a way to cope with it, 26th Technologies Program, LED Street Lighting, City of San Francisco, California.
Annual IEEE Semiconductor Thermal Measurement and Management
Symposium (SEMI-THERM) (2010) 283–288.