Energy and Buildings 93 (2015) 1622

Contents lists available at ScienceDirect

Energy and Buildings

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

Technical and economic evaluation of uorescent and LED luminaires

in underground mining. A case study: New mine level of El Teniente
Francisco Chueco , Fernando Lpez, Miguel Bobadilla
Jacobs Engineering, Apoquindo 5400, Las Condes, Santiago de Chile

a r t i c l e

i n f o

Article history:
Received 12 December 2014
Received in revised form 6 February 2015
Accepted 7 February 2015
Available online 16 February 2015
Keywords:
Lighting
LED
Fluorescent
Underground mining
Energy efciency

a b s t r a c t
This paper establishes the methodology used for lighting system planning in underground mining, shows
local data and lighting levels required by the mining industry, and examines and compares the results of
the technical and economic evaluation of using uorescent and LED luminaires in underground mining
in Chile. A case study for a new expansion of the mine called New Mine Level of El Teniente, Codelco,
the biggest underground copper mine in the world, is analyzed, where a large amount of luminaires are
used to illuminate work and transit areas in the mine. Technical comparison among different brands
and both technologies is performed using DIALux software. Cost-effectiveness and sensitivity analysis is
calculated using prices and economic data from this case study.
2015 Elsevier B.V. All rights reserved.

1. Introduction
In the last decades, underground mining has been widely illuminated by uorescent tubes; other technologies such as metal-halide
and high pressure sodium light xtures are also used to illuminate working areas, but currently, uorescent tubes are the most
frequently used in facilities, tunnels, and drifts.
At present, the project New Mine Level El Teniente, Codelco
[1], the biggest copper underground mine in the world, is under
expansion to replace the current mine production with 137,000
tons per day (t/d). This new infrastructure will keep the mine operations for the next 50 years. The expansion will include different
levels and facilities. Undercut level, Production level, Haulage level,
Crushing, Transportation level, Ventilation, and Drainage levels will
require facilities like civic quarters where there are dressing and
locker rooms, shower rooms, food preparation and dining rooms,
ofces and medical-attention areas (45,000 m2 ), workshops for
trucks and other machinery (100,000 m2 ), concrete plants, extraction and discharge points, loading stations, tunnels and drifts,
power stations, and electrical and communication rooms.
Thousands of people are working in this mine every day; this
environment is very complex to illuminate due to underground
mine conditions. At the mine, the building environment for civic

Corresponding author. Tel.: +56 22 9246239.

E-mail addresses: Francisco.Chueco@jacobs.com, franciscochueco@gmail.com
(F. Chueco).
http://dx.doi.org/10.1016/j.enbuild.2015.02.022
0378-7788/ 2015 Elsevier B.V. All rights reserved.

quarters is largely similar to buildings on the land surface. On the

other hand, there are workshops, extraction and discharge mineral
points, and tunnels where there are dust, moving machinery, transiting cars and buses, and other hazards [2,3]. For these reasons,
it is very important to have the right level of illumination, a low
rate of maintenance and reposition of luminaires, and emergency
illumination to ensure lighting of escape routes during crises.
Energy cost has increased dramatically in the last decade in Chile
[4,5]. The Global Energy Architecture Performance Index places
Chile as the 13th country with the highest industrial energy cost,
averaging 150 USD/MWh [6]. Energy efciency must be considered
to cut down on cost and it is necessary to analyze new technologies
and innovations to reduce electricity usage in the mines.
In the last years, light emitting diode (LED) technology has been
developed industrially by many brands offering this technology in
the market. The aim of this paper is to investigate if LED luminaires
are able to ensure the right level of illumination in a cost-effective
way versus uorescent tubes used until now in the mines.
2. Case study
The project called New Mine Level is currently under construction. As part of the engineering work, an electrical and lighting
system was designed for the new mine. Big quantities of luminaires are required to illuminate the new expansion properly.
All luminaires must be able to icker instantly because the mine
uses emergency signals through a lighting system to communicate re and other emergencies with its employees. Traditionally,

F. Chueco et al. / Energy and Buildings 93 (2015) 1622

17

Fig. 1. Lighting arrangement at dining rooms in underground mines.

Fig. 3. Lighting details of vehicle tunnels in underground mines.

over a reference surface can be calculated by the lumen method in

Eq. (1).
ES =

F n N UF LLF
A

Fig. 2. Lighting details in dressing rooms.

uorescent lights are being used, and for this case study, uorescent and LED luminaires for the same arrangement were compared
technically and economically. For general purposes, most of these
facilities are currently illuminated using IP65 uorescent light xtures with two T8 tubes (Figs. 13).

3. Methodology
3.1. Methodology for technical evaluation
The work of lighting design is often taken too lightly but it
is important and necessary to consider a consistent methodology
(Fig. 4) for planning and evaluation prior to construction. DIALux
software [7] was used to plan the lighting of new levels in the
mine. This software is an independent and manufacturer-neutral
application that allows one to model areas using CAD layouts and
photometry les from each manufacturer.
The software calculates daylight in a room or building using the
radiosity method [810], and the average illuminance and luminaire arrangements are calculated using the zonal cavity method
or the so-called lumen method [1113]. The average illuminance
produced by a lighting system to reach a specic value according
to standards [14] and user requirements (Table 1) can be calculated by means of utilization factors. The average illuminance Es

Fig. 4. Workow design lighting system.

(1)

18

F. Chueco et al. / Energy and Buildings 93 (2015) 1622

Table 1
Lighting level for underground mines.
Lighting level [lm/m2 ] according to

Work area classication

Storage rooms, washing area, rubbish area, warming area, food

storage area, general operations and storage areas
Dressing rooms
Restrooms and showers
Service corridors
Electrical rooms and substations
Dining rooms and table area
Dining rooms, delivery area
Food preparation area
Ofces and medical-attention areas
Service area access, general corridors, and tunnels
Workshops
Emergency lighting
a
b
c

CODELCOa

SECb

IESNAc

200

150

200

200
100200
100
200500
150300
300
400
500
50
500

100
100
50
300
150
300
300
500
50
200300
1

200
100
100
200500
50100
300500
200300
500
50
300500
101

Criteria of the National Copper Corporation of Chile.

Superintendence of Electricity and Fuel (SEC) Chilean Standard NCH Elec. 4/2003.
Illuminating Engineering Society of North America.

where,
ES is the average illuminance over a reference surface
F is the initial bare lamp ux (lumens, lm)
n is the number of lamps per luminaire
N is the number of luminaires
UF is the utilization factor for the reference surface s
LLF is the light loss factor, Eq. (2)
A is the work area (m2 ).
LLF = LLD LMF RSMF

extended lifetime until 100.000 h which matches with LLD lower

than 70% (Fig. 14).
3.3. Methodology for economic evaluation

(2)

where,
LLD is the light loss depreciation factor; it is the fraction of the
initial light output of a luminaire after a certain time of working.
LMF is the luminaire maintenance factor (dirty: 0.7, normal: 0.8
clean: 0.9) [15].
RSMF is the room surface maintenance factor (0.75).
3.2. Useful life and rated life discussion
IES recommends using a light loss depreciation (LLD) higher
than 70% of initial lumen (L70) to avoid dramatically noticeable
differences in light levels by users [16].
Fluorescent-rated life is determined according to LM40 [17], and
it is expected that a uorescent lamp will fail at around 20,000 h
[18]; at that point, this kind of lamp will lose around 8% of initial
lumen output [19]. It means that useful life for uorescent lamps is
the same as rated life or expected lifetime because the lamp could
be used until it fails and its LLD will be 92% at that moment, which
is greater than the 70% recommended.
Most of the LED xtures do not fail dramatically; however, LED
lamps decrease the initial lumen output during its life. LM80 [20]
establishes how much an LED lamp loses lumen output in a certain
number of operation hours to obtain lumen maintenance measurements. LED lamp manufacturers claim that LEDs last more than
50,000 h of operation and some of them until 100,000 h [21]. It
means more than 5 years of testing the lamps, which is impractical. For this reason, measurements are obtained until 6000 h or
10,000 h. LLD can be extrapolated beyond this number according to TM-21 [22]. For this kind of lamps, the expected lifetime
is greater than useful lifetime, and in general, for industrial and
building applications, useful lifetime [23] is reached for L70 even
though luminaires will work beyond this number. L50 is allowed
for decorative applications.
L70 may be too conservative for some LED products [19], but
by the moment this is the IES-recommended practices as specied
in the Lighting Handbook. However the sensitivity analysis shows

Life-cycle cost-benet analysis (LCCBA) method [24] is used to

make the right decision about the most cost-effective technology
[2527]. For both the analyzed technologies, the present value cost
(PVC) is calculated as described in Eq. (3). PVC is the present value
of the cost of investment, operation, and maintenance incurred
along its useful life, allowing a comparison of alternatives with
the present values. This method allows us to economically compare each alternative that illuminates work areas at the mine. The
difference between the PVC of each alternative is calculated, and
sensitivity analysis is carried out when the main variables are modied.
To calculate the PVC, a table of annual costs throughout the 20
years of study is developed. Every annual cash ow reects the
considered ination, in this study, 3%, which affects electricity cost
and equipment replacement. From Eq. (4), the equivalent annual
cost (EAC) is obtained [28]; and PVC is multiplied by the capital
recovery factor (CRF). EAC is a very useful economic indicator at
the moment of project evaluation when the PVC among different
alternatives is equal or similar.
PVC = I +

N

n=1

Cn

(1 + r)n

(3)

where,
Cn is the annual cost (energy and replacement costs)
I is the initial investment
N denotes the periods or yearsr is the discount rate
EAC = PVC

r(1 + r)n

(1 + r)n 1

(4)

where,
EAC is the equivalent annual cost
PVC is the present value cost
n denotes periods or years
r is the discount rate
4. Technical comparison
Every area of the mine is calculated and modeled using DIALux
to ensure the right level of lighting according to the type of work.
This technical evaluation is carried out using different luminaires

F. Chueco et al. / Energy and Buildings 93 (2015) 1622

19

Table 2
Characteristics of the modeled luminaires.
Brand

Lum. Flux (lm)

P (W)
b

lm/W

CRI

Rated Lifetime or L70a

Price (USD)

Fluorescent

A
B
C

4214
4567
4871

79
79b
79b

53
58
62

85
>80
>80

12,00020,000
12,00018,000
12,00018,000

63
79
71

LED

D
E
F

3714
3319
3556

40
53
38

93
63
94

>80
70
76

50,000
100,000
60,000

383
527
134

a
b

Rated lifetime for uorescent tubes and L70 for LED lamps.
2 36 W and ballast power (10% of total power).

available at the national market (Table 2). Three brands were evaluated for each kind of light using the manufacturers photometry
(Figs. 5 and 6). They were IP65 uorescent light xtures with two T8
tubes of 36 W each and electronic ballast and IP65 LED luminaires
using LEDs installed in a plate, and not LED tubes [29]. Some LED
xtures allow replacing the light engine or lamps and others have
to be replaced entirely; this variable has to be taken into account in
replacement costs. Vendors offered 5 years warranty for LED luminaires, although their prices show quite high dispersion and should
be paid attention to the quality of the xtures.
Calculations (Table 3) consider illuminance ux at 70% for LED
lamps at the end of the useful lifetime and 92% for uorescent lamps
at the end of the rated lifetime. It means that calculations show
the illumination level when it is time to replace the luminaire. It
ensures the right illuminance level or better than the minimum
required during its useful life.
Uniformity ratio should be more than 4:1 to ensure homogeneous illuminance in the work plane (0.76 m); it is similar among all
the evaluated luminaires. The calculations obtained (Table 3) show
parity in the results with both types of luminaires: uorescent and
LED; it means that there are no dramatically noticeable differences
in the lighting results between both types of equipment; however,
different brands could show other results depending on the quality.
After calculations, it was observed that direct 1:1 replacement
is possible. This allows considering a direct comparison for economic evaluation. Average luminance is very similar for the six

Fig. 6. Photometry of LED luminaire (Brand F).

luminaires modeled for the same arrangement and number of luminaires. Average luminance ux for initial service and at the end of
the useful life was calculated to ensure the right light levels.
As an example, the results for a dining room in the underground
mine and planning layout of electrical substation is presented
where three 30 MVA transformers and switchgears were installed.
The rooms are excavated forming tunnels, and low reecting factor
is shown in the mine due to the color of concrete used to fortify the
galleries (Figs. 79).

Fig. 5. Photometry of the uorescent luminaire 2 36 W (Brand A).

Fig. 7. Rendering of the lighting arrangement model in the dining rooms.

20

F. Chueco et al. / Energy and Buildings 93 (2015) 1622

Table 3
Results of the dining room calculations.

Fluorescent

LED

a
b

Brand

Eaverage (lx) LLD = 1

Eaverage (lx) LLD according to the type of lamp

Uniformity (average/minimum)

A
B
C
D
E
F

290
317
319
286
312
300

267a
291a
293a
200b
219b
210b

2.0
2.1
1.7
1.9
2.3
2.1

LLD = 0.92.
LLD = 0.70.

Fig. 10. Present value cost, uorescent (A, B, C) and LED (D, E, F).

Fig. 8. Details of the dining room model.

5. Economic and sensitivity analysis

Approximately 8000 luminaires are calculated for the rst 3
years in the mine expansion; after that, other facilities will be developed in the subsequent years with 50 years of exploitation. For this
case study, this rst amount of luminaires was considered.
An excel sheet with formulas automatically calculates the PVC
and the difference between both solutions for different variables.
Calculations take into consideration replacement pieces when useful lifetime is completed. Cost pieces are updated using ination for
the year of completion.

Fig. 11. Equivalent annual cost, uorescent (A, B, C) and LED (D, E, F).

5.1. Base case analysis

PVC and EAC are calculated using luminaire brands A to F
(Figs. 10 and 11). Similar gures are obtained for uorescent equipment; however, for LED luminaires, the results are very dependent
on luminaire cost. Just for LED brand F, the project is more costeffective than the uorescent project.

The lowest gures are obtained for brands A and F. PVC of

7,768,444 USD are obtained using uorescent xtures brand A and
PVC of 5,430,135 USD using LED xtures brand F. It means a saving
of 2,338,310 USD for an evaluation period of 20 years (Table 4).
A payback analysis for this case study shows that by using LED
brand F instead of uorescent solution brand A, a higher initial

Fig. 9. Lighting plan of the electrical substation.

F. Chueco et al. / Energy and Buildings 93 (2015) 1622

21

Table 4
Data for comparison between uorescent A and LED F.
Description
Fluorescent luminaire power
Ballast power
Luminaire cost per unit
Replacement cost per uorescent
tube
Ballast and starter cost
Expected lifetime uorescent tube
Expected lifetime ballast and starter

2 36 W
7W
63 USD
2 USD

LED F

LED luminaire power (W)

LED luminaire cost and replacement
costa
Useful lifetime LED (L70)

38 W
134 USD

Both

Number of luminaires
Annual operating time
Energy cost
Evaluation period
Ination
Discount rate

8000
8760 h
118.1 USD/MW h
20 years
3%
10%

Fluorescent A

15 USD
20,000 h
30,000 h

60,000 h

Fig. 13. PVC difference vs. LED luminaire and energy price.

LED luminaire replacement considers replacing the entire luminaire

investment is recovered in the second year (Fig. 12). This is due

to high-energy cost in Chile and high operation time in underground mines. The difference of accumulated cost along 20 years
for a project of 8000 luminaires is 6,851,547 USD.
5.2. Sensitivity analysis
Once checked, direct 1:1 replacement was possible; it was
evaluated economically and a sensitivity analysis was carried out
that compared luminaires A versus F. The analysis graphics
presents three-dimensional surfaces. In one of the axes, the difference between PVC obtained using luminaires F and PVC calculated
for luminaires A is shown. In the other axis, a range for main variables analyzed as electricity price, LED price, and useful lifetime is
shown. By watching these surfaces, it is possible to determine the
cost-effectiveness borders in the comparison.
The main question is: how far is the base case from the costeffective border? The most important parameters are energy price,
LED luminaire price, useful lifetime, and operation hours per year.
However, the parameter operation hours per year are not relevant
in this analysis because the project is in underground mine working
practically 24 h per day throughout the year.
Figs. 13 and 14 show the difference between the PVC of uorescent A project and LED F. The green area indicates that
uorescent PVC is higher than LED PVC, which means that the LED
solution is more cost-effective than the other solution. The red
area indicates that the uorescent solution is economically more
recommendable than the LED project.

Fig. 14. PVC difference vs. LED price and useful lifetime of LED.

The rst analysis faces the LED luminaire price versus the energy
price (Fig. 13). The LED solution is more cost-effective than uorescent project up to 280 USD LED price for the current energy price
(118.1 USD/MWh). On the other hand, for a wide range of LED xture prices, the cost-effective border depends on the energy price,
for example, for the current cost of LED F (134 USD), the energy
should cost more than 40 USD to maintain recommendation; otherwise, the uorescent project is economically better.
The second analysis faces the LED luminaire price versus useful
lifetime of the LED lamp (Fig. 14). For example, to obtain an LED PVC
that is lower than a uorescent PVC, the useful lifetime should be
better than 27,000 h if the LED luminaire price is 134 USD. On the
other hand, if the useful lifetime of LED luminaire is 60,000 h, the
cost-effective border is on 280 USD, and for 50,000 h, the luminaire
should cost less than 234 USD.
6. Conclusions

Fig. 12. Accumulated costs for uorescent A and LED F.

LED technology is emerging fast and is becoming a cost-effective

solution in some industrial and building lighting projects. Lowenergy consumption, long useful life, and low maintenance rates
are some of its benets. However, other technologies like uorescent tubes are still cost-effective, and to determine which
technology is adequate for each project, technical and economic
analyses must be done.
Manufacturers of LEDs claim saving costs due to lower energy
consumption, but the price of LED xtures can make this solution
not competitive, though they consume less energy than other technologies like uorescent tubes. In this study, only one of the three
LEDs analyzed was cost-effective versus uorescent tubes. The

22

F. Chueco et al. / Energy and Buildings 93 (2015) 1622

analyzed uorescent xtures showed similar gures and costs, but

LED prices are very different in the market.
Technically, LED xtures can substitute uorescent xtures
showing similar light levels using similar arrangements and number of luminaires.
For the case of LED brand F, with the lowest cost, this solution is
more cost-effective than the uorescent project because the most
important parameters in this project are high-energy cost and highoperation hours per year. However, there are a lot of parameters
that are necessary to analyze. Other projects where energy cost and
operation hours are lower could yield a different recommendation.
Sensitivity analysis with two parameters allows knowing how
far the project is from the cost-effective border. Surfaces allow
knowing quickly how important a parameter is to take the right
decision and mitigate risks in making decisions. In this way, the
cost of LED selected is 50% under the maximum cost to be a costeffective solution. Useful lifetime claimed by the manufacturer is
more than 200% of the minimum useful lifetime to maintain the
above cost-effectiveness.
In summary, LED xtures that cost more than 280 USD per luminaire are not cost-effective given the current energy price in Chile;
the 60,000 h of useful lifetime; the operational conditions in underground mining; and the prices of current uorescent xtures.
Acknowledgments
The authors are grateful to Jacobs support in preparing this
paper and delivering innovative solutions to its clients. The authors
also thank Codelco for its interest, for the data provided, and its
efforts to provide the best environmental conditions for its workers
in the mine.
References
[1] J. Revuelta, F. Reyes, R. Pozo, El Teniente New Mine Level Project, First International Future Mining Conference, Sydney, NSW, 1921 November, 2008.
[2] M.A. Reyes, S. Gallagher, J.J. Sammarco, Evaluation of visual performance when
using incandescent, uorescent, and LED machine lights in mesopic conditions,
IEEE Trans. Ind. Appl. 49 (5 (SeptOct)) (2013) 1992 (1999).
[3] J.J. Sammarco, T. Lutz, Visual performance for incandescent and solid-state cap
lamps in an underground mining environment, IEEE Trans. Ind. Appl. 47 (5
(SeptOct)) (2011) 2301 (2306).

[4] Alexander Galetovic, Cristin M. Munoz,

Regulated electricity retailing in Chile,
Energy Policy, Volume 39, Issue 10, October 2011, Pages 6453-6465, ISSN 03014215.
[5] Francisco J. Chueco-Fernndez, ngel A. Bayod-Rjula, Power supply for
pumping systems in northern Chile: Photovoltaics as alternative to grid extension and diesel engines, Energy, Volume 35, Issue 7, July 2010, Pages 2909-2921,
ISSN 0360-5442.
[6] The Global Energy Architecture Performance Index Report 2014, World Economic Forum, December 2013, p. 84.
[7] N. Zafar, P.K. Soori, M. Vishwas, The use of simulation tools in teaching lighting
system design, Power Engineering and Optimization Conference (PEOCO), 2013
IEEE 7th International, 34 June 2013, pp.459463.

[8] M.F. Cohen, D.P. Greenberg, D.S. Immel, P.J. Brock, An efcient radiosity
approach for realistic image synthesis, computer graphics and applications,
IEEE 6 (3 (March)) (1986) 26 (35).
[9] E.J. Padron, M. Amor, M. Boo, R. Doallo, High Performance Global Illumination on
Multi-core Architectures, Parallel, 17th Euromicro International Conference on
Distributed and Network-based Processing 2009,1820 Feb, 2009, pp.93100.
[10] D. Xiao, G. Wuyun, Z. Pan, Indoor Daylight Simulation Using Radiosity Method,
Digital Media and Digital Content Management (DMDCM), 2011 Workshop on,
1516 May, 2011, pp.158161.
[11] G. Parise, L. Martirano, Daylight Impact on energy performance of internal
lighting, Ind. Appl. IEEE Trans. 49 (1 (JanFeb)) (2013) 242 (249).
[12] C. Kao, New Concepts in Interior Lighting Design, Ind. Appl. IEEE Trans. IA-20
(Sept) (5) (1984) 1179 (1184).
[13] C.F.F.C. Filho, A.T. de Albuquerque, M.G.F. Costa, Luminance optimization in
closed environments by simulated annealing, Lat. Am. Trans. IEEE (Rev. IEEE
Am. Lat.) 8 (3 (June)) (2010) 229 (235).
[14] Low Voltage Power Installations; Chilean Standard NCH Elec. 4/2003. Instalaciones de Consumo en Baja Tensin; Norma Chilena NCH Elec. 4/2003.
[15] G. Parise, L. Martirano, S. Di Ponio, Energy performance of interior lighting
systems, Ind. Appl. IEEE Trans. 49 (6 (NovDec)) (2013) 2793 (2801).
[16] Lighting Research Center. LED Life for General Lighting: Recommendations for
the Denition and Specication of Useful Life for Light-emitting Diode Light
Sources. ASSIST Recommends, 2006.
[17] Illuminating Engineering Society (IES), LM40-01 Approved Method for Life Performance Testing of Fluorescent Lamps, The Illuminating Engineering Society
of North America, New York, 2001.
[18] M. Neary, M. Quijano, Solid state lighting for industrial locations, Petroleum
and Chemical Industry Conference, 2009. PCIC 2009. 2009 Record of Conference
PapersIndustry Applications Society 56th Annual, 1416 Sept. 2009, pp.1,7.
[19] Pacic Northwest National Laboratory, U.S. Department of Energy, Lumen
Maintenance and Light Loss Factors: Consequences of Current Design Practices
for LEDs. September 2013.
[20] Illuminating Engineering Society (IES), LM80-08 Approved Method for Measuring Lumen Maintenance of LED Light Sources, The Illuminating Engineering
Society of North America, New York, 2008.
[21] M. Schratz, C. Gupta, T.J. Struhs, K. Gray, Reducing energy and maintenance
costs while improving light quality and reliability with led lighting technology,
Pulp and Paper Industry Technical Conference (PPIC), Conference Record of
2013 Annual IEEE, 2327 June 2013, pp.43,49.
[22] Illuminating Engineering Society (IES), TM-21-11: Projecting Long Term Lumen
Maintenance of LED Light Sources, The Illuminating Engineering Society of
North America, New York, 2011.
[23] Philips Color Kinetics, Useful Life Technical BriefUnderstanding LM-80, Lumen
Maintenance, and LED Fixture Lifetime, 2010.
[24] Illuminating Engineering Society of North America (IESNA), 2000. Cap. 25.
Lighting EconomicsLighting Handbook: Reference and Application 9th Edition.
[25] V. George, A. Bagaria, P. Singh, S.R. Pampattiwar, S. Periwal, Comparison of CFL
and LED lampharmonic disturbances, economics (cost and power quality)
and maximum possible loading in a power system, Utility Exhibition on Power
and Energy Systems: Issues & Prospects for Asia (ICUE), 2011 International
Conference and, 2830 Sept. 2011, pp.1,5.
[26] Chin Kim Gan, Ahmad Farid Sapar, Yik Chee Mun, Kuan Eng Chong, TechnoEconomic Analysis of LED Lighting: A Case Study in UTeMs Faculty Building,
Procedia Engineering, Volume 53, 2013, Pages 208-216, ISSN 1877-7058.
[27] Cludio R. B. S. Rodrigues, Pedro S. Almeida, Guilherme M. Soares, Joo M.
Jorge, Danilo P. Pinto, Henrique A. C. Braga, Um Estudo Comparativo de Sistemas de Iluminaco Pblica: Estado Slido e Lmpadas de Vapor de Sdio
em Alta Presso, 2010 9th IEEE/IAS International Conference on Industry
ApplicationsINDUSCON 2010.
[28] K.S. Jardine Andrew, H.C. Tsang Albert, Maintenance, Replacement, and Reliability: Theory and Applications, CRC Press, May 2013, pp. 301.
[29] W. Ryckaert, I. Roelandts, G. Durinck, S. Forment, J. Audenaert, P. Hanselaer,
M. Van Gils, Performance of LED linear replacement lamps, Light Eng. 20 (1)
(2012) 129139.