Survey
* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project
* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project
General Electric wikipedia , lookup
Wireless power transfer wikipedia , lookup
Multidimensional empirical mode decomposition wikipedia , lookup
Telecommunications engineering wikipedia , lookup
History of electric power transmission wikipedia , lookup
Immunity-aware programming wikipedia , lookup
Renewable and Sustainable Energy Reviews 48 (2015) 710–725 Contents lists available at ScienceDirect Renewable and Sustainable Energy Reviews journal homepage: www.elsevier.com/locate/rser Lightning monitoring system for sustainable energy supply: A review Muhammad Abu Bakar Sidik a,c,n, Hamizah Binti Shahroom a, Zainal Salam b,e, Zokafle Buntat a, Zainuddin Nawawi c, Hussein Ahmad d, Muhammad ’Irfan Jambak c, Yanuar Zulardiansyah Arief a a Institute of High Voltage and High Current (IVAT), Faculty of Electrical Engineering, Universiti Teknologi Malaysia, UTM Johor Bahru 81310, Johor, Malaysia Centre of Electrical Energy Systems (CEES), Faculty of Electrical Engineering, Universiti Teknologi Malaysia, UTM Johor Bahru 81310, Johor, Malaysia c Department of Electrical Engineering, Faculty of Engineering, Universitas Sriwijaya, Ogan Ilir 30662, South Sumatera, Indonesia d Department of Electrical Power Engineering, Faculty of Electrical and Electronic Engineering, Universiti Tun Hussein Onn, 101 Parit Raja, Batu Pahat, Johor, Malaysia e Institute of Future Energy, Universiti Teknologi Malaysia, UTM Johor Bahru 81310, Johor, Malaysia b art ic l e i nf o a b s t r a c t Article history: Received 18 February 2014 Received in revised form 8 January 2015 Accepted 9 April 2015 A lightning monitoring system is used to observe, collect and analyse lightning activities so that a preventive measure to protect power equipment from severe damage can be planned. An effective lightning monitoring system is crucial to ensure the reliability and sustainability of the electrical energy supply. Despite numerous published papers on this topic, there seem to be an absence of comprehensive review papers that evaluate monitoring technologies and their performances. Owing to the literature gap, this paper is written to summarise the working principles of the relevant sensors and the various methods of data transmission, storage and analysis, as well as the various ways of predicting the occurrence of lightning strikes. This knowledge will contribute to the development of a new online lightning detection system that will be more efficient, without reducing the effectiveness and sensitivity of the system, by utilizing the available technology for data transmission and analysis. & 2015 Elsevier Ltd. All rights reserved. Keywords: Sustainable energy Lightning monitoring Lightning sensor Contents 1. 2. 3. 4. 5. 6. n Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 711 Overview of lightning monitoring system . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 712 Atmospheric electric field (AEF) and lightning sensor. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 712 3.1. Electric field mill (EFM). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 712 3.2. Integrated electro-optic sensor . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 713 3.3. Micromachined electrostatic field sensor . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 715 3.4. Flat plate antenna . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 715 3.5. Passive optical lightning sensor . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 716 3.6. Photon and infrasound detection . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 716 Lightning location technique. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 716 4.1. Single station lightning location technique . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 716 4.2. Multi-station lightning location technique . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 716 4.2.1. Magnetic direction finder (MDF) method . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 717 4.2.2. Time of arrival (TOA) method . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 717 4.2.3. Improved accuracy using combined technology (IMPACT) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 717 4.2.4. Interferometry technique . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 717 Methods of transmission and storage data . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 718 5.1. Methods of data transmission. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 718 5.2. Methods of data storage . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 720 Methods of observation and analysis data . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 721 Corresponding author. Tel.: +60 7 5535717. E-mail addresses: [email protected], [email protected] (M.A.B. Sidik). http://dx.doi.org/10.1016/j.rser.2015.04.045 1364-0321/& 2015 Elsevier Ltd. All rights reserved. M.A.B. Sidik et al. / Renewable and Sustainable Energy Reviews 48 (2015) 710–725 711 6.1. Methods of data observation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6.2. Methods of data analysis. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7. Discussions. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8. Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Acknowledgement. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Cloud to cloud discharge Intra cloud discharge Signal conditioning (High Frequency) Electric field sensors Signal conditioning (Low Frequency) Signal Processing Lightning sensors Sensors Cloud to ground discharge Data Transmission Reliability is a crucial aspect of modern electrical power supply —regardless whether the system is conventional or distributed generation. A sustainable supply of energy requires that the transmission and distribution of electrical power are guaranteed. Particularly for a distributed generation system, it is inevitable that the available energy from renewable sources (e.g. wind turbines and photovoltaic arrays) will be maximized in order to recover the high capital cost. In addition, with the expected proliferation of smart grid systems, the protection of sensitive equipment, such as telecommunication and control facilities (in smart distributed generation), must be given the priority. Ideally, it is desirable that these power supplies be as robust as possible; however, a zero level of vulnerability is practically unattainable [1,2]. In many cases, system failures or disruptions due to natural disasters, such as severe lightning and thunderstorm events, can interrupt the energy supply to consumers. More recently, the more frequent occurrence of extreme weather conditions due to global warming and the El Niño effect [3,4] have also been identified as factors that exaggerate the problem. For a lightning discharge, the resulting large current and/or voltage fluctuations can temporarily disrupt or permanently damage parts or components of the power supply chain [5–7]. The most common scenario is a direct or indirect strike on exposed transmission cables, with the consequence of switchgears and relays tripping. In a more severe situation, a direct strike on a substation, for example, can cause damage to the components of a distribution system (transformers, relays and switches)—resulting in a long recovery time. In a renewable energy system, the most vulnerable part is the power electronics equipment (dc–ac, dc–dc). A lightning strike (direct or indirect) can result in a total collapse of generation capability. Very likely, the electronics equipment will have to be replaced. As far as a smart grid is concerned, there is a need to adequately protect the telecommunication and control facilities. They are the backbone of the distributed generation, and the loss of communication significantly reduces the effectiveness of the system. From the above discussions, it is clear that lightning discharge must be considered one of the main variables in estimating risk factors [8–10]. It has to be evaluated meticulously in order to adjust the cash flows of power supply plants [11]. Lightning is a transient discharge of static electricity that serves to re-establish electrostatic equilibrium within a stormy environment [12]. It is commonly characterised by an extremely high current, high voltage and short-lived electrical discharge. To reduce the risk of being damaged or affected by lightning, various types of monitoring systems have been developed to detect and alert individuals of its eminent occurrence. It is important to note that lightning phenomena cannot be prevented, but they can be intercepted or diverted to a path that results in less damage [13]. Besides being used as an early warning system, monitored data can be used for the long-term meteorological evaluation, which in turn, can improve the understanding of worldwide climate change. As the interest in this issue has grown, substantial effort has been devoted to designing a reliable and cost-effective lightning monitoring system. With the proliferation of various commercial products using different concepts, the need to understand the principles behind their operations becomes more crucial. This is particularly vital when selecting the appropriate product for a particular meteorological condition or geographical location. Unfortunately, its importance has always been overlooked, even though the cost of a monitoring system is only a small fraction of the total investment of a power plant. Despite the numerous published research studies carried out in this area, there appears to be an absence of a single, comprehensive review paper that evaluates the relevant lightning monitoring technologies and their performances. Owing to the literature gap, this review paper is written with the following goals in mind: (1) to summarise the functions and working principles of the various sensors used in the system, (2) to elaborate on the methods of predicting the occurrence of lightning strikes and their locations and (3) to describe the periphery instrumentation required for data transmission, storage and analysis. At the end of the paper, a brief comparison of the commercially available lightning monitoring systems is also provided. To complement the analysis, an exhaustive list of references is provided for those interested in further probing the issues raised in this paper. Data storage and Data analysis 1. Introduction 721 722 722 723 723 723 Fig. 1. General block diagram of lightning monitoring system. 712 M.A.B. Sidik et al. / Renewable and Sustainable Energy Reviews 48 (2015) 710–725 Rotor Sensor Plate Sensing Plate Rotor Stator Charged sensor plate (Shutter open electric field) Uncharged sensor plate (Shutter closed electric field blocked) Fig. 2. Electric field mill, top configuration. 2. Overview of lightning monitoring system There are two categories of lightning monitoring, namely earth- and space-based systems. In this review, only the former is considered. Generally, lightning monitoring and detection systems have the same objective, that is, to predict the location and intensity of lightning occurrences. The main components are the sensors. In a typical system, the sensor configuration varies. It may consist of (1) several atmospheric electric field (AEF) sensors along with lightning sensors or (2) lightning sensors alone. The advantage of the former configuration is that it can track thundercloud development processes and movements. Therefore, it provides an opportunity to predict the thunderstorm’s location before it appears. Using the latter, such predictions cannot be achieved. Other components of a sensing system include a signal conditioning unit to process the output signal from the sensor. Generally, it consists of a signal amplifier and filter components. It may also involve a microcontroller to manage the input and output signals from the signal conditioning unit. Then, the conditioned signal will be transmitted in the form of an analogue or digital signal, depending on the data transmission protocol adopted. Finally, the data are collected and stored in a computer system. They will be analysed using certain mathematical prediction tools, and the results will be used for further actions. Fig. 1 shows the general block diagram of a typical lightning monitoring system. 3. Atmospheric electric field (AEF) and lightning sensor An AEF sensor is a device that is used to sense the presence of an electric field in the atmosphere. It monitors the magnitude of the atmospheric electric field between the clouds and the ground. The characteristics of lightning can be discovered by measuring the atmospheric electric field (AEF), which is produced by the motion of an electric charge within the clouds. The changes in the magnitude and polarity of the AEF can be observed during thundercloud formation and lightning discharge. The ability to measure the AEF enables the following: (1) identifying the thunderstorm’s duration; (2) differentiating between the various storm stage characteristics; (3) estimating the values of the lightning parameters, such as transferred charge characteristics; and (4) predicting the location of lightning occurrences [14]. Furthermore, by knowing the trends in AEF variation, one can predict the future weather conditions. For example, in normal fair weather, the AEF (close to the ground) is 100–200 V/m, while during a storm, this value can increase by as much as four orders of magnitude. Furthermore, lightning occurrence can be detected by measuring fast variations characterized by short AEF pulses [15]. The electric field mill is the most popular device, but other types, such as electro-optic integrated sensors, flat plate antennas, micro-machined electrostatic sensors, passive optical lightning sensors and photon and infrasound detection systems, are also used [15–26]. For an accurate measurement, the sensor must be installed far from buildings, trees and other tall objects. The lightning sensor has a similar general objective to the AEF sensor, that is, to monitor lightning occurrences. The difference, however, is that the lightning sensors detect this by means of an electromagnetic field, which is the combination of the electric and magnetic fields. The AEF sensor provides critical lightning information, which provides an advanced warning before lightning actually strikes. 3.1. Electric field mill (EFM) One of the most widely-used AEF sensors is the electric field mill (EFM). It functions by detecting the charge induced on the sensor electrode. In principle, the EFM measures the electric field by alternately exposing and shielding the conductor in the atmosphere, as shown in Fig. 2. The fixed induction conductor is known as a stator, while the rotated conductor that alternately shields the stator is known as the rotor. Therefore, the periodic changes in the electric field signals can be acquired accordingly. When the sensor plate is shielded by the rotor, the sensor plate is charged with an electric field [15,16,18,19]. The number of vanes in the stator and rotor can be varied, depending on the design of the EFM. The charges induced from the electrostatic field can be expressed as follows: [27] Z q¼ D dA ¼ εo EA½C ð1Þ The output voltage of EFM, Vo, is given by Vo ¼ εo EA ½V 2C ð2Þ where D is electric flux density, εo is the permittivity of the medium of the measuring field, E is the electric field, A is the area and C is the sensor variable capacitance. The electrostatic electric field can be calculated as shown in Eq. (3) [17]. Xn Q i :H i E¼ ð3Þ 3=2 ½V=m i¼1 2πεo : H 2i þL2i where Qi is the volume of the charge region, Li is the horizontal distance between the observation point and each charge region M.A.B. Sidik et al. / Renewable and Sustainable Energy Reviews 48 (2015) 710–725 713 Table 1 Summary of types and features of electric field mills. Author Type of EFM Features Hui [16] Three-dimensional EFM A double electrode difference structure was applied in both the axial (z) stator and the radial (x, y) stators The axial stator has two interlaced induction electrodes, which have the same size and shape as a rotor, while the radial stator has two pairs of planar electrodes with the same shape and size that are perpendicular to one another Can reduce the measuring error effectively and improve the sensitivity, precision and reliability Is simplified in that the cost and energy consumption are reduced by excluding the earthing brush and photoelectricity synchrodetector Optimized version of the field mill sensor The instrument body of the EFM is mostly made of aluminium, but some critical parts are made from stainless steel: the Fort [21] Boltek EFM-100 Data logging capabilities and GPS receiver Fort [21] Vaisala EFM 550 Data logging capabilities and GPS receiver Zhu [17] AMEO340 Detects lightning location in the range 0–20 km and has a measuring range from 100 kV/m toþ 100 kV/m Bennet [22–23] JCI 131 No earthing of rotating chopper Has large gaps and long insulation tracking paths in the sensor unit to operate in a wet environment Piper [24] Campbell scientific CS110 Uses a reciprocating shutter that is electrically connected to the ground potential Aranguren [25] Field mill network of Bogota Can cover 196 km2 area by using five electric field mills, with which it can measure the vertical electric field network Each mill can detect electric fields between 20 kV/m and 20 kV/m at a rate of 500 samples per seconds Fort [15] Renno [26] Miniature sensor of cylindrical field mill shutter, disk sensor and body cap Waterproof and easy to assemble Consists of a rotating shutter, sensors, shield, front end, motor control electronics and acquisition processing data storage Developed an isolated electric field mill that can remove the error caused by wind-borne charge transfer The electric field ranges between 50 V/m and 1 106 V/m Waveguide 1 Y coupling E a V φo Input Pi Output Po b electrode Waveguide 2 LiNbO3Crystal Fig. 3. Structure of integrated electro-optic sensor head using a Mach–Zehnder interferometer. and Hi is the total distance to each charge region. The measured electric field will become larger and larger as the charges of the cloud increase and approach the electric field mill. The atmospheric electric field is approximately perpendicular to plane surface [20,28]. Even though the electric field mill is far from perfect, the technology has improved over time. Table 1 shows the summary of the types of EFMs and their features. 3.2. Integrated electro-optic sensor AEF measurement using an integrated electro-optic sensor can reach close to 1000 kV/m [29]. This sensor exhibits several advantages: (1) It has a non-lensing system, which is necessary in a high-voltage environment; (2) it is easy to assemble and (3) it does not require earthquake-proofing, because the system is fully integrated before being placed in the high-voltage environment. The structure of an integrated electro-optic sensor head utilises a Mach–Zehnder interferometer, as shown in Fig. 3. The input light is divided into two equal parts by a left Y-coupling and is then transmitted through two horizontal waveguides, namely Waveguides 1 and 2. The light that travels through Waveguide 1 is modulated by an electric field. Then, the waveguides interfere with each other at the right Y-coupling. The output power of the sensor is an optical signal that is proportional to the applied signal. There is a path difference of one quarter of the wavelength between these two waveguides, which forms an optical bias of π/2. This is desirable because it allows the sensor to be driven without a DC bias voltage and allows the output to vary linearly with the input voltage. The popularity of integrated optical electrical field sensor continued to grow until 2011; three variations have been developed [30–32]. These sensors have been designed and fabricated with various features, as shown in Fig. 4: (a) with a vertical dipole with two electrodes, (b) with a horizontal dipole with two electrodes and (c) with a single-shield electrode. These sensors are fast and strong enough to measure AEFs up to 89.0 kV/m, 12.5 kV/m and 268 kV/m, respectively [32]. In addition, they are a very compact, immune to external EMI noise, and have a wide broadband response (DC to GHz). 714 M.A.B. Sidik et al. / Renewable and Sustainable Energy Reviews 48 (2015) 710–725 The fundamental principle of electro-optic modulation is based on the phase shift of the light wave during propagation, i.e., the intrinsic phase difference between the two optical paths is π/2. Po ¼ n πrLГ V ð4Þ φ¼ λd where n is the intrinsic refractive index of the crystal. The refractive index of some electro-optic crystals, such as lithium niobate (LiNbO3), will vary when exposed to the electric field. The variable r is the electro-optic coefficient corresponding to the polarization of the electric field, L is the length of the electrode, λ is the wavelength, V is the voltage across the gap d and Г ( o1) is the overlap factor of the electric field with the light wave in the crystal. The modulator output of the optical power can be expressed as follows: 3 Po ¼ Pi ½ 1 þ cos ðΔφo þ φÞ 2 ð5Þ where Po and Pi are the output and input laser power, respectively, and Δφo is the intrinsic phase difference between the two waveguides. The output of the optical power can be verified when Pi ð1 sin φÞ 2 ð6Þ If φ{1, Eq. (4) can be expressed as Po ¼ Pi ð1 φÞ 2 ð7Þ Finally, the half wave voltages, Vπ, can be identified as Vπ ¼ λd Гn3 rL ð8Þ The design of the integrated electro-optic sensor has been improved further by Ben Niu [33]. The improvement focused on the phase difference between the two waveguides. The integrated electro-optic sensor with LiNbO3 crystal substrates is used to measure high-power impulse electric fields. It can measure an Sensing Electrode (+) Combdriven structure Waveguide Grounded Shutter electrode Folded Beam Vertical dipole with two-electrode design Si Subtrate Waveguide Sensing Electrode (-) electrode Fig. 6. Schematic structure of a micromachined electrostatic field sensor. E Horizontal dipole with two-electrode design Shutter Waveguide Sensing Electrode (+) Sensing Electrode (-) electrode Single-shield electrode design Si Subtrate Fig. 7. Operating principle of a micromachined electrostatic field sensor. Fig. 4. Variations of integrated electro-optic sensors. Waveguide 1 Y coupling a Output P1 φo Input Pi b Output P2 electrode Waveguide 2 LiNbO3Crystal Fig. 5. The structure of the integrated electro-optic sensor. M.A.B. Sidik et al. / Renewable and Sustainable Energy Reviews 48 (2015) 710–725 715 Fig. 8. Flat plate antenna with RC circuit [31,32]. electric field of more than 1100 kV/m and is very effective for the measurement of lightning impulse electric fields. Fig. 5 shows the structure of the integrated electro-optic sensor proposed by Ben Niu. The output power of the each arm can be expressed as follows: Po ¼ Pi ð1 7 φÞ 2 Fig. 9. Parallel flat plate antenna [32]. ð9Þ 3.3. Micromachined electrostatic field sensor The micromachined electrostatic field sensor (EFS) can be used to measure AEF [34]. This sensor works at the resonant frequency for maximum sensitivity. The sensor architecture has three blocks, a comb-driven electrode, vibration shutter and sensing electrode, as shown in Fig. 6. The operating principle of the electrostatic field sensor is shown in Fig. 7. The shutter oscillates back and forth, thus covering the side walls of either the positive or negative sensing electrodes. This sensor is constructed via the silicon-oninsulator (SOI) process, in which the sensing electrode and grounded shutter are positioned in the same plane. The combdriven electrodes in this sensor are designed to verify that the grounded shutter works at its resonant frequency automatically. The micromachined electrostatic field sensor features include low cost, small size and low power consumption. Furthermore, the advantages of this electric field sensor are high quality factor (Q), accuracy, sensitivity and reliability. The differential output current of the electrodes of the electric field sensor can be expressed as follows: iout ¼ εo E dA dt ð10Þ where εo is the permittivity of free space, A is the effective area of the sensing electrodes and E is the electric field. 3.4. Flat plate antenna A flat plate antenna is used to detect the AEF and thus determine the lightning strike distance [35]. The flat plate antenna is capable of determining the location of a lightning strike within a short-range radius of 10 km by measuring electric field intensity. The plate is considered over ground level and the wavelength of the electric field is larger than the flat plate size [35,36]. Fig. 8 shows the flat plate antenna with the measuring circuit. Laser Meander Antenna Optical Fibre Fig. 10. The passive optical lightning sensor. The changing of the electric field will induce voltage on the flat plate antenna. The induced charge is expressed as follows: Z Z D:ds ¼ pdv ð11Þ s v D ¼ Eεo εr ð12Þ D:S ¼ Q ð13Þ By substituting Eq. (12) into Eq. (13), the normal electric field can be expressed as follows: En ¼ Q εo εr S ð14Þ Hence, the voltage between the flat plate antenna and the ground is shown below Z d Z d Q Q :d ð15Þ En dx ¼ ð 1Þdx ¼ Vg ¼ ε ε ε S o r o o εr S 0 0 The equation can be simplified by substituting Eq. (13) into Eq. (14). V g ¼ En :d ð16Þ 716 M.A.B. Sidik et al. / Renewable and Sustainable Energy Reviews 48 (2015) 710–725 where D is the electric flux density, S is the area of the plate, En is the electric field normal, Q is the induced charge, εo is the free space permittivity, εr is the relative permittivity and d is the height with respect to the ground. From Eq. (16), the voltage between the parallel flat plate antenna and the ground is directly proportional to the electric field normal to the plate and its height with respect to the ground. The height d of the antenna is modified to obtain the effective height deff. The effective height is classified as the height between the measuring circuit and the antenna. The antenna has an effective height of approximately 0.25 m and a physical height of 1.5 m [36]. The measuring circuit is attached to the flat plate antenna, the RC circuit of which is shown in Fig. 8. Due to the effect of the RC circuit, Vm becomes less than Vg. R can be neglected because R is assumed to be very large compared to C. Cg Cg V m ¼ En d ¼ Vg Cg þ C þ Cc Cg þ C þ Cc z Flat plate φ y ð17Þ Fig. 11. (a) Flat plate antenna, (b) cross loop antenna. Eq. (17) shows that the measured voltage Vm is the voltage Vg times the capacitive divider, composed of the capacitance C, the antenna capacitance with respect to the ground Cg and the capacitance of the coaxial cable Cc. The standard value of Cg is 59 pF, while Cc is 100 pF for a 1 m coaxial cable [36]. The value of Cc may change according to the length of the coaxial cable from the lower antenna to the input of the measuring circuit. Fig. 9 shows the structure of the parallel flat plate antenna; the diameter of the plate is 0.5 m, and the gap distance between the two plates is 0.03 m. 3.5. Passive optical lightning sensor Rosolem [37] proposed using a passive optical lightning sensor to detect lightning’s electromagnetic pulses. This sensor was mainly designed to indirectly monitor lightning strikes on overhead power lines, substations or power generation plants. Dissimilar to others, this sensing technique does not require solar panels, batteries or electronic control circuits. It is a viable solution that can allow the power utility to provide fast power recovery and the precise preventive maintenance of transmission lines. This passive optical lightning sensor is fabricated using a dipole meander antenna [38], a low-cost Fabry–Perot semiconductor laser operating at 1310 nm and a single-mode optical fibre to carry the detected signal to a remote optical receiver, as shown in Fig. 10. The process of lightning detection can be described as follows: An electromagnetic field is created by the intense currents of lightning discharged over a wide frequency. Then, the electromagnetic fields propagate away from the lightning channel in the form of radio waves, and their attenuation depends on the frequency of the radiation. A transitory voltage will be induced in the antenna when the radio waves reach the sensor. This voltage generates an electrical current that instantly modulates a semiconductor laser attached directly to the antenna. The optical power emitted by the laser is proportional to the intensity of the lightning-induced electromagnetic field. Therefore, it can be transmitted to other locations by using optical fibre as the laser output. 3.6. Photon and infrasound detection A combination of photon and infrasound detection is another approach to lightning detection. Photon detection has a high sensitivity and a fast response speed [39]. It involves photomultiplier tubes (PMT), a voltage biasing circuit, and a signal modulating circuit. The characteristics of the PMT allow it to be used with faint light signals, as well as fast-pulse light signal. Therefore, the PMT can be used to detect lightning. Meanwhile, infrasound detection focuses on the audible sound and infrasound of thunder. Infrasound detection involves an infrasound sensor, a transformer circuit and a signal pretreatment circuit. In order to distinguish the thunder’s sound with other natural phenomena, such as wind, typhoons, earthquakes and volcanic eruptions, an infrasound acquisition and analysis system is required. When the lightning signal occurs, PMT first attain it. Then, the infrasound sensor receives the signal. The channels in the PMT calculate the lightning flash azimuth and the time between the light and sound. The time between the light and infrasound arriving at the system, measured via the GPS-clock, is used to compute the lightning flash distance. Therefore, the exact longitude and latitude to the point of reference can be computed using this data. 4. Lightning location technique The lightning location technique can be divided into two categories: single-station and multi-station lightning location techniques. 4.1. Single station lightning location technique The single-station lightning location (SSLL) technique requires a single observation station to locate lightning. The implementation of this technique involves estimating the lightning distance that could be obtained by using several methods: (1) an amplitude-based lightning signal, (2) wave line theory, (3) electric field (EF) and magnetic field (MF) phase difference, (4) delay time difference, (5) group delay, (6) Schumann Resonance and (7) the wave impedance method. The wave impedance method is the latest SSLL method introduced by Mingli [40]. However, the singlestation lightning technique is not so popular, as the accuracy of the SSLL still requires further improvement compared with a multistation lightning location system. 4.2. Multi-station lightning location technique In order to monitor the lightning, the most common electromagnetic radio-frequency-locating methods are magnetic field direction (MDF), time of arrival (TOA), IMPACT and interferometry. M.A.B. Sidik et al. / Renewable and Sustainable Energy Reviews 48 (2015) 710–725 These methods are reviewed in Sections 4.2.1, 4.2.2, 4.2.3 and 4.2.4, respectively. These methods provide the best real-time reallocation detection of individual lightning strikes. 4.2.1. Magnetic direction finder (MDF) method Magnetic direction finder (MDF) is a method used to determine the location, movement and intensity of lightning. The observational data can be displayed textually or/and graphically. The MDF method can be implemented by using a vertical and orthogonal magnetic loop/cross loop antenna and a flat plate antenna [41]. Fig. 11 shows the types of MDF antennas [35,42]. There are two types of cross loop MDF (CL-MDF) antennas utilised in lightning detection: narrow-band (tuned) CL-MDF and gated wideband CL-MDF [43]. Narrow-band CL-MDF operates in a narrow frequency band, with the centre frequency commonly in the range of 5–10 kHz. At this frequency, the lightning signal energy is quite high, and the attenuation in the earth ionosphere waveguide is quite low. However, there are several disadvantages of narrow-band CL-MDF when lightning distance is less than 200 km. The inherent azimuthal error (polarization error 101) will occur. This error results from the detection of magnetic field components from the non-vertical (horizontal) channel. The magnetic field lines of circles in a plane are perpendicular to the non-vertical channel section. Also, the fact that the magnetic fields are improperly oriented for the direction finding of the ground strike – which occurs due to ionospheric reflections and sky waves – causes error in narrow-band CL-MDF. A gated wideband CL-MDF has been introduced to overcome the problem of large polarization errors in narrow-band CL-MDF. The north–south and east–west components of the initial peak of the return strike magnetic field were sampled to achieve gated wideband CL-MDF. Gated wideband CL-MDF has an operating bandwidth from a few kilohertz to 500 kHz. Both narrow and gated wideband CL-MDF are susceptible to site error [44]. Site error can be defined as a systematic function of direction, but generally, it is time-invariant. The presence of unwanted magnetic fields – non-flat terrain and nearby conducting objects, such as underground and overhead power lines and structures – is the source of the errors. To avoid this problem, the area surrounding the CL-MDF must be flat, and no conducting objects can be close to the area. 4.2.2. Time of arrival (TOA) method The time-of-arrival (TOA) method is able to determine the location of lightning accurately by using at least two stations [45]. The distance is obtained by calculating the arrival time of the lightning’s electromagnetic signal at the stations because the velocity of signals in space is a constant [46–48]. The time-ofarrival differences among remote stations will be calculated, and this will result in several time-difference hyperbolas. The intersection of the hyperbolas will be assumed to be the location of the lightning strike [49]. The TOA method uses three types of lightning location, which are as follows: (1) very short baseline (ten to hundreds of meters), (2) short baseline (tens of kilometres) and (3) long baseline (hundreds to thousands of kilometres). Very short baseline and short baseline commonly operate at VHF 30–300 MHz, whereas long baseline usually operates at VLF 3–300 kHz. Generally, the VHF is related to the air breakdown process [50,51], while the VLF signal results from current flow in the lightning channels [52]. For the very short baseline, the time difference between the arrivals of an individual lightning strike’s VHF pulse at the receivers is shorter than the time between the pulses; it is in the microsecond to hundred microsecond range. Short baseline generally purports to offer electromagnetic images of lightning 717 channels and to study the spatial and temporal development of the discharge [43]. Long baseline is commonly used to classify the ground strike point and the location of the flash. The long baseline system is known as the Lightning Position and Tracking System (LPATS). LPATS has been established since the 1980s, and it can operate at VLF/LF. The LPATS sensor has a higher sensitivity in terms of detecting lightning strikes as compared to IMPACT. This sensor is expensive due to the high operating cost [49]. Another method is a wideband sensor, which employs the TOA method. The Earth Networks Weather Bug Total Lightning Network (WTLN) is a lightning detection network that uses a wideband sensor to detect both cloud-to-ground (CG) and inter-cloud (IC) flash signals [53]. A real-time lightning cell tracking system and the subsequent dangerous alert system can be provided by using WTLN total lightning data 4.2.3. Improved accuracy using combined technology (IMPACT) Improved accuracy using combine technology (IMPACT) is a method that has been developed to combine the MDF and TOA methods’ techniques. In this approach, the direction finding and absolute arrival time provide azimuth information and range information, respectively [47,54]. Hence, IMPACT allows for an optimized estimate of location because it has redundant information regarding timing and angle. The combination of MDF and TOA provides some advantages compared to independent MDF or TOA. For example, a discharge that occurs along the baseline between two IMPACT sensors will be more accurately located by the intersection of two direction vectors. The Global Lightning Detection network, GLD360, was implemented using the magnetic direction finder (MDF) and long baseline timeof-arrival (TOA) methods, combined with a proprietary lightning recognition algorithm in the VLF band. This GLD360 detects 80% of all large negative strokes (420 kA in magnitude) according to the National Lightning Detection Network (NLDN) [55]. The application of MDF and TOA methods are summarized in Table 2. 4.2.4. Interferometry technique The interferometry technique is a technique used to locate the VHF pulses associated with lightning discharges. In the interferometry technique, no identification of individual pulses is needed, because the interferometer measures the phase difference between narrow-band signals corresponding to the noise-like bursts that occur during lightning. Then, these signals are received by two or more closely spaced sensors [43]. The principle of interferometry consists of determining the direction of the source of radiation using a combination of phase measurements in a relatively small bandwidth. The simplest interferometric system consists of two antennas located a few meters away from one another. Generally, the antenna is made up of an array composed of VHF dipoles that is connected with the interferometric receiver in order to determine the phase differences over a large dynamic range. The phase difference is commonly acquired by mixing the signals from several antennas, and the accuracy is achieved by integrating them over the desired time resolution. Finally, the phase differences are processed in order to obtain an azimuth from two-dimensional (2D) networks or azimuth and elevation from three-dimensional (3D) networks of the source. The information from several sensors is received by the central processor, which will determine the location of each individual event via triangulation [56–58]. The System de Surveillance et d’Alerte Foudre Par Interferometrie Radioelectrique (SAFIR) network is based on the VHF interferometry technique, which is able to observe lightning in 3D. In Table 3, a summary of research using the interferometry technique is presented. 718 M.A.B. Sidik et al. / Renewable and Sustainable Energy Reviews 48 (2015) 710–725 Table 2 Summary of lightning location using the MDF and TOA methods. Author Receiver Ibrahim [41] Orthogonal magnetic loop Magnetic direction finder (MDF) antenna and flat plate antenna Rakov [43] VHF (30–300 MHz) (Very short baseline and short baseline) Time of arrival (TOA) Electric field whip antenna Time of arrival (TOA) Rakov [43] Hussein [49] Method Features Requires at least two stations Stations are separated by distances of up to a few hundred km The intersecting point provides the source location The time difference between the arrivals of an individual lightning VHF pulse at the receivers is short compared to the time between pulses, which is a microsecond to a hundred microseconds Long baseline, also known as the Lightning Position and Tracking System (LPATS) The stations are 200–400 km apart in order to determine the location by measuring the differences between the signal arrival times at the stations A microprocessor-based system is used to store the location data, which is displayed on a video terminal The output data of this sensor are serial number, polarity, latitude and longitude Amir [56] Chonglin [53,57] Broadband circular flat antenna Time of arrival (TOA) Wideband sensor Time of arrival (TOA) of striking points, error between actual time differences reported, the identity of the three receivers used in the solution and lightning current amplitude The efficiency is in the range of 40–55% within 400 km of the detection station Three broadband circulating flat antennas having diameters of 30 cm each are used to locate the position The Earth Networks WeatherBug Total Lightning Network (WTLN) is a lightning detection network that uses a wideband sensor Has a detection frequency in the range of 1 Hz to 12 MHz The WTLN consists of an antenna, a global positioning system (GPS) receiver, a GPA-based timing circuit, a digital signal processor (DSP) and on-board storage and Internet communication equipment The WTLN sensor works when the signal is detected and the waveform is recorded. It then sends the waveforms to the central lightning detection server via the Internet The waveform of arrival time and signal amplitude can be used to determine the peak current of the strike and the exact location of the flash, including the latitude, longitude and altitude The WTLN has good IC detection efficiency, which is up to 95% Cummins [54,58] IMPACT sensor TOA and MDF In the range of 0.4 Hz to 400 kHz, lightning can be detected by an IMPACT sensor Provides azimuth and range information Estimated parameters: longitude, latitude and discharge time Ibrahim [59] Small active antenna TOA and MDF Lojou [55] Orthogonally oriented magnetic loop antenna TOA and MDF, combined with a lightning waveform recognition algorithm Known as the Global Lightning Detection network, or the GLD360 A worldwide detection of flashes with significant detection efficiency and Known as the Pekan Lightning Detection System (PLDS) Consists of two orthogonal magnetic loops Does not contain polarization error Simple, reliable and easy to use Able to detect and analyse real-time lightning data, which can be displayed both textually and graphically location accuracy of the CG flashes, along with the polarity of the measured discharges and an estimate of peak current Able to accurately locate most of the lightning discharged around the world A long-range global network that uses multiple terrestrial sensors to calculate lightning discharge locations (latitude and longitude) and times In a 24-h period, 588,200 events were detected Provides accurate, early detection and tracking of severe weather It has no data gaps, as satellite and radar systems do. Also, there is no need for maintenance or expenditures The data received from the GLD360 can be derived in a week, which is quite quickly, thus providing better end-user service In terms of accuracy, the GLD360 outperforms all other long-range lightning detection networks, including satellites It correctly identified the polarity in 145,901 events, which represented 94% accuracy 5. Methods of transmission and storage data 5.1. Methods of data transmission Data transmission is an important process in a lightning monitoring system. Several methods of data transmission collected from approximately 22 published works are reviewed in Section 5.1. Furthermore, because the received data must be stored securely, several methods of data storage are reviewed in Section 5.2. Data transmission is the physical transfer of data from one point to another point through a particular channel [59]. Some of the channels used to transmit data include coaxial cable [35,60–62], fibre-optic cable [63–66] and wireless communication [67–70]. Coaxial cable has a good bandwidth, low error rate and relatively low resistance. It can M.A.B. Sidik et al. / Renewable and Sustainable Energy Reviews 48 (2015) 710–725 719 Table 3 Summary of research using the interferometry technique. Author Network Qing [63] Baran [64] SAFIR 3000 Features Able to observe lightning in 3D Located at over 1000 m in altitude The interferometric array uses the differential phase measurement of lightning’s electromagnetic waves for the long-range detection of lightning activity The SAFIR LF discrimination sensor is a wideband antenna that is capable of identifying the cloud-ground characteristics of lightning Has the capability to provide lightning warnings via a moving trace of thunderstorm cells and also to detect the threedimensional distribution of lightning discharges CG and IC lightning can be located, and this information is then displayed in three dimensions at a high temporal and spatial resolution Provides lightning warnings based on a moving trace of thunderstorm cells almost 30 min in advance Ahmad [49] SAFIR Has the highest location accuracy and detection efficiency and also provides the maximum warning time The detection efficiency is 95%, and the location accuracy is in the range of 1.0 to 2.0 km The typical range of interferometry is about 50–150 km, and an interferometer is composed at least three sensors Lojou [55] Total lightning sensor TLS200 Uses five dipole arrays to determine the direction of the lightning sources based on phase differences of the signal at each dipole Can be used for accurate 2D mapping of cloud flashes, as well as efficient CG flash location Provides an increase in the detection range of the VHF total lightning sensors of more than 30%, as well as more stable operations Coaxial Cable 40 m Magnetics Loop Antenna LabVIEW 2011 Coaxial Cable 40 m Wire Antenna Fig. 12. Coaxial cable connections in a single lightning detection station. support multiple channels and a wide range of services, i.e., voice, data, video and multimedia. The transmission speeds achievable via coaxial cable are as high as 10 Gbps. Coaxial cable is strong, but there are limits to its length; it will stop working if these limits are exceeded. Usually, coaxial cable is capable of reaching hundreds of meters in length. This type of cable is also suitable for transferring high-speed signals without high losses. However, coaxial cable has problems with deployment, such as bus topology [71]. The bus topology will be susceptible to congestion, noise and security risks. The coaxial cable may be a path via which the lightning current can infiltrate equipment or devices and cause damage if adequate maintenance is not performed on the lightning protection system. Fig. 12 shows the example of the coaxial cable connection in a single lightning detection station [41]. Fibre-optic cable uses thin strands of glass in its architecture. The pulse of light travelling along the core of the fibre allows it the highest available data transmission speed. Usually, the transmission speed of fibre-optic cable is between 100 and 2000 Mbps. Fibre-optic cable offers high bandwidth and long-distance terminals to transmit data, as well as resistance to lightning disturbances. However, the cable can only stretch over a certain number of kilometers before the signal becomes impossible to detect. Also, fibre-optic cable is expensive to install and fragile. Wireless Local Area Networking (WLAN) technology is another option for data transmission. The coverage area of WLAN is close to 20 km2 [62]. The advantages of using WLAN in data transmission are its flexibility (it can transmit data within the radio coverage area) and the fact that its nodes can communicate without further restriction. Ad hoc wireless networks allow for communication without planning, which makes them more convenient than wired networks [72]. WLAN technology is robust in that it is disaster-resilient [73]. In theory, the speed of standard WLAN is rated according to its maximum bandwidth, where 802.11b offers up to 11 Mbps, 802.11a and 802.11g offer up to 54 Mbps and 802.11n offers up to 300 Mbs. However, the performance of WLAN is not as good as these theoretical values. In reality, it is only capable of achieving 5.5 Mbps, 20 Mbps and 100 Mbps, respectively, according to the standard [74]. Nevertheless, WLAN also has disadvantages, such as lower quality of service (QoS), lower bandwidth due to limitations in transmission (1/10 Mbit/s) and high error rates due to interference from other sources [75]. Cordless phones, Bluetooth devices and motion detectors all employ radio frequencies that can interfere with WLAN [76]. This interference can 720 M.A.B. Sidik et al. / Renewable and Sustainable Energy Reviews 48 (2015) 710–725 Battery 9 Vdc Oscilloscope Voltage Regulator 5 Vdc I I E ∫ Vout Monostable Multivibrator PIC 16F877A Microcontroller Indicator and SCI circuit PC tor uc nd co Fig. 13. A lightning flash counter with an SMS data sender. Table 4 The comparison of transmission used for collecting data of lightning. Transmission type Coaxial cable Quality and significant difference Cost benefit Good bandwidth, low error rate and relatively low resistance EMC/EMI susceptible without proper protection system Suitable for transferring high-speed signals without high losses Moderate installation cost Fibre optic High bandwidth and long-distance terminals to transmit data EMC/EMI resistance Moderate speed Frigile High installation cost WLAN Flexible—it can transmit data within the radio coverage area Disaster-resilient EMC/EMI susceptible Moderate speed Low installation cost GSM–GPRS EMC/EMI susceptible yes Low speed reduce the quality of WLAN service, and it also can weaken WLAN security. Data can also be transmitted by using a remote file transfer protocol (FTP) server via a Global System for Mobile–General Packet Radio Service (GSM–GPRS) modems [77,78]. A GSM–GPRS modem can exist in the form of an external unit or a PC card. The external unit could be connected through a serial cable, a USB cable, Bluetooth or infrared. The GPRS has much higher data transmission speed in that large amounts of data can be transferred to and from the mobile device over the Internet. One advantage of GPRS is that it has a great backup option. The portability factor has diminished with the emergence of much faster data cards. GPRS can also be used to transmit SMS. The data can also be transmitted via modular phone SMS [79,80]. The SMS transmission speed that can be achieved is about 30 SMS messages per minute. A GPRS modem is required to send and receive SMS via GPRS. In this way, the exact date of the lightning event, polarity, current crest amplitude, rise time and decay time can be recorded and sent to users via SMS. Hussein Ahmad presented an SMS system as a method of data transmission [81]. The time, date and the number of lightning strikes are required so that a lightning counter can send this information to a specified mobile phone via SMS. The AT command plays an important role in sending out SMS signals in that it is important to focus on the auto SMS function. There are a few procedures that must be followed before writing the AT command. The mobile phone must perform a test in order to determine the types of SMS modes that the mobile phone can support before it Zero installation cost Monthly subscription payment can perform the task. There are two types of SMS modes, which are PDU mode and text mode. The message contents in PDU mode must be converted to Hex code before being sent. Meanwhile, for text mode, an SMS can be written normally in alphabetical format. Fig. 13 shows a lightning flash counter with an SMS data sender. The AEF measurement information can also be transmitted using wireless data transmission by using the Global System Mobile Communication (GSM) network [77]. The coverage of GSM, which can reach across distances of hundreds or thousands of kilometres, is able to transmit the data through an Internet provider to a PC. This method offers low cost and the ability to transmit data using worldwide coverage. In addition, it will reduce the manpower needed by eliminating the routine inspections that were previously necessary to collect the data manually, and the computing system does not need to be installed on-site. In Table 4, types of transmission related to the quality and cost benefit is presented. 5.2. Methods of data storage It is important to safely store the data collected before analysis is carried out. The acquired data can be stored in a secure digital card, also known as an SD card. This can protect important data from being accidentally overwritten. These SDs card are provided with a mechanical switch that acts as a write protector. One advantage of using an SD card in data storage is its non-volatile memory, meaning that it keeps data stable on the card. The data on this card are not threatened by the loss of power and do not M.A.B. Sidik et al. / Renewable and Sustainable Energy Reviews 48 (2015) 710–725 721 Cloud +q i +q1 +qo +q2 +q3 t/2 -q1 -qo -q2 -q3 -q i Fig. 14. Dynamic cloud model. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.) need to be periodically refreshed. Furthermore, such cards are free from mechanical difficulties and damages. An SD card does not produced any noise while at work, allows immediate access and is also easy to keep track of. Moreover, it has a relatively large amount of storage space compared to old backup devices. Unfortunately, the SD card also has the disadvantages of being easy to break, lose, misplace or smash. In addition, SD cards may be affected by various forms of electronic corruption, such as viruses, making the entire card unreadable or broken. In addition, the digital signal oscilloscope (DSO) is used to capture and store the electric field signal [35]. The DSO provides advanced trigger, storage, display and measurement features. The data also can be stored by using a MySql database, which is a database management system. The data sent from the GSM module will be sent to the MySql database and displayed as a Hypertext preprocessor (PHP) page [77,82]. Using this method results in better data storage because the data are saved in an HTML page, which is viewable via a web browser. Thus, it is easier for users to read the data collected. 6. Methods of observation and analysis data The data obtained from lightning monitoring systems can be analysed further. The data are not only used for detecting lightning but can also be used for observing and analysing climate changes in the atmosphere. Discovering useful information within the observational data allows us to suggest conclusions and supports decision making. Several methods of lightning data analysis and observation have been compiled in Sections 6.1 and 6.2. The informative data from the sensors can be used in various ways via these methods. 6.1. Methods of data observation Several types of methods are used to observe the data. An isokeraunic-level (IKL) map is used to estimate the values from the mapped thunderstorm data yearly because it is difficult to measure the ground flash density directly [83]. The IKL map is related to the number of flashes per unit area per year. Thus, it is possible to determine the frequency of lightning in an observed region quite easily. Even so, the accuracy of this approach is low. Because the data collection depends only on the sound of thunder being heard by observers in weather observation stations, only thunderstorms occurring near the stations are registered. When the lightning location is far enough from the station, the sound of thunder is very quiet. Consequently, some data may be lost for that location. In addition, the IKL map shows the average figures for thunderstorm days gathered over many years. The IKL map only shows the variation in lightning activities in a given area per year. In order to present developing storms, a dynamic cloud model is introduced [84]. In this model, the colours blue and red are defined as the positive charge and negative charge, respectively. The arrow pointers represent their random positions, and their magnitudes correspond to the sizes of the circles. Some important parameters are required to obtain the dynamic cloud model, i.e., (1) the positive and negative charges (7 Q), (2) the distance between the positive and negative charges and (3) the cloud height. These are used to calculate the electric field for a particular time interval. Fig. 14 shows the dynamic cloud model, in which the positions and magnitudes of the positive and negative charges change with time as the thundercloud propagates. In the Beijing and Heibei areas, the meteorological data, as well as the locations of lightning discharges and the number of flashes, were collected by using the SAFIR 3000 network. [85]. The recorded lightning discharge locations were related to the storm distribution along a southwest-northeast frontal disturbance at which the two independent thunderstorm cells moved separately to the north and south. Furthermore, a prototype of a single-station lightning locating system (S-LLS) has been developed that consists of three configurations. There are lightning signal sensing (LSS), lightning data acquisition (LDA) and lightning data display (LDD) modules [86]. The LSS module is responsible for electromagnetic signal capture. This module consists of a flat-plate antenna, two crossed-loop antennas, a cloud-to-ground/cloud-to-cloud (CG/CC) logic circuit and a lightning simulator for testing. Then, the results from three analog signals and CG/CC logic are transmitted to the LDA module for processing. The LDA module is a micro-computer-based data processor with embedded A/D converters. This module processes the signals from the LLS based on a proposed lightning-locating theory. The CG/CC logic circuit in the LLS module provides information regarding the lightning impulses. Meanwhile, the LDA provides the time, location, strength and flash type for the lightning events to an on-board memory and then transmits these data to the LDD module via the Internet. The LDD module is a software pack installed on a PC. The LDD module provides realtime lightning monitoring and off-line data playback functions, as well as the visualization, statistical analysis and permanent storage of lightning events. High-speed video cameras are used to record the lightning flash propagation. The temporal upward and downward leaders’ progress can be captured by using a high-speed video camera. One kind of high-speed video camera can capture 5000 f/s at the highest resolution of 1024 800 [79,87]. Instead of using a high-speed video camera, a high-speed charge-coupled device (CCD) video camera could be deployed to capture the development process and physical form of lightning discharges [88]. Given the high frame rate and large resolution, the lightning can be observed clearly. The recording of a high-speed CCD video camera system was used to synchronize the picture output with electromagnetic field waveform data. The synchronization accuracy between the picture output from the high-speed CCD video camera and the electromagnetic field waveform data can be nearly 100 ns. A low-speed video camera, such as one that captures 50 f/s, could be also utilised in observing lightning, but its function may be limited to detecting significant upward lightning strikes on the 722 M.A.B. Sidik et al. / Renewable and Sustainable Energy Reviews 48 (2015) 710–725 tall telecommunication antenna and locating low-current lightning strikes [79]. A recent research was carried out by Yang et al. [89] worked on to integrate Geographic Information System (GIS), databases, Remote Sensing (RS), visual reality, 3D visualisation, spatial information grid and computer networking to develop a Digital Lightning Prototype System (DLPS) from the perspective of the digital earth. 6.2. Methods of data analysis Data analysis is important in obtaining conclusions from research. Therefore, several methods have been used to analyse data regarding lightning events. LabVIEW 8.5 software was used to collect, calculate and analyse lightning data by developing a virtual instrument. Using a statistic function provided in LabVIEW, the time maximum for each signal from the receiver – a VHF antenna (2 MHz cut-off frequency) and a microphone – was determined. The arrival time difference (time delay) between the electric field and acoustic signals was calculated. Meanwhile, the noise in the captured signals was eliminated digitally by using a high pass filter (HPF) in LabVIEW [60,90]. Another virtual instrument developed using LabVIEW was used to monitor lightning activity within a 25 km radius from Pekan [35,91]. The incoming lightning strikes’ signals were analysed by using various statistical functions and techniques. Afterwards, the lightning strike locations were estimated by calculating the ratio of EMF to EF. The Pekan Lightning Detection System (PLDS) system analyses the raw lightning data and transforms it into real-time lightning data [92]. The change of colour to either yellow or red in the border of the window is used to provide an alert. A yellow alert will be triggered when the strike is within a 50 km radius, while a red alert will be triggered when the strike is within a 10 km radius of the sensor location. A special ‘storm’ is depicted to indicate the detected flashes. When one of these storms is shown, there is a good chance that lightning will occur in that area. The threat assessment window is a description of storm activity over the previous 5–20 min. This window shows information about the average position of all lightning flashes detected in the previous 5– 20 min. To calculate the placement of the coloured area in the threat assessment window, the average range of all strikes in a similar direction is used. Moreover, the strike rate graph shows the rate of change in lightning activity at a glance. The window rate is also analysed, and this analysis is shown in the form of six counter displays. The six counters are the total rate counter, the strong strike rate counter, the noise rate counter, the strike rate percentage change counter and the strike rate counter. In addition to LabVIEW, Microsoft Visual Basic 6.0 was used to develop a graphical user interface (GUI) that was intended to record and count every time lightning struck a specific structure. The GUI was connected with a database that was created using Microsoft Access. The database consisted of two columns. The first column is used to store the frequency of lightning current discharge through a conductor, while the second column is used to record the date and time of each discharge [81]. Another method of analysing the data used a numerical simulation. The numerical simulation was developed to investigate the relationship between AEF and cloud height. The partial differential equation (PDE) toolbox in Matlab was applied during this analysis to simulate the static electric field distribution in a courtyard [20]. Furthermore, the data collected from lightning events can be analysed using artificial intelligence techniques (AI), which consist of fuzzy logic analysis, neural network analysis and neuro-fuzzy analysis [93]. The fuzzy logic method can be used to analyse the data with a multiple-input multiple-output (MIMO) system. Because the neural network develops learning, the input-output relationship can be used to determine the threat level in real time. By using a neural network, the exact location of positive and negative charges based on the AEF can be predicted. The neural network consists of an input layer, a hidden layer and an output layer [39,94]. Neuro-fuzzy analysis is derived from two reliable techniques: fuzzy logic and neural networks. The output is usually presented in polynomial form with input variables x and y. There are two types of neuro-fuzzy modes: Sugeno and Mamdani. Neuro-fuzzy analysis can be performed using several methods, such as mean percentage error, root mean square error and mean squared error [95]. The area of a lightning strike can be found based on latitude and longitude. Analysis using the fuzzy logic method created a graph of lightning level and also a graph of the number of lightning incidences corresponding to the type of lightning. From the results obtained, the lightning was mapped onto the Malaysian map using Google Earth. The characteristics of the lightning were represented by blue, purple and red icons to indicate low, medium and high levels of current, respectively. Nevertheless, there are some limitations when mapping lightning characteristics by using Google Earth; the map merely displays the lightning current value, the level of current and the region represented. 7. Discussions It is difficult to determine the best sensors/methods for several reasons, such as the following: (1) in the literature, system sizes, working principles and sensor structures vary; (2) the efficiency of sensors/methods vary; (3) their limitations vary; (4) there are differences in location installation and (5) the hardware used in the sensors/methods vary. Based on these difficulties, a summary of the working principle of every sensor/method is provided. The electric field mill (EFM) is based on the electromagnetic induction principle, which uses a rotor and stator to expose and shield the sensor plate. The working principle of this sensor is approximately the same as that of a micromachined electrostatic field sensor, which oscillates the shutter back and forth to measure the electric field. The integrated electro-optic sensor uses light to drive the sensor, while photon detection also uses a fast-pulse light signal to obtain the lightning signal. The passive optical lightning sensor’s working principle is based on detecting the radio frequency emitted by lightning. In lightning location networks, the magnetic direction finder (MDF) measures the voltage signal on the loop; two antennas are used in this method. The direction of the flash can be determined by comparing the voltage signals from both antennas. The time-ofarrival (TOA) method also uses a minimum of two antennas; the intersecting hyperbolas at the two stations are used to locate the lightning flash. The TOA method uses the time difference between sensings of the arrivals of the electric pulse emitted by lightning. Ideally, the TOA method is more sensitive than MDF. On the other hand, the interferometry method also can be performed using a minimum of two antennas. It can locate the VHF pulse in order to determine the direction of the source using a combination of phase measurements. In order to transmit the data, an effective method of transmission is crucial. The connection between the sensor/antenna and the data acquisition system requires a cable, WLAN technology or GSM–GPRS. In terms of length and coverage area, fibre-optic cable is capable of operating over longer distances than coaxial cable. Given the short distance between the sensor/antenna and the data acquisition system used to monitor the signal onsite, the researchers preferred to use coaxial cable. The coverage area for WLAN can M.A.B. Sidik et al. / Renewable and Sustainable Energy Reviews 48 (2015) 710–725 reach 20 km2, while GPRS-GSM can transmit information for hundreds or thousands of kilometres via an Internet provider. Fibre-optic cable has the highest transmission speed, around 100 to 2000 Mbps. Coaxial cable is only capable of reaching transmission speeds of 10 Gbps. WLAN can reach a high transmission speed, but it is not as good as fibre-optic cable or coaxial cable, because many factors limit WLAN speed, such as radio interference, physical obstructions of the line-of-sight between devices, the distance between devices and multiple devices communicating over the network simultaneously. GPRS has a slower speed than the other methods, having an average rate of 40 to 50 kps. In terms of the installation process and cost, fibre-optic cable is expensive and fragile. Coaxial cable is affordable, and its installation is easier than that of fibre-optic cable. The wireless method is preferable because of the ease of connection. GSM–GPRS usually has associated charges based on the volume of data sent. Generally, data storage can be performed on a device or online. The SD card and digital signal oscilloscope (DSO) are devices used to store data. MySQL is the online form of storage in which data are stored in a database. The SD card offers many types of memory capacity, i.e., 4 GB, 32 GB, etc. Unfortunately, it can be exposed to viruses and can be easily to lost or broken because it is small in size. The DSO cannot store data permanently, because the device has storage limitations. If the memory limit is exceeded, the oldest data will be deleted. MySQL is a safer way to store the data. It is typically easier to obtain the data online than to find the device on which the data are stored. Currently, the development of worldwide multimedia has made it easier to acquire information online. Observational data are important in monitoring behaviour, events or physical characteristics. The iso-keraunic-level (IKL) dynamic cloud model, SAFIR 300, and single-station lightninglocating system (S-LLS) are methods of collecting thunderstorm data in order to estimate the location of strikes, lightning impulse information and meteorological data. High-speed video cameras, low-speed video cameras and high-speed charge-coupled device video cameras are able to record the physical characteristics of lightning. Generally, the data analysis is used in discovering useful information in order to support decision making. Some of the methods are simple, while others are complex. For LabVIEW, the analysis is related to graphical programming because it uses a visual programming language. The program contains a block diagram, a front panel and a connector panel, which contain the graphical source code. It is very interactive compared to writing in a traditional language. On the other hand, Microsoft Visual Basic and Matlab have been used to develop a graphical user interface (GUI). A GUI can be created interactively or programmatically [96]. Nevertheless, C/Cþ þ is the most common programming language used to develop software applications. The artificial intelligence method of neural networks is complicated because it includes a multi-layer network that must be trained in order to obtain the prediction results. On the other hand, fuzzy logic is much simpler than neural networks, but its performance is lower. In order to improve the performance of fuzzy logic, it can be combined with neural networks. The combination of both methods increases the complexities involved. 8. Conclusion This paper reviews and summarizes lightning monitoring systems according to their technologies and data acquisition systems based on various academic journals. This paper provides the working principal and structure of each sensor/method in every lightning monitoring system that has been constructed. Furthermore, it reviews the methods of data transmission, data 723 storage, data observation and data analysis. It is difficult to discuss the details of every sensor/method because every sensor/method has its own benchmarking. Therefore, this paper includes only a general discussion of the structure and working principle of each method/sensor, as well as the advantages and disadvantages of the methods of transmitting and storing data, observational methods and algorithms of the analysis methods. It is hoped that the information in this paper can be a source of information for future research. Acknowledgement The authors gratefully acknowledge the support extended by Ministry of Higher Education (MOHE), and the Universiti Teknologi Malaysia (UTM) for providing the seed money as for financial support during the course of this investigation on the Research University Grant (GUP), no. Q.J130000.2523.03H59. References [1] Verbruggen A, Laes E, Lemmens S. Assessment of the actual sustainability of nuclear fission power. Renewable Sustainable Energy Rev 2014;32:16–28. [2] Kandilli C, Ulgen K. Review and modelling the systems of transmission concentrated solar energy via optical fibres. Renewable Sustainable Energy Rev 2009;13(1):67–84. [3] Bo Z, et al. An analysis of previous blackouts in the world: lessons for China' s power industry. Renewable Sustainable Energy Rev 2015;42:1151–63. [4] Mabel MC, Raj RE, Fernandez E. Analysis on reliability aspects of wind power. Renewable Sustainable Energy Rev 2011;15(2):1210–6. [5] Saidur R, et al. Environmental impact of wind energy. Renewable Sustainable Energy Rev 2011;15(5):2423–30. [6] Eltawil MA, Zhao Z. Grid-connected photovoltaic power systems: technical and potential problems—a review. Renewable Sustainable Energy Rev 2010;14 (1):112–29. [7] Cooperman A, Martinez M. Load monitoring for active control of wind turbines. Renewable Sustainable Energy Rev 2015:189–20141 2015:189–201. [8] Joselin Herbert G, Iniyan S, Amutha D. A review of technical issues on the development of wind farms. Renewable Sustainable Energy Rev 2014:619–4132 2014:619–41. [9] Hameed Z, et al. Condition monitoring and fault detection of wind turbines and related algorithms: a review. Renewable Sustainable Energy Rev 2009;13 (1):1–39. [10] Kaundinya DP, Balachandra P, Ravindranath N. Grid-connected versus standalone energy systems for decentralized power—a review of literature. Renewable Sustainable Energy Rev 2009;13(8):2041–50. [11] Wu C, et al. Multistage stochastic inexact chance-constraint programming for an integrated biomass-municipal solid waste power supply management under uncertainty. Renewable Sustainable Energy Rev 2015:1244–5441 2015:1244–54. [12] Karthick, S, G Jason. Lightning as atmospheric electricity. In: Canadian conference on electrical and computer engineering 2006 (CCECE’06); 2006. [13] Sidik MAB, et al. Study on the effectiveness of lightning rod tips in capturing lightning leaders. Electr Engin 2013;95(4):367–81. [14] Aranguren, D, et al. Estimation of lightning electrostatic parameters based on atmospheric electric field measurements and genetic algorithms. In: Proceedings 2006 ILDC/ILMC conference, Tucson, AZ, USA; 2006. [15] Fort, A, et al. Design and modeling of an optimized sensor for atmospheric electric field measurement. In: IEEE sensors applications symposium (SAS), 2010; 2010. [16] Hui, J, Z Bihua, L Yanxin. The latest design and development of the field mill used as atmospheric electric field sensor. In: Fifth Asia-Pacific conference on environmental electromagnetics, 2009 (CEEM 2009); 2009. [17] Zhu, H, Z Yang, T Yang. The application of electric field data combined with other observed data in lightning warning. In: International conference on multimedia technology (ICMT), 2010; 2010. [18] Montanya, J., Basis for lightning warning systems based in natural electric field measurement; 2012. [19] Wang ZH, et al. A study of the electrostatic field networking in three isolated thunderstorms. Appl Mech Mater 2013:775–84239 2013:775–84. [20] Yang, B, B Zhou, T Gao. Research on distribution of electric field near ground during thunderstorm process. In: International symposium on electromagnetic compatibility, 2007 (EMC 2007); 2007. [21] Fort A, et al. Design, modeling, and test of a system for atmospheric electric field measurement. IEEE Trans Instrum Meas 2011;60(8):2778–85. [22] Bennett AJ, Harrison RG. Variability in surface atmospheric electric field measurements. J Phys Conf Ser 2008;142(1):012046. [23] Chubb, J, J Harbour. A system for the advance warning of risk of lightning. In: Proc ESA annual meeting; 2000. 724 M.A.B. Sidik et al. / Renewable and Sustainable Energy Reviews 48 (2015) 710–725 [24] Piper IM, Bennett AJ. Observations of the atmospheric electric field during two case studies of boundary layer processes. Environ Res Lett 2012;7(1):014017. [25] Aranguren D, et al. On the lightning hazard warning using electrostatic field: analysis of summer thunderstorms in Spain. J Electrostat 2009;67(2– 3):507–12. [26] Renno NO, et al. A miniature sensor for electrical field measurements in dusty planetary atmospheres. J Phys Conf Ser 2008;142(1):012075. [27] Kil G-S, et al. Development of a Lightning Warning System. J Energy Power Eng 2012;6(2):287–92. [28] Wang Z, et al. The electrostatic field networking in three isolated thunderstorms. TELKOMNIKA Indonesian J Electr Eng 2013;11(1):63–72. [29] Rong, Z, et al. Novel integrated electro-optic sensor for intensive transient electric field measurement. In: IEEE international symposium on electromagnetic compatibility, 2006 (EMC 2006); 2006. [30] Rong Z, et al. Design and application of an integrated electro-optic sensor for intensive electric field measurement. IEEE Trans Dielectr Electr Insul 2011;18 (1):312–9. [31] Rong Z, et al. Measurement of electric field distribution along composite insulators by integrated optical electric field sensor. IEEE Trans Dielectr Electr Insul 2008;15(1):302–10. [32] Rong, Z, et al. The development of integrated electro-optic sensor for intensive electric field measurement. In: IEEE international symposium on electromagnetic compatibility, 2007 (EMC 2007); 2007. [33] Ben, N, et al. An electro-optic integrated sensor for lightning impulse electric field measurements. In: Asia-Pacific symposium on Electromagnetic compatibility and 19th international Zurich symposium on electromagnetic compatibility, 2008 (APEMC 2008); 2008. [34] Chunrong, P, et al. Measuring atmospheric electric field using novel micromachined sensor. In: IEEE international conference on nano/micro engineered and molecular systems (NEMS), 2011; 2011. [35] Ibrahim, WI, et al. Measurement of vertical Electric fields from lightning flashes using parallel plate antenna. In: International conference on electrical, control and computer engineering (INECCE); 2011. [36] Ahmad NA. Broadband and HF radiation from cloud flashes and narrow bipolar pulses. Uppsala University: Sweden: Faculty of Science and Technology; 2011. p. 64. [37] Rosolem JB, et al. A passive opto-electronic lightning sensor based on electromagnetic field detection for utilities applications. Meas Sci Technol 2010;21(9):094032. [38] Rashed J, Chen-To T. A new class of resonant antennas. IEEE Trans Antennas Propag 1991;39(9):1428–30. [39] Wu, Y, et al. Design of lightning location system based on photon and infrasound detection. In: Eighth international conference on electronic measurement and instruments, 2007 (ICEMI ’07); 2007. [40] Mingli C, Lu T, Du Y. An improved wave impedance approach for locating close lightning stroke from single station observation and its validation. J Atmos Sol Terr Phys 2014. [41] Ibrahim, W, M Ghazali. Measurements of electric and magnetic fields due to lightning strokes based on single-station detection. In: IEEE Asia Pacific conference on applied electromagnetics (APACE2012), Avillion Legacy Melaka; 2012 . [42] Sharshar, AA, SH Zainud-Deen, SM El-Halafawy. Crossed square loop antenna. In: IEEE antennas and propagation society international symposium Digest, 1997; 1997. [43] Rakov V. Electromagnetic methods of lightning detection. Surv Geophys 2013:1–23. [44] Tao, L, C Mingli, D Yaping. Site errors estimation and correction for MDF/TOA combined lightning location network. In: International conference on lightning protection (ICLP); 2012. [45] Nastos P, Matsangouras I, Chronis T. Spatio-temporal analysis of lightning activity over Greece—preliminary results derived from the recent state precision lightning network. Atmos Res 2014:207–17144 2014:207–17. [46] Fagan, D, R Meier. Intelligent time of arrival estimation. In: IEEE forum on integrated and sustainable transportation system (FISTS); 2011. [47] Zeng Q, et al. The application of lightning forecasting based on surface electrostatic field observations and radar data. J Electrostat 2013;71(1):6–13. [48] Yoshida S, et al. Initial results of LF sensor network for lightning observation and characteristics of lightning emission in LF band. J Geophys Res Atmos 2014;119(21) 12,034-12,051. [49] Ahmad, H, M Alam. Review of lightning warning system: technical and economic aspect; 2008. [50] Takai H, et al. Lightning detection at the telescope array cosmic ray observatory. Bull Am Phys Soc 2014:59. [51] Behnke S, McNutt S. Using lightning observations as a volcanic eruption monitoring tool. Bull Volcanol 2014;76(8):1–12. [52] Dentel LM, da Rocha BRP, de Souza JRS. Evaluation of STARNET lightning detection performance in the Amazon region. Int J Remote Sens 2013;35 (1):115–26. [53] Chonglin, L, S Heckman. Using total lightning data in severe storm prediction: global case study analysis from North America, Brazil and Australia. In: International symposium on lightning protection (XI SIPDA); 2011. [54] Cummins, KL, MJ Murphy, JV Tuel. Lightning detection methods and meteorological applications. In: IV International symposium on military meteorology, Malbork, Poland; September, 2000. [55] Lojou, J, et al. Latest developments in global and total lightning detection. In: Seventh Asia-Pacific international conference on lightning (APL), 2011; 2011. [56] Lojou J-Y, et al. Nowcasting of thunderstorms using VHF measurements. In: Lightning: principles, instruments and applications. Springer; 2009. p. 253–270. [57] Elbaghdady L, et al. One site three dimensions lightning location system using VHF broadband interferometers. J Atmos Electr 2013;33(2):91–105. [58] Elbaghdady L, et al. Correlated one site three dimensions VHF broadband digital interferometer algorithm and phased array radar (PAR) observations. J Atmos Electr 2014;34(1):27–40. [59] Hsueh YH, et al. A vectorized data communication system for localization in Zigbee sensor networks. Adv Sci Lett 2012;9:487–92. [60] Ibrahim, WI, ZA Malek. Time-to-thunder method of lightning distance determination. In: IEEE international conference on power and energy (PECon), 2010; 2010. [61] Amir, A, W Ibrahim, Multi-station short baseline lightning monitoring system. In: 16th Asian conference on electrical discharge, Johor Bahru, Malaysia ; 2012. [62] Aiming, Y, et al. Application of ground electric field in smart power supply against lightning of wireless access base station. In: International symposium on electromagnetic compatibility, 2005 (EMC 2005); 2005. [63] Dudley SEM, Quinlan TJ, Walker SD. Ultrabroadb and Wireless #x02013; optical transmission links using axial slot leaky feeders and optical fiber for underground transport topologies. IEEE Trans Veh Technol 2008;57 (6):3471–6. [64] Khotiaintsev S, et al. Optical intensity-type refractometer for remote measurements via fiber-optic cables. J Lightwave Technol 2012;30(8):1073–9. [65] Schanda J. Optics to solve the information problems of mankind. Measurement 1991;9(1):2–9. [66] Cranch GA, et al. Remotely- interrogated three-axis fiber laser magnetometer. IEEE Sens J 2013;13(3):890–9. [67] Deli Q, Gursoy MC, Velipasalar S. Effective capacity of two-hop wireless communication systems. IEEE Trans Inf Theory 2013;59(2):873–85. [68] Quang-Dung H, Yue G, Tho L-N. Challenges and research opportunities in wireless communication networks for smart grid. IEEE Wirel Commun 2013;20:3. [69] Boquete L, et al. A portable wireless biometric multi-channel system. Measurement 2012;45(6):1587–98. [70] Zhou T, et al. High-precision and low-cost wireless 16-channel measurement system for multi-layer thin film characterization. Measurement 2013;46 (9):3600–11. [71] Goleniewski L. Telecommunications essentials: the complete global source. 2nd ed.Addison-Wesley Professional; 2006. [72] Bucci G, et al. Architecture of a digital wireless data communication network for distributed sensor applications. Measurement 2004;35(1):33–45. [73] Juljaroenkul J, Chantrapornchai C. Wireless system design experience: case study of a manufacturing factory. Int J Multimedia Ubiquitous Eng 2013;8 (2):35–44. [74] Mitchell, B. How fast is a Wi-Fi Network? 12/01/2014. Available from 〈http:// compnetworking.about.com/cs/wireless/f/wirelessspeed.htm〉; 2014. [75] Natkaniec M, et al. A survey of medium access mechanisms for providing QoS in Ad-Hoc networks. IEEE Commun Surv Tutor 2013;15(2):592–620. [76] O’Keeffe M. RF standards for short-range wireless connectivity. Electron Prod (Garden City, New York) 2007;49:11. [77] Nawawi Z, et al. Data transmission system of rotating electric field mill network using microcontroller and GSM module. J Teknol 2013;64(4):109–12. [78] Ionel R, Vasiu G, Mischie S. GPRS based data acquisition and analysis system with mobile phone control. Measurement 2012;45(6):1462–70. [79] Berger, G, et al. Recent progress in lightning studies at the Pic du Midi observatory. In: Seventh Asia-Pacific international conference on lightning (APL); 2011. [80] Kabasa, B, E Rashayi, E Chikuni. Online monitoring of lightning incidence and severity in rural Zimbabwe. In: IEEE international conference on industrial technology (ICIT); 2013. [81] Ahmad H, Sidik MAB, Salam MA. Development of a new lightning surge counter with SMS system. Int J Electr Eng 2012;19(1):29–34. [82] Al-Ali AR, Zualkernan I, Aloul F. A mobile GPRS-sensors array for air pollution monitoring. IEEE Sens J 2010;10(10):1666–71. [83] Shui Ming C, Du Y, Meng F Ling. Lightning data observed with lightning location system in Guang-Dong Province, China. IEEE Trans Power Delivery 2004;19(3):1148–53. [84] Singye, J, et al. Thunderstorm tracking system using neural networks and measured electric fields from few field mills. In: IEEE international symposium on circuits and systems, 2005 (ISCAS 2005); 2005. [85] Qing, M, et al., Research on lightning warning with SAFIR lightning observation and meteorological detection data in Beijing-Hebei Areas. WMO TECO2005 instruments and observing methods report; 2005(82). [86] Mingli, C, L Tao, D Yaping. Experimental study of single-station lightning locating technique. In: Seventh Asia-Pacific international conference on lightning (APL); 2011. [87] Antunes, L, et al. Characterization of lightning observed by multiple highspeed cameras. In: International symposium on lightning protection (XII SIPDA); 2013. [88] Nianwen, X, G Shanqiang. A precisely synchronized platform for observing the lightning discharge processes. In: Asia-Pacific power and energy engineering conference (APPEEC); 2011. [89] Yang S, et al. A digital lightning prototype system: DLPS/HUST. Int J Digital Earth 2013;6(Suppl.2):155–71. M.A.B. Sidik et al. / Renewable and Sustainable Energy Reviews 48 (2015) 710–725 [90] Sidik, MAB, et al., Atmospheric electric field data logger system. J Teknol 2013;64(4):119–23. [91] Wan Ismail, I, et al., Development of single-station early warning lightning alarm system; 2013. [92] Ibrahim, W M Ghazali. An overview of Pekan lightning detection system (PLDS). In: IEEE international conference on power and energy (PECON 2012), Kota Kinabalu, Sabah Malaysia; 2012 . [93] Hassan MK, et al. Lightning strike mapping for Peninsular Malaysia using artificial intelligence techniques. J Theor Appl Inf Technol 2011;34(2):202–14. 725 [94] Jayendra, G, et al. Intelligent lightning warning system. In: Third international conference on information and automation for sustainability, 2007 (ICIAFS 2007); 2007. [95] Mohd Basri MA, Othman MF, Husain AR. An approach to brain tumor MR image detection and classification using neuro fuzzy. J Teknol 2013;61:2. [96] Takayanagi, Y, A Takashi. Lightning observations using Hitachi Lightning Monitoring System (HLMS). In: AGU fall meeting abstracts; 2013.