Monitoring system for photovoltaic plants: A review

Renewable and Sustainable Energy Reviews 67 (2017) 1180–1207

Contents lists available at ScienceDirect

Renewable and Sustainable Energy Reviews journal homepage: www.elsevier.com/locate/rser

Monitoring system for photovoltaic plants: A review Siva Ramakrishna Madeti n, S.N. Singh Alternate Hydro Energy Centre, Indian Institute of Technology Roorkee, Uttarakhand 247667, India

art ic l e i nf o

a b s t r a c t

Article history: Received 7 October 2015 Received in revised form 14 September 2016 Accepted 20 September 2016

The Photovoltaic (PV) monitoring system collects and analyzes number of parameters being measured in a PV plant to monitor and/or evaluate its performance. In order to ensure the reliable and stable operation of any PV system, an effective monitoring system is essential. Moreover, the monitoring system keeps track on various electricity generation indices and fault occurrences. The cost and complexity of existing PV monitoring systems restricts their use to large scale PV plants. Over the past decade, different aspects of PV monitoring systems were reported in wide range of literature. In this paper, a comprehensive review of various PV monitoring systems is presented for the first time. This includes the detailed overview of all the major PV monitoring evaluation techniques in terms of their relative performances. Major aspects of PV monitoring systems which examines in this paper are: sensors and their working principles, controller used in data acquisition systems, data transmission methods, and data storage and analysis. The acquaintance of all these aspects are crucial for the development of effective, low cost, and viable PV monitoring systems for small and medium scale PV plants without compromising on the desired performance. & 2016 Elsevier Ltd. All rights reserved.

Keywords: Photovoltaic system Monitoring system Data acquisition

Contents 1. 2. 3.

4.

5.

6. 7.

n

Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Overview of PV monitoring system . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Photovoltaic system configurations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.1. Grid-connected PV system . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.2. Stand-alone solar PV system . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.3. Hybrid PV system . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.4. Stand-alone hybrid AC solar power system with generator and battery backup . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Characteristics of monitoring system . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.1. Monitoring parameters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.2. Measurement of monitoring parameters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.2.1. Current measurement . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.2.2. Voltage measurement . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.2.3. Solar radiation measurement . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.2.4. Temperature measurement . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.3. Controller . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Major instruments used in PV monitoring system . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5.1. Current sensors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5.2. Voltage sensors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5.3. Solar radiation sensors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5.4. Temperature sensors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . DAQ system . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Methods of data transmission, storage and analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7.1. Methods of data transmission. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

Corresponding author. E-mail address: [email protected] (S.R. Madeti).

http://dx.doi.org/10.1016/j.rser.2016.09.088 1364-0321/& 2016 Elsevier Ltd. All rights reserved.

1181 1181 1182 1182 1182 1183 1183 1183 1183 1184 1184 1189 1190 1192 1193 1193 1193 1194 1195 1195 1195 1196 1196

S.R. Madeti, S.N. Singh / Renewable and Sustainable Energy Reviews 67 (2017) 1180–1207

7.2. Methods of data storage . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7.3. Methods of data analysis. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8. Opportunities for PV monitoring systems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9. Challenges and prospects of PV system monitoring: Current and Future. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10. Discussions. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11. Recommendations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12. Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

1. Introduction The energy demands of almost all the countries around the globe are on rise due to its large scale industrial expansions, increasing population, and continuous growth in energy consumption per capita. It should be noted that major portion of energy requirement is in form of electricity. On the other hand, use of fossil fuel based electricity generation came to saturation levels due to increased environmental concerns and limited resources. Thus, the gaps between the demand and generation in future are to be met by renewable energy sources (RES). In line with this objective, RES such as solar, wind, biomass, micro-hydro, and geothermal are being converted into electrical energy and delivered either to demand centers or utility grids [1–4]. The other motive behind promoting the dependency on renewable is to provide safe, clean and sustainable energy. In tropical countries, solar energy is deemed as the most reliable and viable options among all RES [5]. Owing to the developments in photovoltaic (PV) technologies and various financial subsidies being provided by the government bodies to electrical energy generation sector using PV technology has seen a rapid evolution during the past few decade. This is evident from the increase in cumulative installed PV capacity (MW) of countries participating in International Energy Agency- Photovoltaic Power Systems Programme (IEA-PVPS), from 103 MW in 1992–139795.2 MW in 2014 [6]. More recently, urban population and industrialists have shown their interests in PV energy generation in view of sustainable development. This will ultimately make the PV technology overcome their set back of lower power density by expanding its foot print to urban/populated areas. Therefore, a steep rises in PV systems/ plants are expected in the years to come. At this verge, it is very much essential to develop the technologies, which keep track on PV energy production from a given PV plant and keep up its production in every possible dimension. The objective of such PV monitoring technologies is to predict/sense different undesirable situations, which may plunge the energy production levels from available solar irradiation [7]. The issues that are to be tackled to achieve the desired objective could be optimal control/design, climatic conditions, surrounding objects, and geographical locations etc. It was reported that the annual energy loss due to partial shading is about 10–20% [8]. In addition to this, there will be a production loss associated with the occurrence of each of grid fault. All such issues pertaining to different domains have to be addressed by a single solution. Such an endeavor requires the complete knowledge of meteorological data (climatic conditions, which affects energy production) of the area where the system will be installed [9]. Consequently, it is necessary to develop techniques, which help in estimating the true potential (power) of RES in the installed area in real-time. Moreover, continuous monitoring of PV system(s) health are very crucial to detect the causes, which hamper the desired performance [10]. A comprehensive solution for all these problems is being termed as PV monitoring system, whose job is to maximize the operational reliability of PV system with minimum system costs.

1181

1200 1200 1201 1201 1203 1203 1204 1204

The PV monitoring systems are aimed to provide/report information about the energy potential, energy extracted, operating temperature analysis of different of faults that might occur, and energy loss associated with them. The data being monitored can also be used for early detection/warning, evaluating the climatic changes etc. There has been a growing interest and importance in these issues. Due to which, significant expertise has been devoted in developing the effective, low cost, and viable PV monitoring systems for small and medium scale PV plants without compromising on the desired performance. Each monitoring system would comprise of several commercial products integrated within it. With the rapid increase in different commercial products based on various principles/concepts, it’s very important to examine the operation and characteristics of each of them. The selection of appropriate product for a particular climatic condition is vital for an effective PV monitoring system. Over the past decade, different aspects of PV monitoring systems were reported. This paper presents an overall review on PV monitoring systems covering all the important factors and components associated with it. This paper is organized as follows: Section 2 provides an overview of PV monitoring system. Classification of PV based systems is given in Section 3. In Section 4, the different characteristics of monitoring system are discussed. While major instruments used in PV monitoring system has been reviewed in Section 5. In Section 6, various data acquisition systems used to handle the output data of sensors are presented. Section 7 different methods used for data transmission, storage and analysis have been reviewed and summarized, and Sections 8 and 9 addresses the major challenges and opportunities in PV monitoring systems. The results of analysis are discussed in Section 10. Major recommendations and suggestions have been presented in Section 11. Section 12 summarizes the conclusion of this work.

2. Overview of PV monitoring system The general block diagram of PV monitoring system is shown Fig. 1. The PV monitoring systems can be broadly classified as ground based or space based monitoring systems. The former approach is more prevalent due to its quick response and accuracy in monitoring the PV system health. Thus, it provides a chance to enhance PV system performance by detecting the possible energy losses from changes in operating condition and/or faults, before they have a considerable effect on energy production and/or system health. Major components used in the ground based systems are sensors, which measure the variable in real-time in the monitoring system. In this aspect, space based systems could be economical due to the absence of sensors. The bottleneck for space based systems is their low accuracy in estimation and is greatly affected by the climatic conditions, which is clearly undesirable. Therefore, the scope of this PV monitoring systems review is confined to ground based systems. Another important unit in a PV monitoring system is the signal conditioning unit. This unit performs signal amplification and

1182

S.R. Madeti, S.N. Singh / Renewable and Sustainable Energy Reviews 67 (2017) 1180–1207

3. Photovoltaic system configurations Photovoltaic (PV) systems are mainly classified according to their configurations, functions, and connection topology. Two principle classifications are stand-alone systems and grid connected PV systems. PV systems can be designed to supply DC and AC loads. These systems can also be connected with energy storage systems and other energy sources. Fig. 2 gives a brief classification of PV system configuration.

Solar Panels

Sensors 3.1. Grid-connected PV system

Signal Conditioning

Signal Processing

Data Transmission

Fig. 3 shows the block diagram of a grid connected solar PV system. The major components of this system are PV module, power conditioning unit (PCU), and an on-site distribution panel. PV array produces DC power from the incoming solar radiation using photovoltaic effect. The PCU converts the DC power output from PV array into AC power, according to voltage magnitude, frequency and power quality requirements of grid. An on-site distribution panel provides a bi-directional interface between the PCU output circuit and connected electric utilities. It allows the system to supply AC power to either on-site electrical loads or to utility grid, when PV system output is greater than the load demand. This attribute increases the safety to grid connected PV system by stopping continuous supply of power to grid during down time. 3.2. Stand-alone solar PV system

Data Storage and Data Analysis

Fig. 1. General block diagram of PV monitoring system.

sensor measurement filtering for further processing. The signal conditioning unit also consists of a microcontroller that transmits the outputs of signal conditioning unit to a personal computer (PC) in real-time using an adopted protocol. The PC uses data for analysis, display and storage. Based on the internal analysis and external commands by users, PC delivers the commands to system control unit for future actions.

Fig. 4 shows the block diagram of a stand-alone PV system or direct coupled PV system. It can be designed to supply AC and DC loads. The PV array output is directly connected to load, and is thus called a direct coupled system. There is no energy storage element used in directly coupled PV systems, due to which it can supply energy to the load during sunny hours. For better utilization and to extract maximum power from the available PV array, a maximum power point tracker (MPPT) is used in between the PV array and load. In some other stand alone PV systems, batteries are used as storage elements. Fig. 5 shows the block diagram of a stand-alone PV system with provision of battery to supply energy to AC and DC loads. It is similar to systems without battery except the fact that few additional components are required to provide battery charge stability. The charge controller regulates the output current of a PV array and stops voltage values from exceeding the maximum level for battery charging. The output of charge controller is connected to DC load as well as a battery with a dual DC cut-off switch, which

Solar PV system

Grid connected PV system

Stand alone PV system

Hybrid PV system

Large scale production (without battery)

With Battery (e.g. for houses and industries )

Wind-PV hybrid system

With Battery (Smart Grid concept )

Without Battery (PV water Pump)

PV-Diesel hybrid system

Fig. 2. Classification of PV system.

PV based utilities

Solar Lamp, Solar mobile charger etc.

S.R. Madeti, S.N. Singh / Renewable and Sustainable Energy Reviews 67 (2017) 1180–1207

AC Load

Power Conditioning Unit

Solar PV Module

Solar PV Array

Charge Controller

DC Load

Rectifier

Battery Bank

Inverter

Distribution Panel

Grid Fig. 3. Block diagram of grid-connected solar PV system [11].

Solar PV Module

1183

DC / AC Load Diesel Generator / Wind Turbine

Fig. 4. Direct coupled solar PV system [12].

Solar PV Array

Charge Controller

Generator / Grid Supply

AC Load

Fig. 6. Block diagram of photovoltaic hybrid system [14].

DC Load 3.4. Stand-alone hybrid AC solar power system with generator and battery backup

Battery

Inverter

AC Load Fig. 5. Block diagram of stand-alone PV system with battery storage [13].

disconnects the PV array and load simultaneously during faulty conditions. During proper insolation, PV array output is supplied to load and battery simultaneously. The controller thus ensures that the DC output from PV array is sufficient to sustain the connected load while battery sizing. 3.3. Hybrid PV system Hybrid PV system commonly refers to PV systems integrated with wind turbines, diesel generators, or any other non-conventional or conventional energy sources. Fig. 6 shows the block diagram of a Hybrid PV system. In this system, PV is generally sized to supply a base load demand, where as the alternative source is used for conditions when the load demand reaches its peak value. This helps in achieving low operation and maintenance costs, in addition to a reliable supply. Hybrid systems can also be useful in situations where demand peaks are significantly higher than the base load demand. It makes little sense to size a system to meet demand entirely with solar PV in a case where, say normal load is only 10% of the peak demand. By the same rule, a diesel generator-set sized to meet the peak demand would be operating at inefficient part-load for most of the time. In such a situation, a PV-diesel hybrid would be a good option.

A stand-alone hybrid AC solar power system is functionally similar to PV system, but AC inverters are used in these systems to convert DC into AC. The inverter output contain harmonics for which harmonic reduction filters must be used. In some cases, an AC transfer switch is incorporated which is capable to allow an output from AC type standby generators. These inverters have special electronic devices, which transfer power from generator to the load.

4. Characteristics of monitoring system The main characteristics of any monitoring systems are categorized into following types: monitored parameters, sensors, controller, data transfer mechanism, program development software, and monitoring method. 4.1. Monitoring parameters Due to intermittent nature of solar energy, the power output of a PV system may increase or decrease drastically which leads to increased stress on the grid or sometimes causes power outages [15]. Since, PV achieves high penetration levels on utility grid, compelling it to monitor the parameters for ensuring reliability. An important cogitation of any monitoring system is the choice of parameters to be measured. These parameters are selected according to British Standard BS IEC 61724 [16]. Depending on the type of PV system configuration, a list of parameters is given in Table 1. It can be distinguished as grid-connected and stand-alone PV system. After studying previous works on PV monitoring systems, it is noticeable that the most eminent operational and metrological parameters are solar radiation, temperature, PV voltage and current, while other parameters are configuration dependent. Fig. 7 (a) and (b) shows the locations of different sensors of a typical grid connected and stand alone PV system. The monitoring system consists of numerous sensors, which provide the information of different assets under various conditions. This information can be used by the operators in making decisions related to utilization, replacement, and system reliability.

1184

S.R. Madeti, S.N. Singh / Renewable and Sustainable Energy Reviews 67 (2017) 1180–1207

4.2. Measurement of monitoring parameters

Table 1 Parameters to be measured. PV system type Parameters

Grid connected

Metrological

Electrical

(i) Total irradiance, in the plane of array GT (ii) Ambient temperature (iii) Module temperature (iv) Wind speed (v) Wind direction (vi) Humidity (vii) Barometric pressure

Photovoltaic array: (i) Output voltage (ii) Output current (iii) Output power (iv) Output energy

Stand alone

Utility grid: (i) Grid voltage (ii) Current to utility grid (iii) Current from utility grid (iv) Power to utility grid (v) Power from utility grid (vi) Utility grid impedance Load: (i) Output voltage (ii) Output current (iii) Output power

Operational and metrological parameter measurements are the process of measuring the electrical and physical properties of the PV system and atmospheric conditions where the system is installed. Various techniques used for the measurement of essential parameters of PV monitoring system are discussed below: 4.2.1. Current measurement The information of current flow is essential for healthy monitoring system, in order to improve its stability. Many current measuring techniques are available depending on the type of application and its requirements in terms of precision, cost, size, and bandwidth. Nowadays, for digital monitoring and control system, this current measurement needs to be in digital form which calls for an analog to digital converter (ADC) to be used at the output of a particular current measuring technique. Owing to its small size and low cost, a shunt resistor has been extensively used in power electronics to measure the current. The analog control loops are being replaced by digital control loops to achieve high efficiency, integration level and easy utilization of sophisticated control techniques. The voltage drop across the shunt is very small which

AC power from inverter

Horizontal solar radiation Ambient air temperature

PV module temperature

DC power from Array

Energy to grid

In-plane solar radiation PV array

Inverter

Grid

(a) Grid connected

AC power from inverter

Horizontal solar radiation Ambient air temperature

PV module temperature

DC power from Array

In-plane solar radiation PV array

Inverter

(b) Stand alone Fig. 7. Monitored parameters for the PV system.

Consumer units

S.R. Madeti, S.N. Singh / Renewable and Sustainable Energy Reviews 67 (2017) 1180–1207

1185

Current Measurement

Based on Ohm’s Law A. Shunt Resistor

Based on Faraday’s law of induction

Based on Magnetic Field effect

A. Rogowski Coil

A. Magnetic Field (i) Hall effect sensors

(i) High-performance Coaxial Shunt (ii) Low-cost Surface-Mounted Device

B. Current transformer

(ii) Fluxgate principle

Based on Faraday effect A. Polarimeter Detection Method B. Interferometer Detection Method

(iii) Magneto Resistance Effect a) Anisotropic Magneto Resistance (AMR) b) Giant Magneto Resistance (GMR)

B. Trace Resistance Sensing

Fig. 8. Classification of current measurement based on working principle.

needs to be amplified as it alters the bandwidth and leads to increase in size and cost of device. Moreover, with the increase of power density level in power converters, power losses in shunt resistor become troublesome. Researchers have been seeking an alternative to shunt resistors with accuracy similar to other current measuring methods with low power losses and provision for an ADC. It is important to understand the working principles and technical limitations of various current measuring techniques. These are classified based on different physical principles as shown in Fig. 8. 4.2.1.1. Current measurement based upon of Ohm’s law. Ohm’s law states that voltage drops across two points of a conductor is proportional to current flow through two points. It is basically obtained by using simplification of Lorentz force law.

J = σ ( E + V * B)

(1)

where, J is the current density, σ is material conductivity, E is electric field, v represents the charge velocity, and B is the magnetic flux density acting on the charge. In most cases, the charge velocity is sufficiently small, and can be neglected, then Eq. (1) can be written as:

J = σE

(2)

The relation given in Eq. (2) is known as Ohm’s law and can be used for current measurement. This technique is reliable, simple and cost effective. 4.2.1.1.1. Shunt resistor. A simple approach for current measurement is to use a shunt resistance as shown in Fig. 9. To measure the current flow, voltage across the shunt resistance is measured and construed in form of current. It can be used to measure both dc and ac quantities. This method of current measurement is avoided to measure high currents as considerable power loss occurs [17].

4.2.1.1.2. Trace resistance sensing. An intrinsic resistance is a possible alternative to use as a conducting element in the circuit instead of using a dedicated shunt resistor. It promises to provide a new approach for current measurement with low power loss at very low cost. The resulting voltage drop across the trace resistance is very small because the resistance of copper trace is very low [18]. An amplifier with high gain is required to get a useful output signal. This current measuring technique is limited in its applications due to its limited gain-bandwidth- product, which changes the performance of the method. Only a few publications are available on this technique; hence, it has limited applications. This method may become more popular on all power conversion industries due to its high efficiency in sensing and to the increase power density.

4.2.1.2. Current measurement by means of Faraday’s Law of induction. The current measurement based on operating principle of Faraday’s laws of induction is mainly used in two devices: Rogowski coils and current transformers (CTs). The key feature of these devices are having inherent characteristic of electric isolation between the current being measured and output signal. This facilitates the measurement of currents at high and floating voltage potentials provided a ground reference to output signal. Isolated current measurement is indispensable in most of the applications because the safety standards demand electrical isolations. 4.2.1.2.1. A. Rogowski coils. Fig. 10 shows the schematic diagram of Rogowski coil with a nonmagnetic core material. The working principle is based on Ampere’s and Faraday’s laws. The amperes law defines that path integral of the magnetic flux density B around the closed curve is proportional to the current ic flowing through surface S (enclosed by curve C), as follows:

Current to be measured

Rshunt Voltmeter Fig. 9. Current measurement using shunt resistance.

Fig. 10. Schematic of a Rogowski coil with nonmagnetic core material.

1186

S.R. Madeti, S.N. Singh / Renewable and Sustainable Energy Reviews 67 (2017) 1180–1207

∮c B.⃗ dl⃗ = μoic

(3)

For a theoretical analysis let radius ‘ r ’ of the Rogowski coil is much greater than the cross sectional diameter. The magnetic field intensity around the closed curve when current ic is centered inside the coil is given by:

B=

μoic (4)

2Πr

Applying Faradays law of induction, the induced voltage across Rogowski coil can be given as:

v=−N

NAμo dic dϕ dB = − NA =− dt dt 2Πr dt

(5)

where, ic is the measured value of current, v is proportional to derivative of the primary current ic , A is the cross sectional area of coil, N is the number of turns. In order to yield accurate results, the above equation is further solved by considering infinite input impedance with integration constant k .

vout=−

NAμo 2Πr

k

∫t

NAμo dic .dt + v( 0) = − k ic + vout( 0) out dt 2Πr

(6)

Eq. (6) is only valid to those coils, which are not circular in shape [19]. However, a variation in area of turns and non-uniform winding density around the coil leads to substantial increase in measurement errors [20]. These errors are profound at junction, coil end and near gaps in coil turns. Several companies compensate the error due to gaps, by increasing winding density around the coil. Nevertheless, when primary current wire is installed as far as possible from junction of the coil, this effect can be minimized. It is evident from Eq. (6) that direct currents can also be measured by using Rogowski coil. It is based on the principle of detection of change in magnetic field which is proportional to change in current. The reconstruction of DC component is only possible by knowing the value of current at t¼0, which is represented at v(0)out in Eq. (6). The integrators used in practical applications possess a small input DC offset voltage, which causes change in frequency response, so it should be compensated to have reduced gain at low frequency. These changes further lead to reduce the gain at low frequency. Therefore, practical Rogowski coils are not suitable for measurement of currents at low frequency as depicted in Fig. 11. [19,21,22]. It has been proposed that in order to measure direct currents, the combination of open-loop

Fig. 12. Current transformer with one turn in primary side and multiple turns in secondary.

magnetic field sensor and Rogowski coil can be used as it extents the measurement range. [23]. 4.2.1.2.2. Current transformer (CT). Current transformer (CT) also works on the principle of Faraday’s law of induction similar to Rogowski coil, to measure current. CT usually has single turn in primary side and multiple turns in secondary side, like Rogowski coil. The core is chosen of high relative permeability material on which secondary winding is wound as shown in Fig. 12. The major difference between a Rogowski coil and CT is that secondary winding of the current transformer is loaded with a sense resistor R s through which current is will flow. This current produces a magnetic flux that acts opposite to the flux produced by primary current. Eq. (5) derived for Rogowski coil can be modified as follows:

vs= − N

μμ dϕ d = − NA o r ( ic −Nis) dt lm dt

(7)

where, vs is induced voltage across the coil terminals, N is the number of coil turns, μ o is the permeability of free space, A is coil area, μ r is relative permeability of the material used for core (for air cored coils μr is 1), lm is the magnetic path length, ic represents the current flowing in primary winding, and is is the current through R s.

is =

ic l − 2 m N N Aμoμ r

∫t vs.dt

(8)

The second term in Eq. (8) represents expression of inductance, and is known as magnetizing inductance L m .

is =

Fig. 11. Frequency/current limits of Rogowski coils.

ic 1 − N Lm

∫t vs.dt

(9)

Basic equivalent circuit diagram of CT based on Eq. (9), using a theoretical DC transformer is shown in Fig. 13 by neglecting core loss, stray inductances, and winding resistance. It includes a magnetizing inductance L m , which requires mean voltage applied to transformer winding to be zero, otherwise transformer saturates. The secondary winding capacitance C w limits the bandwidth, especially at high number of secondary turns. The second term in Eq. (9) makes CT unsuitable for measuring

S.R. Madeti, S.N. Singh / Renewable and Sustainable Energy Reviews 67 (2017) 1180–1207

v=

RhIB d

1187

(11) 1 n*q

Fig. 13. Equivalent circuit diagram of current transformer [17].

direct currents. If any DC component is present in primary current, then increase in magnetizing current takes place until full DC component flows through L m. Hence, CT in standard configuration is unable to measure direct currents. When the frequency is very high, then second term in Eq. (9) becomes small, and consequently secondary current is directly proportional to primary current. It can be measured by means of a shunt resistance Rs . This method for current measurement provides various benefits such as low losses, isolation, simple working principle, and no special need for further amplification of the output voltage [24]. 4.2.1.3. Current measurement by using Magnetic Field effect. Faraday’s law of induction does not hold good to measure current, which produces static magnetic field. On the other hand, current measuring devices based on magnetic field effect are capable to sense both static and dynamic magnetic fields. Consequently, these techniques provide a good alternative for current measurement. 4.2.1.3.1. Hall-Effect. The Hall-Effect was discovered by Edwin Hall in 1879. The basic working principle behind this effect is the Lorentz force, which is illustrated in Fig. 14. When a current I flows through a slab of semiconductor or conductive material penetrated by a magnetic field, a voltage is induced perpendicular to both current and magnetic flux density B.

v=

IB nqd

(10)

where, v is the voltage induced across the end of semiconductor, I is the current flowing through conductive material, B is the magnetic flux density, n is the carrier density, q is the charge of current carrier, and d is thickness of the sheet. Eq. (11) can be rewritten as:

where, Rh = and is known as hall coefficient [26]. Eq. (11) shows that induced voltage v is directly proportional to magnetic flux density B . Hall-Effect sensors can be placed in air gap of a magnetic core, which concentrates the flux linking with a current carrying asset. The main drawback in Hall-Effect sensor is presence of offset voltage at output even when magnetic field is zero, known as misalignment voltage. In order to compensate the misalignment voltage and different thermal drift, a special circuitry is required when using Hall-Effect in current measurement [27]. A typical problem associated with the use of Hall-Effect based current measurement devices are that it interferes with the magnetic field leakage arising from close currents when it is placed inside a ferromagnetic core. Nowadays various theoretical and commercial solutions are available for improving accuracy, sensitivity and bandwidth of these devices [28]. Common applications of these devices include motor drives, power conversion systems, radar devices, welding equipment, and in electro-winning industry etc. 4.2.1.3.2. Fluxgate principle. Flux gate technology has been widely used to sense the magnetic fields accurately since 1930s [29,30]. Its basic operating principle is based on utilization of nonlinear characteristic exhibited by the magnetic material between magnetic field intensity ( H) and magnetic flux density (B). Vacquier type fluxgate sensor as shown in Fig. 15 has the arrangement of two parallel rods identical in shape on which excitation winding with same number of turns are wounded in opposite direction, and connected in end to end manner. These two parallel rods are placed in a pickup coil in order to detect the amount of magnetic field produced by these two rods. The amplitude and polarity of induced magnetic filed in two rods depends upon the given external field. If the external field does not exist, then the induced magnetic filed in two rods have same magnitude but opposite polarity. On the other hand, if external field does exist, then the polarity of inducted magnetic field is shifted and voltage inducted in the pickup coil is proportional to difference between the rate of change of flux in two rods.

⎛ dB dB2 ⎞ vs= − 2NA⎜ 1 + ⎟ ⎝ dt dt ⎠

(12)

where, N is the number of turns of pickup coil, A is the area of cross section of a rod. The time deviation of magnetic flux density of each core in the above Eq. (12) can be represented in terms of their permeability μ .

μ=

dBHext ± Ho d(Hext ± Ho )

(13)

Due to nonlinear characteristics of magnetic core, μ depends on the field H = Hext ± Ho as depicted in Fig. 16. It is a combination of

Fig. 14. Schematic representation of Hall-Effect sensor [25].

Fig. 15. The Vacquier fluxgate [29].

1188

S.R. Madeti, S.N. Singh / Renewable and Sustainable Energy Reviews 67 (2017) 1180–1207

Fig. 17. Barber pole structure [34].

Fig. 16. The fluxgate method takes advantage of the fact that the permeability μ of a magnetic core material depends on the applied magnetic field H [32].

demagnetizing effect and properties of core material [31]. Combining Eqs. (12) and (13) gives:

⎛ d( H + H ) d( Hext − Ho ) ⎞ ext o ⎟⎟ vs= − 2NA⎜⎜ μ 1 + μ2 dt dt ⎠ ⎝

(14)

For an external static field, Hext in Eq. (14) becomes:

vs= − 2NA

dHo ( μ 1 − μ 2) dt

(15)

The differential permeability μd is given as:

μd = μ 1 − μ 2 =

dBH + Ho d(Hext + Ho )

−

dBH − Ho d( Hext − Ho )

(16)

The voltage induced across the pickup coil is expressed as:

vs= − 2NAμd

dHo dt

(17)

When excitation field Ho is greater than the provided external static filed Hext , then peak induced voltage is proportional to external static filed, which can be used to measure the magnetic fields. 4.2.1.3.3. Magneto resistance effect (MR). Magneto resistance is the property of magnetic material in which the value of electrical resistance will change as a function of external applied magnetic field. The devices based on these materials are used to measure magnetic fields. The main application of this technique is to read a magnetic recording head, but now they are used for other applications like intensity of electric current measurement. There are various physical effects that can cause a change in resistance in materials under magnetic field. Among these, only two MR effects are most popular in measuring the electric current which are discussed here: 4.2.1.3.3.1. Anisotropic magneto resistance (AMR) This effect was discovered by Lord Kelvin in 1856, and effect is related with hall elements i.e., direction and magnitude of applied magnetic field. When a ferromagnetic material, such as Permalloy carries a current I it experiences a change in resistance that depends upon angle between the direction of magnetization M and direction of current flow [33]. The resistance value becomes minimum when flow of current is perpendicular to magnetic field and maximum when current is flowing parallel to magnetic field M . The AMR effect is sensible to the direction of magnetic field when current I is forced to flow at an angle of 45° to the direction of magnetic field through a deposited aluminum bars over a Permalloy magnetic field. This structure is called as barber poles. The main drawbacks of an AMR are the permanent change in the direction of

initial magnetization of Permalloy ferromagnetic material due to strong external magnetic field, which produces permanent measurement error till new Permalloy re-magnetization is achieved in proper orientation as shown in Fig. 17. The bandwidth of these sensors is limited by the frequency range of conditioning circuit used for signal amplification. 4.2.1.3.3.2. Giant Magneto Resistance (GMR). Gruenberg and Fert discovered the effect of Giant Magneto Resistance (GMR) [35,36]. GMR effect is an alternative technique to sense the static and dynamic magnetic fields. It describes that when nonmagnetic and ferromagnetic materials are exposed to a magnetic field, then there is a greater change in its electrical resistance, upto 12.8% at room temperature. The magnetic field in case of AMR effect directly affects the electrical resistance which leads to change in resistance upto 2–4% as compared to GMR effect [37]. Thus it is possible to detect the change in electrical resistance to measure magnetic field four times greater than that measurable with the GMR devices. This skill has been widely used to ameliorate the performance to read a magnetic recording head. Owing to its high sensitivity, the GMR effect sensors can be used to detect very small currents. This technology is cheaper; and hence, mass production using standard semiconductor technology is possible. Nevertheless, this technology has major drawbacks including distinct thermal drift, high non-linear behaviour and permanent change in its behaviour due to strong external magnetic fields. 4.2.1.4. Current measurement by means of Faraday Effect. Faraday discovered that by applying a magnetic field parallel to direction of light propagation, a circular birefringence can be introduced in the material [38]. For negligibly small values of this birefringence, the rotation of plane θ of linearly polarized light is proportional to the applied magnetic field H along path s as given in Eq. (18)

Ө=Vk

⃗ ⃗ ∫ H.ds

(18)

where, Vk is a constant of proportionality known as Verdet constant. Employing this effect in optical current devices, direct currents over 100 kA can be measured easily. These devices provide excellent electrical isolation and require negligible amount of space and energy [39]. Techniques developed on this principle have been discussed here. 4.2.1.4.1. Polarimeter detection method. Fig. 18 shows the schematic representation of polarimeter detection method. This method uses linearly polarized light passed through a fiber optical coil, which carries the current to be measured. The rotation of polarized light can be related to the magnitude of current by following relation:

Ө=VkNic

(19)

where, θ is the angle of rotation, Vk is the proportionality constant, N is the number of turns in coil and ic is the magnitude of current to be measured. The advantage of using a fiber optic coil is that its measurement accuracy is not affected by stray magnetic fields. The

S.R. Madeti, S.N. Singh / Renewable and Sustainable Energy Reviews 67 (2017) 1180–1207

1189

Fig. 18. Polarimeter detection method.

Fig. 19. Beam splitting in polarimeter at 45°. Fig. 20. Schematic of a Sagnac interferometer.

output light is polarized at 45° to the original polarization by a polarizer in the analyzing circuit. This output light intensity is given by following relation:

Id =

Io ( 1 + sin2Ө) 2

(20)

where, Id is output intensity and I0 is input intensity and θ is the angle of rotation. For small rotation, sin2θ can be linearized. Wollaston prism is a polarizing beam splitter that can be used to check the dependency of output intensity on input intensity by set at 45° to split the beam equally, as shown in Fig. 19. Thus the ratio between difference and sum of output I1 and I2 of two detectors is calculated as:

S=

I1 − I2 =sin2Ө ≈ 2VNic I1 + I2

(21)

. From Eq. (21) it is clear that output signal S is independent of input intensity I0 [40]. The linearity in this method is restricted to small value of Ө as sin Ө can be approximated to Ө for large arguments. Birefringence occurs due to bending of optical fiber cable which reduces the accuracy. This problem of bending stress can be mitigated by using bulk glass [41]. 4.2.1.4.2. Interferometer detection method. This method employs two counter propagating light beams to estimate the Faraday Effect. This method is superior in terms of scale factor stability, zero point stability and measurement range as compared to polarimeter detection method [42,43]. In this method, a Sagnac interferometer is used to linearly polarize and split a light beam in two equal parts, which are then circularly polarized using quarter wave retarders ( λ /4) [44,45]. When light beams come out of the coil, they are turned back into linear light; due to Faraday Effect, current ic becomes proportional to phase difference. The circular polarized beams are altered by Faraday Effect such that one beam travels faster than other. This results in a phase difference between two light waves, which can be measured in terms of magnetic field or current. The phase shift and its relation to current can be described by following Eq. (22) [45]:

∆ϕs = VkN

⃗ ⃗=2V Ni ∮c H.ds k c

(22)

where, ∆∅s is the phase difference, ic is the current to be measured, H is the magnetic field intensity, N is the number of optical coil turns and Vk is the Verdet proportionality constant. The schematic representation of this method is shown in Fig. 20(a) and (b) for open loop and closed loop Sagnac interferometer. The phase shift is measured in open loop Sagnac interferometer by interference of linear polarized light. The interfering light beams have a phase shift of 180°, and hence cancel each other. The light intensity obtained from the interference can be given by following relation:

Id =

Io 1 + cos∆ϕs 2

(

)

(23)

where, I0 is the input light intensity [40]. This relation is approximate as the losses in components have not being considered. A major disadvantage of this method is that it has very small sensitivity when phase shift approaches zero. To tackle this problems, a periodic phase modulation is provided, which uses the ratio of first and second harmonic amplitude to generate linear output [39,45]. Closed loop utilizes the frequency shifter to compensate phase shift till beams are in same phase. Phase shift is directly related to frequency control signal, which is linear over wide range as compared to open loop system [39]. Feeding light into both ends reduces the bending stress interior of the fiber but it is still susceptible against vibrations and thermal drift. Temperature sensors can be used to contain the thermal drifts [45–47]. A sensing accuracy of 0.1% can be achieved by using these techniques for current measurement. 4.2.2. Voltage measurement Depending on the type of PV system configuration, voltage measuring devices are used to measure various voltage levels like low, medium, high voltage utility assets across various points such as PV panel output, inverter output, cables, switchgears, transformers, conductors and capacitor banks for reliable and efficient operation. In case of grid connected PV system, the voltage level is expected to vary which may increase the stress on existing grid assets due to intermittent nature of solar energy. It is expected that the electricity consumption may be increased upto 33% by

1190

S.R. Madeti, S.N. Singh / Renewable and Sustainable Energy Reviews 67 (2017) 1180–1207

(iv) Algorithm for voltage sensing is complex which requires considerable computational time [57,58].

R1

VPV

Further, these voltage sensors are costly and are developed for revenue-grade metering applications where the error in measurement is in the range of 0.1–1% [60]. Thus, there is a need to develop an economical voltage sensor with high accuracy.

R2

VOUT

Fig. 21. Simplified potential divider with voltage conditioning [7].

2040, with increasing mean annual of 0.9% per year [48]. So the operating performance and reliability of a grid connected PV system will reduce if the grid assets were not designed for such increase in system stresses. Therefore, voltage measurement of utility assets has enormous value as it allows utilities to monitor the assets performance and to detect system outages [49,50]. A resistive potential divider is generally used for direct voltage measurement in which a voltage divider referenced to ground is created by connecting two resistances in series. The value of one resistance should be very high compared to other. The input voltage is applied across the two resistances and output voltage is measured across low value resistance, which is connected to conditioning circuit. This brings input voltage level to measurable range. A typical potential divider circuit for DC voltage measurement with conditioning is shown in Fig. 21. where, R2 is selected as very low compared to R1, VPV is the PV output voltage and VOUT is the measured voltage. The output voltage equation for this circuit can be written as:

VOUT R =1+ 2 VPV R1

(24)

Traditionally, potential transformers (PT) and capacitive coupled voltage transformers (CCTV) are used for medium and high alternating voltage measurement on utility network [51]. A high value of PT and CCTV requires oil to provide high insulation and cooling, which increase the cost and regular maintenance of the measuring system. Electro-optical voltage sensors (EOVT) are also used in some other utility networks, but due to their limited life span and higher cost, its utilization is limited [52]. In present research, new technologies are presented for voltage measurements which make use of a floating sensor for measurement of high voltage asset due to which the insulation requirement is very less. A donut shaped overhead line voltage sensor was presented by authors in [53], which was used to monitor voltage though electric field around the conductor but it had a complex design. In [54], authors presented the same approach for hexagonal structure. Authors presented a cylindrical shaped sensor for voltage measurement in [55] by displacing a multiple current sensors to cancel voltage disturbances. However, the computational effort for this voltage sensing algorithm is high. Others research works related to voltage measurement by using the same principle are presented in [56–59], and these works are suffering with at least one of the following drawbacks. (i) Field calibration of sensors required to find error in measurement is more expensive. (ii) Increased cost due to complex and bulky design. (iii) Depending on the type of configuration and their application, sensors are constrained [55,59].

4.2.3. Solar radiation measurement In order to estimate the potential of solar PV power generation, measurement of solar radiation at a particular site is essential. This section discusses various instruments used in solar radiation measurement. Radiometer is a device, which is used to measure the electromagnetic radiation. A radiation sensor is the key component in a radiometer. Various sensors are used in solar instruments, which work on following principles: thermo-electric, calorimetric, thermo-mechanical, and photoelectric [61]. 4.2.3.1. Thermo-electric. It consists of two dissimilar metallic plates whose ends are connected together. When two junctions are at different temperature, an electromotive force (emf) will be induced which is proportional to the type of material used and temperature difference [62]. This device is used in radiometry to measure solar radiation by exposing one plate to incident radiation while other one is shielded from it. 4.2.3.2. Calorimetric. This device consists of a high conductivity metal, which is coated black paint. When an incident solar radiation falls on the metal, the radiant energy is converted into heat and can be measured by various means [63]. The heat gained by the body of calorimeter body Q̇ with small increment of temperature ∆T and V̇ of its volume is given as:

̇ ( T )c ( T )∆T Q̇ = Vh av p av

(25)

where, h represents the heat transfer coefficient, heat capacity at constant volume is denoted by Cp , and Tav is the average temperature. 4.2.3.3. Thermo-mechanical. When a bimetallic strip is heated, having two metals of unequal coefficient of thermal expansion, it leads to bending of bimetallic strip. In such instruments, two metallic strips are present, having one end fixed while other is free. The fixed end is coated with reflectivity material and free end is coated with highly absorptance black paint. These two strips are insulated from each other to prevent heat flow [64]. When an incident solar radiation falls on the blackened strip, it causes bending of a bimetallic strip. This bending can be used as the basis for measurement of solar radiation. 4.2.3.4. Photoelectric. Photovoltaic instruments are widely used for measurement of solar radiation among different photoelectric devices. The photoelectric effect is based on the formation on p-n junction semiconductors. In such semiconductors, two atoms, one having an excess electron while the other being deficient is combined together such that a hole is created and excess electron is free to move. The electron moves within the semiconductor leaving a hole at its previous location. This movement of free electron leads to flow of current in photovoltaic devices. The main drawback of these devices is that the spectral responses in infrared and red portions are strong [65]. These instruments are faster in response and economical. 4.2.3.4.1. Measurement of direct irradiance: Pyrheliometers. An instrument used for the measurement of direct solar radiation flux is known as Pyrheliometer. This instrument is normally connected to an electrically driven equatorial mount for the sun, where two-

S.R. Madeti, S.N. Singh / Renewable and Sustainable Energy Reviews 67 (2017) 1180–1207

1191

Fig. 22. (a) Schematic of a Pyrheliometer, (b) A Hukseflux DR01 first class Pyrheliometer [67].

axis sun tracking mechanism is employed. A thermopile multi junction detector with a quartz window protection is placed at the bottom of a collimating tube as shown in Fig. 22(a). In order to minimize the fluctuations in ambient temperature, a compensator is provided, and a black paint is coated to detector to absorb solar energy with wavelength range of 0.280–3 mm [66]. The aperture angle of a Pyrheliometer is 5°. It measures solar radiations coming from the sun within a limited circumsolar region but excludes all the diffused radiations coming from the sky. To give an instantaneous value of direct beam radiation, a read out device can be used. A first class pyrheliometer has been shown in Fig. 22(b). 4.2.3.4.2. Measurement of global irradiance: Pyranometer. An instrument, which measures global solar radiation with a solid angle of 2Π on a flat surface is known as Pyranometer. A schematic representation of pyranometer is shown in Fig. 23(a). It consists of two glass made hemispherical transparent covers and white disk, which limits the acceptance angle to 180°. The spectral sensitivity of the instrument lies in the wavelength range of 0.29–2.8 mm [66]. It can also be used to measure diffused solar radiation if contribution of direct beam component is eliminated. In order to ensure continuous shading on pyranometer, a small shaded disk is mounted on the automatic solar tracker which can prevent direct solar radiation reaching the sensor. The arrangement is shown in Fig. 23(b). In this instrument, most common sensing elements are based on thermo-mechanical, photovoltaic and thermoelectric principles. Unlike pyrheliometers, the sensing elements are flat in shape. Based on principle of thermoelectric, the Eppley Laboratory manufactures black & white and spectral precision type of pyranometers. In case of black & white type pyranometer, a radially wire-wound plated differential thermopile kind of detector is used

in which three separate black segments are coated with 3 M black, and three white segments coated with barium sulphate [69]. This device is thus called as “Black & White” pyranometer. In order to compensate the effect due to ambient temperature fluctuations, an inbuilt circuit was fitted. The Eppely black & white pyranometer is shown in Fig. 24 which transmits radiation uniformly from 0.285 to 2.8 mm. The Eppley precision spectral pyranometer consists of a spherical multi junction thermopile, and sensing material in this instrument is coated with a highly absorptance paint of all wavelengths such as Parson's black lacquer. Two semi-circular covers of WG295 Schott glass was fitted to this instrument and outer cover can be changed by semi-circular Schott glass filters, which transmit radiation uniformly within a specific frequency bands, hence called as precision spectral pyranometer. This is more precise than the black & white pyranometer. Solarimeter is a popular pyranometer manufactured by Kipp & Zonen and consists of Moll thermopile detectors [70]. A number of commercial pyranometers are available based on principle of thermo mechanical which are generally called Pyranograph or Robitzsch actionograph. In this case, the free end of bimetallic strip gets magnified and starts moving due to mechanical linkages and is recorded by pen on a drum. This instrument is independent of external power sources because the drum is powered by spring. Hence, a remote area installation is possible. A typical bimetallic actionograph type pyranometer is shown in Fig. 25. A bimetallic actionograph type pyranometer is generally more popular due to various benefits like portability, simplicity, and sturdiness and accuracy. There are varieties of commercially available silicon photovoltaic cell type pyranometers. These have low accuracy but fast

Fig. 23. (a) Representation of a pyrheliometer, (b) LPPYRA 12first class pyranometer [68].

1192

S.R. Madeti, S.N. Singh / Renewable and Sustainable Energy Reviews 67 (2017) 1180–1207

calibration of photovoltaic pyranometers is a problem because the complete solar cell response is obtained within a narrow spectral band.

Fig. 24. Eppley black and white pyranometer [61].

4.2.4. Temperature measurement Major component of solar radiation absorbed by PV module does not reciprocate as electrical energy but leads to increase in module temperature, and thus reduces its overall efficiency. The temperature of PV module depends on various parameters such as thermal dissipation, packing box material, amount of solar radiation absorbed, and environmental conditions of modules like wind speed and ambient temperature [74]. The band gap of a solar cell gets reduced if the temperature of PV module increases. This leads to increase in short circuit (SC) current about 0.1%/°C, and decrease in open circuit (OC) voltage of approximately 2 mV/°C [75]. Thus it can be said that few parameters of PV module get strongly affected with the module temperature. Hence, module temperature is an important parameter to measure in order to predict the performance of PV system. Various temperature sensors based on different principles used for measuring temperature in PV applications are discussed as follows: 4.2.4.1. Thermocouple. One of the widely used temperature sensor is the thermocouple. It is very strong and economical which can operate over wide temperature range. Fig. 27 shows a thermocouple circuit. When two dissimilar metals having different thermal coefficients are joined together with a common junction, an emf develops across its junction on heating. This effect is known as Seebeck effect, which was discovered by Thomas Seebeck in 1821 [77]. With respect to change in temperature, the induced voltage varies nonlinearly, but for small change in temperature it can be considered to vary linearly as shown in Eq. (26).

Fig. 25. Robitzsch (bimetallic) pyranograph [71].

(26)

∆V ≈ S∆T

where, ∆T is the change in temperature, S is the Seebeck coefficient, and ∆V is the change in voltage. 4.2.4.2. Resistance Temperature Detector (RTD). RTDs are used as a standard device for measurement of temperature. It has various advantages like high stability, linearity and accuracy close to 70.1 °C over a large range of temperature. It works on the basis of change in resistance due to change in temperature. If temperature rises, then electrical resistance of the metal increases and conversely, if temperature decreases the resistance also decreases. A small diameter coil is the sensing element of an RTD, which has high purity and is usually made of platinum, nickel or copper. Such a configuration is known as wire-wound element. Standard Platinum Resistance Thermometers (SPRTs) is the highest-accuracy RTD and also fragile as shown in Fig. 28.

Fig. 26. Spectral response of silicon cell against spectral irradiance [73].

response [72]. Without external power supply they can provide an output signal and are cheaper as compared to thermopile instruments. Measurement inaccuracies originate from the spectral-selective characteristics of PV cells. Fig. 26 shows the spectral response of photovoltaic sensors, and it can be observed that comparatively sharp peaks occur near 1.0 mm with sensitivity range 0.4–1.1 mm. Effect in global spectral irradiance due to different atmospheric parameters are also shown in the same diagram. The

4.2.4.3. Thermistor. Thermistor is a device in which resistance is a function of temperature. Unlike RTD, thermistor is a ceramic semiconductor, which has low temperature coefficient of resistance 0.4–0.5% /°C) [79]. A thermistor may have a large positive temperature coefficient (PTC device) of resistance or large negative temperature coefficient (NTC device) of resistance depending upon the material used.

Fig. 27. Thermocouple circuit of materials A and B [76].

S.R. Madeti, S.N. Singh / Renewable and Sustainable Energy Reviews 67 (2017) 1180–1207

Fig. 28. A standard platinum resistance thermometer [78].

The relation between the resistance versus temperature of a thermistor is in the form: ⎡

⎤

RT = R Tre B⎣ ( 1/T) − ( 1/Tr)⎦

1193

Engineering Workbench (LabVIEW) data acquisition software. Above described methods have common characteristics that are a microcontroller and data-logging unit is used to measure the signal of interest. This collected data is transmitted to PC through RS-232 serial port. However, for data-acquisition at high sampling rate and advance controlling, the data transmission through RS232 is not enough and it is necessary to use specifically designed buses. Koutroulis [85] developed a computer-based DAQ in order to eliminate the above mentioned limitation. This approach consists in substituting the microcontroller and commercial data-logging unit by a commercial available DAQ card as shown in Fig. 31. The collected data are transmitted to PC through PCI bus for future data processing.

(27)

where, RT is the thermistor resistance at temperature T , R Tr represents the resistance at reference temperature, B is a constant that depends on the thermistor material, T is the thermistor temperature, and Tr is the reference temperature, usually 25 °C. 4.2.4.4. Silicon Temperature Sensors. Integrated circuit (IC) temperature sensors are different from conventional sensors in many ways. An IC temperature sensor works in the temperature range of 55 °C to þ150 °C. While some devices go beyond this range, others stick to this range owing to package or economy constraints. Looking from the functional perspective, an IC temperature sensor includes efficient signal processing circuitry within the package [80]. Cold junction compensation and linearization circuits are not required in this case. Comparator and ADC circuits are also not required unless extremely specialized circuit requirements are there, because such functions are already inbuilt in commercial ICs. 4.3. Controller Data Acquisition (DAQ) system refers to a controller that is employed widely in RES to collect data from various sensors of a PV plant, and then this data is sent to central computer for further process and controls the data. Data acquisition is the process of gathering information from real world in analog form and digitalizes the signal for presentation, analysis and storage on a personal computer. This process involves several stages including sensors, signal conditioning and ADC. The basic block diagram of a simple DAQ system is shown in Fig. 29. Different DAQ used for monitoring the performance of PV water-pumping system [82] and battery charging [83] is shown in Fig. 30(a). The set of sensors are connected to ADC interface with microcontroller unit, which records the sensor data. The collected data are stored in EPROM, with an RS-232 series port, these collected data is transmitted to personal computer (PC) and processed. A different approach has been proposed by Wichert et al in [84], shown in Fig. 30(b). A commercial data-logging unit has been used instead of ADC and microcontroller for measuring the operational and metrological parameters of a hybrid photovoltaic– diesel system. Like in previous case, the collected data is transmitted to personal computer with a RS-232 serial port and is further processed by using Laboratory Virtual Instrumentation

Analogical signal

Sensor

5. Major instruments used in PV monitoring system The instruments used to monitor the above mentioned parameters having range from a low range to wide range of possibilities. Proper selection and function of a monitoring system requires that the user should understand the capabilities and functional limitations of instrument, its response to environmental and operational variations and data analysis for specific objectives. Depending on the objectives and monitoring location, the instruments are selected. Major instruments, which are used in PV monitoring system to measure the above mentioned parameters with high precision and accuracy, are listed below: (i) (ii) (iii) (iv) (v) (vi) (vii) (viii) (ix) (x) (xi) (xii)

   

Current sensor Voltage sensor Solar radiation sensor Temperature sensor Anemometer wind speed sensor Hygrometer sensor Barometer pressure sensor DAQ system Data transmission instrument Data storage and analysis tool or instrument Software tools Miscellaneous

Analog voltmeter, ammeter Connecting wires Oscilloscope Power supply

5.1. Current sensors Current sensors play a crucial role in PV monitoring system and are necessary for the purpose of control and protection. It is a device that detects the current flowing through the measured path and converts it proportionally into measurable voltage [86]. A simple to complex range of current measurement is possible with the current sensors. Conventional current measurement requires a wide bandwidth (typically from DC to 100 kHz) and galvanic isolation tends to increase the complexity and cost of measuring system. Hence, current measurement using current sensors are

Signal Conditioning

A/D Converter

Fig. 29. Block diagram of a DAQ system [81].

Digital signal

1194

S.R. Madeti, S.N. Singh / Renewable and Sustainable Energy Reviews 67 (2017) 1180–1207

Sensors

A/D Converter

EPROM

Microcontroller RS-232 A microcontroller-based system

Weather sensors

RS-232

Data logger unit

RES Plant a data logging unit connected to a PC Fig. 30. Data-acquisition architectures for RES systems.

National Instruments Connector Block

Sensor Interface Electronic Circuits

WINDOWS 98, LABVIEW

National Instruments DAQ CARD

Weather & RES operation Monitoring sensors

Hybrid Photovoltaic/ Wing generator system

Fig. 31. Architecture of a computer based DAQ system.

advantageous compared to conventional methods. It has enormous applications like over current protection, condition monitoring, current regulation, etc., Comparison of the performance characteristics of different current sensors available in market is listed in Table 2:

5.2. Voltage sensors A voltage sensor is a device that measures DC and/or AC voltage levels. It receives voltage signal as input and provides output in form of analog voltage or current signal. It is a self powered device [87]. Electrical voltage sensors vary in terms of optional features, environmental conditions and performance specifications. It has variety of applications like power failure detection, power demand

S.R. Madeti, S.N. Singh / Renewable and Sustainable Energy Reviews 67 (2017) 1180–1207

1195

Table 2 Comparison of the performance characteristics of different current sensors. Current sensor

Model

Measurement Range ( A )

Temperature Range (°C)

Cost

Sensitivity

Linearity

Integratibility

Shunt

INA28x series SMD 2-tab LTC6102

 14 to þ 80 15 15

 40 to þ125  65 to þ250  40 to þ125

Low

mV/A

Very good

Excellent

Rogowski

XH-SCT-Tx series Shinhom-391188-1 MRS04

100 to 100k 1 to 100k 1 to 100k

 20 to þ60  20 to þ80  25 to þ125

Low

mV/A/μs

Very good

Excellent

CT

CT1052 CT12 CT1267-RC

0.25 to 20 1 to 300 0.25 to 40

 10 to þ 60  40 to þ85  40 to þ70

Medium

1 V/A

Fair

Good

Hall

RS 650-548 CSLA series CSNP661 CLSM-50 ACS712 ABL series

1m 1 to ± 100 1 to ± 90 1 to ± 400 1 to 15 1–16 m

 40  25  40  25  40  50

High

10 Gauss

Poor

Fair

Medium

10  2 Gauss

Fair

Excellent

GMR

Table 3 Comparison of different characteristics of various voltage sensors. Voltage sensor

Model

Measurement Range ( V )

Temperature Range (°C)

Cost

Voltage divider

ACPL-C870000E DVL 1000 CE-VZ0232MS2-0.5 DARE DC

3 to 5.5 (DC)

 40 to þ 105

Low

50 to 2000 (DC) 3 to 75 (DC)

 40 to  85  40 to þ 50

3 to 500 (DC)

 40 to þ 85

control, load sensing, etc., the following Table 3 offers a comparison of different characteristics of various voltage sensor types. 5.3. Solar radiation sensors The solar radiation sensor measures global solar radiation, which is the sum of beam and diffuse and reflected solar radiation. This sensor converts the incident solar radiation into electric current that can be measured by various means [88]. This instrument is very crucial in order to test the performance of the PV modules. Conversely, the operating temperature of PV module is highly influenced by the radiation level. Table 4 provides the information of different solar radiation sensors.

to to to to to to

þ125 þ85 þ85 þ85 þ80 þ150

thermocouple, used to measure temperature by means of electrical signal. It responds instantly and gives results with good precision [89]. The major concern when measuring the temperature of any medium is that there should not be any influence of measuring device on the medium it is measuring. Measurement errors and steam-effect can be reduced by proper selection of sensor size and lead configuration. Some of the commercially available temperature sensors and their differences in performance characteristics are given in the Table 5. As mentioned earlier, the characterization of PV module depends on the observation and measurement of operational and metrological conditions. Thus the collection of variables that affect the performance of modules such as solar radiation, temperature, voltage and current of PV module among others are essential. Analytical procedures and several test methods for characterizing the electrical performance of PV modules are proposed by King [90]. There are only few study noticed in literature describing the design criteria for developing the monitoring system. Hence, in order to design an effective monitoring system, the first step is to study different solutions proposed by the authors for their own system. Table 6 gives the details of various sensors used in different studied PV monitoring systems.

6. DAQ system 5.4. Temperature sensors A

temperature

sensor

is

a

device,

typically

RTD

or

Table 4 Different characteristics of various solar radiation sensor types. Solar radiation sensor

Model

Measurement Range ( W/m2 )

Temperature Range (°C)

Cost

Pyranometer

CS-300L SP230-L LI200X-L LP02-L CMP3-L CMP6-L CMP11-L CMP21-L CMP22-L

0 0 0 0 0 0 0 0 0

 40  40  40  40  40  40  40  40  20

Low

CHP1-L MS-56 SHP1

0 to 4000 0 to 4000 0 to 4000

Pyrheliometer

to to to to to to to to to

1750 1750 3000 2000 2000 2000 4000 4000 4000

to to to to to to to to to

þ 70 þ 70 þ 65 þ 80 þ 80 þ 80 þ 80 þ 80 þ 50

 40 to þ 80  40 to þ 80  30 to þ 60

Medium

High Very high Very high

Controllers play a vital role in all the monitoring systems and are used to handle the output data of sensors. Therefore, appropriate selection of suitable controller is of utmost important. Many researchers have used microcontroller, data-logger, DAQ card and module as controllers to gather signals from sensors and digitize the signal for storage, analysis and presentation on a personal computer (PC). Depending on the type of PC technology, a variety of DAQ systems provides flexibility for test, automation and measurement applications. Common examples of such systems are- Peripheral Component Interconnect (PCI), PCI Express, PCI eXtensions for Instrumentation (PXI), Personal Computer Memory Card International Association (PCMCIA), Universal Serial Bus (USB), Ethernet, and wireless data acquisition. Microcontroller and data-logger are cheaper in comparison to DAQ cards and modules. These are easily programmable, and have been used earlier in many studies [85,101,107,110,112,121]. A commercially available data-logger controller has been used in [84], to acquire a set of operational and metrological parameters of a hybrid photovoltaic– diesel system. However, compared to DAQ card, a data-logger unit

1196

S.R. Madeti, S.N. Singh / Renewable and Sustainable Energy Reviews 67 (2017) 1180–1207

Table 5 Performance characteristics of different temperature sensors. Temperature sensor

Model

Temperature Range (°C)

Cost

Linearity

Thermocouple

THM-J series TTD25-B TSD25-N THM-K series

0 to 482  17.8 to 148.9 32 to 212 0 to 1149

Low

Most types non-linear

RTD (PT-100)

RTD1-Cxx-01 series RTD1-Cxx-02 series RTD1-Hxx-01 series RTD1-Rxx-01

 50  50  50  40

300 300 300 85

Wire-wound – High

Fairly linear

Thermistor (NTC)

QT0805 series QT1206 series T020 series T100/E100 T200/E2oo

 65 to  55 to  40 to  50 to 0 to 70

þ150 125 125 150

Low to moderate

Exponential

Semiconductor

LM 135 LM 235 LM335

 55 to 150  40 to 125  40 to 100

Moderate

Linear

to to to to

has lack of flexibility and cannot be used for renewable energy system control [122]. In case of microcontrollers, an interface circuit is used to measure the sensor output. Sensors generate signals, which are often very difficult to measure directly due to noisy environment or because of extremely high or low level signals. Hence, a signal conditioning circuit is essential to increase the accuracy of a data acquisition system. For digital monitoring and control systems, operational and metrological parameters need to be in digital form. This calls for an ADC connected to output of a particular sensing technique. This accuracy also depends on the resolution of an ADC. For small size and low cost PV systems, a microcontroller with 8-bit ADC was found to be sufficient in [123]. However, for large and long time period PV systems, better microcontrollers such as a 10-bit and 12-bit ADC were used. Due to its high resolution, the measurement accuracy of ADC is not affected and a measurement error is governed by sensor accuracy. Table 7 gives the summary of controller techniques used by the researchers.

7. Methods of data transmission, storage and analysis The process of data transmission is important in PV monitoring systems. A number of previous works, related to data transmission have been reviewed and summarized in Section 7.1. Data storage is also an important factor since it can be used for performance analysis and future reference. Hence, several methods related to data storage have been presented in Section 7.2. 7.1. Methods of data transmission A physical transfer of data from one point to another through a proper channel is known as data transmission [126]. Every communication system consists of transmitter, receiver and a transmission channel. The transmitter prepares the received information from sensors, for transmission through proper channel. This transmitted information is finally detected and transformed by receiver in order to record, visualize, and analyse further. Some channels used by researchers to transmit data are: (i) wired communication, which has two subcategories, (a) coaxial cable [127–129] and (b) fiber-optical cable [130–135], (ii) wireless communication [136–139], and (iii) power line communication (PLC) [140–143]. Coaxial cable has low resistance, low error rate and good bandwidth. It supports multiple channels and various services

Film - Low

including data, voice, video and multimedia. Its data transmission speed is as high as 10 Gbps. It consists of solid copper conductor surrounded by a dielectric insulating layer and is enclosed by a metallic shield. Generally, the voltage is applied to centre conductors to carry the useful signal by keeping shield at ground potential. The benefits of using this coaxial design are that the leakage of electrical and magnetic fields is less which in turn reduces the interference effect. It is able to carry weak signals without any interference from the environment, but it has a limitation to its length and its deployment i.e. bus topology [144]. The bus topology is prone to noise, congestion and security risks. However, it is suitable for data transmission of long distance with high speed and less losses. Fig. 32 shows the example of coaxial cable connection in a PV monitoring system. A Fiber-optic cable consists of thin strands of glass in its architecture to carry light. The light pulse travels through core of the fiber which allows highest data transmission speed. Generally, the speed of data transmission of a fiber-optical cable is from 100 Mbps to 200 Mbps. It offers long distance terminals with high bandwidth to transmit data as compared to copper wire [145]. These cables can transmit data over several kilometers without fading. Fiber optic cables are fragile, expensive and their installation is a tough task. Another option for data transmission is Wireless Local Area Networking (WLAN). It covers a wide area of approximately 2000 ha [146]. It has flexibility in data transmission (without any radio coverage area it can transmit data) and additionally the nodes in WLAN can communicate without any future restrictions. Ad-hoc wireless networks allow communication without planning, which makes them better as compared to wired networks [147]. Various 802.11 WLAN standards and their corresponding maximum speed are given as: 802.11 n standard provides speed upto 300Mbps, 802.11 g and 802.11a provide speed upto 54Mbps, and 802.11b provides speed upto 11Mbps. However, in reality, the performance of WLAN is not as good as theoretical values. It can only achieve upto 100Mbps, 20Mbps, and 5.5Mbps respectively [148]. There are few disadvantages, such as lower band width, lower quality of service (QoS), due to interference errors at higher rate [149]. Motion detectors, cordless phones, Bluetooth devices and radio frequency devices cause interference with WLAN [150] which leads to reduction in quality and security of WLAN services. File transfer protocol (FTP) server is another option for transmission of data via a Global System for Mobile–General Packet Radio Service (GSM-GPRS) modems. It can exist in form of PC card or external unit and can be connected through a USB cable, a serial

Table 6 Sensors used in the different studied systems. Author

Sensors

Features Solar Radiation (G) sensor

Temperature (T) sensor

Shunt resistor

Resistive potential divider

Eppley, Kipp, and Zonen

Thermistor

I: The voltage drop across the shunt resistor is used as a proportional measure of current flow. Shunt resistors have been used extensively to measure transient current pulses with fast rise-times and high amplitudes. V: Simple construction of sensor provides good accuracy and feasibility. G: Kipp and Zonen are independent, and give accurate reading of solar radiation with large response time and small temperature coefficient. T: Enclosed with Hermetic seal that can increase the ruggedness and eliminates the moisture induced sensor failure.

Mukaro [91]; Mukaro [92]

x

x

Eppley pyranometer

x

G: Eppley pyranometer has a copper-constantan thermopile which can withstand severe mechanical shocks and vibrations.

Duryea [93]

Hall effect

Resistive potential divider

x

Silicon temperature sensor (IC-LM335)

I: Fast response time, minimum core temperature rise, magnetic field stability. T: It is precise, easily-calibrated integrated circuit temperature sensor. Operating as a 2-terminal zener, the LM335 has a breakdown voltage directly proportional to absolute temperature at 10 mV/°K. With less than 1-Ω dynamic impedance, the device operates over a current range of 400–5 mA with virtually no change in performance. When calibrated at 25 °C, the LM335 has typically less than 1 °C error over a 100 °C temperature

Koutroulis [85]

Hall effect

Resistive potential divider

GS1 pyranometer

RTD (MP 100A)

G: The sensor has high quality, blackened thermopile with a glass covering dome acting as a filter which allows solar radiation. It has a flat spectral response in the range 0.3–3.0 micro meters. T: Very robust, therefore provides long-term stability. Cable length compensation up to 100 m.

Pietruszko [94]

Shunt resistor

Resistive potential divider

Silicon cell pyranometer

RTD (PT 100)

Krauter [95]

Hall effect

x

CM 3 pyranometer

x

Papadakis [96]

Hall effect

Resistive potential divider

pyranometer

RTD

G: Is compact in size and easy to install. They offer high performance and reliability. Silicon cell pyranometers are very cost-effective. T: It has accuracy about 8̄ 0.10 °C, high linearity over limited temperature range, Wide temperature range:  250–600 °C. G: Due to its flat spectral sensitivity from 300 to 3000 nm, it can be used in natural light, under plant canopies, and can measure reflected solar radiation. –

Forero [97]

Shunt resistor (in built in Resistive potential divider (in SP LITE Kipp & Zonen NI- FP-AI-100 module) built in NI- FP-AI-100 module) pyranometer

Thermistor

V, I: It consists of 8 voltage or currents inputs having 12-bit resolution. It can measure medium to very small voltage and current signals. G: SP Lite2 is easy to use. It can be directly connected to voltmeter or data logger. Direct readout in Watts per square meter (W/m2) can be derived from the measured voltage divided by the calibration coefficient.

Soler-Bientz [98]

x

x

Silicon photodiode pyranometer

K type Thermocouple

Mondol [99]

x

x

pyranometer

T type Thermocouple

So et al. [100]

Hall effect

Resistive potential divider

CMP 21 Kipp & Zonen pyranometer

RTD (PT 100)

G: Is small in size and has low cost. The value indicated by these pyranometer may differ from the “true” broadband solar irradiance by over 10%. T: It exhibits good corrosion resistance and provides wide range of temperature measurement upto 12600C. T: It is very stable and provides low temperature measurements from 0 to 350°C G: It has a low-dome thermal offset error, long term stability of sensitivity. It offers low temperature dependent performance and linearity as compared to CM 11 pyranometer.

Benghanem [82];

Maafi [83];

1197

Voltage (V) sensor

S.R. Madeti, S.N. Singh / Renewable and Sustainable Energy Reviews 67 (2017) 1180–1207

Current (I) sensor

1198

Table 6 (continued ) Author

Sensors Current (I) sensor

Features Temperature (T) sensor

Shunt resistor (in built in Resistive potential divider (in NI- FP-AI-100 module) built in NI- FP-AI-100 module) x x Shunt resistor (in built in Resistive potential divider (in NI- FP-AI-100 module) built in NI- FP-AI-100 module)

Silicon cell pyranometer

RTD (PT 100)

Silicon cell pyranometer Thermo electric pyranometer (NOVALYNX model 240-8101)

x X

Benghanem [104]

x

x

Kipp and Zonen pyranometer

Carullo [105]

Hall effect

Resistive potential divider

CMP 22 Kipp & Zonen pyranometer

Silicon temperature sensor (IC-LM335) RTD (PT 100)

Chouder [106]

Shunt resistor

Resistive potential divider

Ayompe [107] Ranhotigamage [108]

x Shunt resistor

x Resistive potential divider

CMP 11 Kipp & Zonen pyranometer Sunny sensor box Silicon Photodiode pyranometer (PDB-C139) Silicon cell pyranometer CMP 11 Kipp & Zonen pyranometer CMP 21 Kipp & Zonen pyranometer Silicon cell pyranometer CMP 11 Kipp & Zonen pyranometer x Silicon Photodiode pyranometer (Davis SRS-100) Silicon cell pyranometer Silicon cell pyranometer Silicon Photodiode pyranometer (LDR07) Kipp & Zonen Radiometer x

Rosiek [102] Boonmee [103]

Ammar Mahjoubi [109] Hall effect Alessio [110] Hall effect

Hall effect Resistive potential divider

Wittkopf [111]

x

x

Martín [112] Silvestre [113]

Hall effect Hall effect

Resistive potential divider Resistive potential divider

Peter [114] Tina [115]

Hall effect Hall effect

Resistive potential divider Resistive potential divider

Sánchez-Pacheco [116] Schill [117] Andò [118]

Hall effect x x

Resistive potential divider x x

Moreno [119] Han [120]

Hall effect Shunt resistor

Resistive potential divider Resistive potential divider

_

G: It consists of 12 wedge-shaped radially arranged thin copper sectors. 6 white and 6 black sectors are arranged alternately. Output from the thermopile is approximately 15 mV/Wm2. _

K type Thermocouple

G: CMP 22 has all the features of CMP 21 but uses very high quality quartz domes for a wider spectral range, improved directional response, and reduced thermal offsets. Because of the high optical quality of these domes the directional error is reduced below 5 W/m2. It is better than CMP 11. –

RTD (PT 100) Thermistor (B57164K472J)

– _

RTD RTD (PT 1000)

_ _

RTD (PT 100)

_

RTD (PT 100) K type thermocouple

_ _

x Silicon temperature sensor (IC-LM35) RTD (PT 100) RTD (PT 100) Silicon temperature sensor (TMP36) RTD (PT100) Silicon temperature sensor (IC-LM35)

_ _ _ – _ _ _

S.R. Madeti, S.N. Singh / Renewable and Sustainable Energy Reviews 67 (2017) 1180–1207

Solar Radiation (G) sensor

Gagliarducci [101]

Voltage (V) sensor

S.R. Madeti, S.N. Singh / Renewable and Sustainable Energy Reviews 67 (2017) 1180–1207

1199

Table 7 Summary on different types of controllers. Author

Controller

Features

Othman [124]

NI DAQ

1. NI-DAQ contains two drivers- Traditional NI-DAQ and NI-DAQmx, each having their own hardware-software configurations and application programming interface (API). 2. It can perform a variety of functions such as analog to-digital (A/ D) conversion, digital-to-analog (D/A) conversion, digital I/O, and counter/timer operations. 3. Increased performance, including faster single-point analog I/O and multithreading; and a simpler API for creating DAQ applications with fewer functions.

Anwari [125] ; μC (PIC 16F877A, ATmega Rosiek [102] 16) Lopez [112]

DSP

1. High performance Reduced Instruction Set Computer (RISC) CPU. 2. Operating speed: clock input (200 MHz), instruction cycle (200nS).3. Low power- high speed CMOS flash/EEPROM. 4. Only thirty seven instructions to remember5. Its code is extremely efficient, allowing the PIC to run with typically less program memory than any controller. 1. This highly integrated device has a 16-bit Modified RISC processor with modified Harvard architecture, and incorporates a 12-bit 200 ksps A/D converter.2. This MCU also includes a digital output serial peripheral interface (SPI), capable of communicating with the RF radio module.

cable, Infrared or Bluetooth [151,152]. High speed of data transmission is possible with GPRS devices, and thus huge volume of data is transferred to and from the mobile devices through Internet. The advantage of GPRS over GSM is that it provides great backup option. With the advent of new and faster data cards, the portability factor has become easy, and also has provision to transmit Short Message Service (SMS). Peersman [153] has presented an SMS system as a method of data transmission. Modular phone SMS can also be used to transmit the data with an average transmission speed of about 30 SMS per minute [154]. A GPRS modem is required to send and receive SMS via GPRS. In this way, the amount of solar radiation received at a particular area during a specific time period, and corresponding amount of voltage and current generated by PV module, module temperature can be recorded and sent for use via SMS. The ATtention (AT) command plays a significant role for working of GSM module, to send SMS signal and focus on auto SMS function. Before writing the AT command, there are different procedures that must be followed. In order to determine whether the mobile phone can support a type of SMS mode, a test should be performed on the mobile phone before any task can be performed. There are two basic types of

SMS modes: sending and receiving SMS messages, which are text mode and protocol description unit (PDU) mode. The message in text mode, an alphabetical format is used to write an SMS. Whereas, the message in PDU mode should be changed into Hex code before it can be sent. Fig. 33 shows an embedded system with an SMS data sender. The information related to metrological and operational parameters can also be transmitted by using Global System Mobile Communication (GSM) network [155]. GSM can cover distances of thousands of kilometers and can transmit the data from an Internet provider to a PC. This method offers worldwide coverage of data transmission with low cost. It reduces the need of manual data collection and routine inspection, which is necessary in case of traditional systems and also avoids the need to provide on-site computing. Power Line Communication (PLC) is only wireline technology with comparable price to wireless systems. These systems can operate at high frequency band (2–30 MHz) with maximum speed of data transmission around 200 Mbps [156]. For the purpose of signaling, fiber optics are deployed along with high voltage (HV) network, but are not being fully utilized presently. Only a

DC/AC Inverter Temperature sensor

Solar radiation sensor

Tamb GI,P

GH,P

Vdc, meas Idc, meas

NI PXI-6254

Vac, meas Iac, meas

Fig. 32. PV monitoring with Coaxial cable.

1200

S.R. Madeti, S.N. Singh / Renewable and Sustainable Energy Reviews 67 (2017) 1180–1207

# AT Init # # AT Sent # PV system

Microcontroller

AT command code

PC

Mobile device

Fig. 33. Embedded system with an SMS data sender.

fractional capacity of fiber optic network is required; and hence, it can be used to form an extended telecommunications network, incorporating PLC systems operating on the MV and LV networks. Liu [157] has explained the advantages of PLC lines extensively. 7.2. Methods of data storage Data storage is very important to carry out any performance analysis. The complete set of acquired data is stored in a Secure Digital (SD) card. Thus, original data can be prevented from any accident. Mechanical switch are provided with SDs card for write protection (WP) and card detection (CD) mechanisms. The main advantage of SD card is a type of non-volatile storage, which means data remains stable even if power is turned off. In addition, these SD cards are free from difficulties due to mechanical damage. SD cards do not produce any noise interference and provide an easy way of data storage and tracking. It has more space for data storage as compared to other storage devices. There are few disadvantages encountered in SD cards for data storage such as: fragility, misplacing and easily affected by virus. In order to capture the voltage, current in some applications, digital signal oscilloscope (DSO) is used which can measure the electric field signals [158]. It also provides extra features such as data storage and display. My Sequel (MySql) data base can also be used to store data, which is a database management system. The data sent from GSM module is sent to MySql database and displayed as a Hypertext preprocessor (PHP) page [159,160]. This method results in better data storage since the data is saved in HTML format. This HTML can be easily read by users using any browser. In Table 8, different types of transmission and their qualities are presented. 7.3. Methods of data analysis Data analysis is essential for drawing conclusions from research. Hence, different methods have been used for quantitative data analysis to determine the performance of various PV system configurations. The evaluation of energy performance of a PV plant is not easy because the operation of a PV plant is affected by many variables. The main problem is due to system response, which is

strongly dependent on the factors like solar radiation, cell temperature, ambient temperature, humidity, pollution, cloudiness and air velocity. Many models have been reported in literature to evaluate the effects of different uncertainties [163–168]. Nowadays, in order to assess the overall performance of a PV system, standard benchmarks [169] are used which are represented in terms of solar resource, energy production and system losses but they have following drawbacks: (a) It provides inaccurate information about the overall performance of PV system. (b) It does not allow any assessment of behaviour of individual parts in a PV system. To enhance the features of available software, a two-step procedure has been proposed in [170]. It has been implemented in Matrix Laboratory (MATLAB) with advanced statistics. These procedures are based on descriptive and inferential statistics. The first step consists of defining performance benchmarks and utilizing the population of energy data (offline supervision). The second step consists of verifying the well-operation of PV plant and utilizing the sampled energy data (real-time monitoring). These two procedures allow detecting and locating small misoperations, which is not possible by standard benchmarks. Another procedure is described in [171], in which the information related to population can be extracted even in absence of entire year data. This approach is based on a bootstrap technique as described in [172]. By using the above mentioned techniques, it is possible to evaluate results successfully by the expert users, since they possess required knowledge and skills to accomplish the tasks. LabVIEW interface provides an alternative by combining the procedures described in [170–172] to study the operation of PV system. LabVIEW is a graphical user-interface (GUI), which supports data acquisition, manual and automatic control of system parameters. In order to collect, calculate and analyse the operational and metrological data of a PV system, LabVIEW is used which is developed by virtual instrument (VI). It has ability to interact with MATLAB software. In addition, it also allows system designer to control and manage the PV system optimally. Meanwhile, noise present in the captured signals can be eliminated by using high pass filter in LabVIEW [173]. A VI was developed with the help of LabVIEW to monitor standalone solar PV plant in Bogota [97]. The data of incoming solar radiation was stored in MS-EXCEL and

Table 8 Different types of transmission used and their qualities. Author

Transmission type Significant differences in quality

Wang [127]

Coaxial cable

Luecke [161]

Fiber optic

Ranhotigamage [108] WLAN

Rosiek [102]

GSM–GPRS

Krauter [95]

Satellite

Han [162]

PLC

1. Low error, Good bandwidth and relatively low resistance. 2. Able to carry weak signals without any interference from the environment. 3. Without high loss in signal it can be able to transfer data with high speed. 1. It offers long distance terminals with high bandwidth to transmit data compared to copper wire. 2. Immunity to Electromagnetic Interference (EMI).3. Reasonable data transfer speed but less than coaxial cable. 4. Expensive to install and fragile 1. It can cover wide area to transmit data but it has lower band width, lower quality of service (QoS), due to interference errors at higher rate. 2. Speed of data transmission is moderate. 3. Disaster resistant. 4. It is susceptible for electromagnetic radio interference. 1. High speed of data transmission is possible with GPRS devices. 2. Due to the low cost and diffusion of the GSM devices, the transmission system is fairly cheap. 3. Electromagnetic radio interference- susceptible yes. 1. It has a great spatial and temporal coverage2. It is a very expensive method.3. Speed of data transmission is less.4. It is susceptible for electromagnetic radio interference. 1. Good bandwidth 2. It is cost effective compared to other data transmission systems. 3. Moderate speed of data transmission. 4. Moderately susceptibility to EMI.

S.R. Madeti, S.N. Singh / Renewable and Sustainable Energy Reviews 67 (2017) 1180–1207

analyzed by using various techniques and statistical function. The efficiency of PV system and of inverter was then calculated by using LabVIEW. The Harmonic analysis of AC signal generated by the inverter was done by using Fourier analysis with the help of LabVIEW package tool. A new method for data analysis which is used to test PV systems periodically is by using numerical simulation. It can be done by using Personal Computer Simulation Program with Integrated Circuit Emphasis (PSpice) [174] and MATLAB [175]. It has an advanced mathematical manipulation toolbox, which is very helpful in PV system simulations. MATLAB is very flexible as compared to PSpice, and is able to simulate complex dynamic systems; it has been successfully used in the modeling and simulation of many PV systems [176]. The operational and metrological data collected from PV system can be analyzed by using artificial intelligence (AI) techniques, such as Artificial neural network (ANN) analysis, fuzzy-logic, genetic algorithms (GA) and fuzzy-neural analysis [177]. ANN is a learning technique popular for time series forecasting and can be used to recognize the optimal behaviour of a PV system. Although NN parameter tuning may produce different prediction results, it is easy to design an NN based environmental and operational parameter predictor with reasonable precision to detect undesirable situations which may plunge energy production levels from available solar irradiation [178]. By using this, fault localization can be achieved when behaviour of the system diverts from the expected one [103,179–184]. It is typically arranged in layers called input layer which receives data for processing. Its output layer shows its response to training data. A hidden layer is present between the input and output layers where the actual processing is done as shown in Fig. 34. Fuzzy set (FS) theory was introduced by Zadeh in 1965, which is generalization of conventional set theory [185]. It contains a mathematical tool, which can deal with linguistic variables associated with natural languages. Flow chart of fuzzy inference system is shown in Fig. 35. It can be used for data analysis by having multiple-input multiple-output (MIMO) system. This technique reduces the outliers influence on the fitting function, and thus ensures accurate estimation of parameter prediction [186]. GA is another option for data analysis. It considers many points in search space simultaneously for fault classification, diagnosis and control. GA is the most popular technology based on Darwinian thinking of natural genetics and natural selection [188]. This technique avoids the local traps and searching of entire parameter sphere to achieve optimal model parameters, thus offsetting the impacts of measurement noise [189]. Fuzzy-neural architecture was derived by Ishibuchi et al. [190] based on two reliable

Fig. 34. Feed-forward neural network [177].

1201

techniques, such as fuzzy logic and neural networks. Mamdani and Sugeno [191] are the two types of neuro-funny modes in which for input variables x and y for output can be represented in polynomial form. Several methods can be used to perform fuzzy-neural analysis including root mean square error, mean percentage error and mean square error [192,193]. This technique provides a powerful modeling for complex and dynamic systems, which are often used to resolve the faults encountered in PV systems. The established fuzzy models can readily estimate the location of fault in real time applications [194].

8. Opportunities for PV monitoring systems Various organizations such as International Atomic Energy Agency (IAEA), World Energy Council (WEC), International Energy Agency (IEA), US Energy Information Administration (EIA), have given different energy demand predictions/projections by years 2020, 2030 and 2050 in [195]. According to their estimates, percentage of world’s electricity generated from coal will be reduced from 40% in 2008 to 37% and 33% in 2030 and 2050, respectively. With this decreasing trend in use of fossil fuels for energy generation, meeting the increasing future energy demand creates several local, regional and environmental challenges. The sustainable way to meet the gap between the future demand and generation is by using RES for generation. The European Union (EU) member states and many other developing countries have placed more priority on non-conventional energy sources [196,197]. By 2020, EU envisages 20% annual energy consumption to come from RES. Moreover, EU has put forward the plans to raise this number to 27% and 34% by years 2030 and 2050 respectively. Among all RES, solar energy has high potential and most abundant source in the world which can be used to meet the future energy demands. It is projected that the solar power production across the globe is to be increased to 402 TW h by 2030. At this verge, it is very much essential to develop technologies, which can keep track a PV energy production from a given PV plant and keep up its production in every possible dimension. Consequently, it is also necessary to develop techniques, which can help estimating the true potential (power) of RES at the installed area in real-time. Moreover, continuous monitoring of PV system health is very crucial to detect the causes, which could hamper the desired performance.

9. Challenges and prospects of PV system monitoring: Current and Future There has been a significant progress in PV monitoring techniques over past few years. Availability of large and open data sets and development of new suitable algorithms, which ensure reliable and stable operation of PV system have been the reasons behind progress. However, their practical application for PV health monitoring is hindered by the challenges associated with different components integrated within it. These new challenges in PV monitoring system practicality have been addressed in this paper. A brief summary of challenges in the field of PV health monitoring reported in the literature are discussed below: 1) Harsh environmental conditions: The conventional wired PV monitoring systems to collect measurements suffers from several limitations. It could be reduction in lifespan due to continuous exposure to sunlight or reduced sensors reliability due to highly corrosive and caustic environments, RF interference, dust, high humidity levels, vibrations and/or other conditions that challenge performance [198,199]. Such harsh environmental conditions might result the portion of cable failures and sensors to

1202

S.R. Madeti, S.N. Singh / Renewable and Sustainable Energy Reviews 67 (2017) 1180–1207

Knowledge base

Database

Input

Rule base

Fuzzification interference

Defuzzification interference

(Fuzzy)

Output

(Fuzzy)

Decision making unit Fig. 35. Flow diagram of fuzzy inference system [187].

malfunction or render the information they gather obsolete [200]. 2) Reliability and latency requirements: Quality of various sensors used in PV monitoring systems is well described by specifications in terms of reliability, latency, network throughput, etc. Moreover, since the data collected by the sensors tends to be time sensitive, it should be delivered to the controller node in a timely manner [201]. 3) Efficiency degradation: Most of the PV modules are manufactured by the silicon material and comes with a warranty for 25 years. However, it is very hard to find a complete PV conversion system (including all electrical interfaces) with an equivalent lifetime in actual applications. Also, no generalized test protocol is in place to validate the complete PV monitoring system lifetime. Furthermore, it is not easy to predict the degradation of various parts in complete PV conversion systems under different environmental stresses. A research study in [202] has reported that the average annual degradation of PV modules is to be around 0.8% per year, as depicted in Fig. 36. The impact of this degradation on electrical characteristics of PV modules was explained by Chamberlin et al. [203]. The main factors which affect this degradation are levels of metal oxidation, corrosion of connectors, discoloration of busbars and increase in series resistance of a PV module. All these are progressive effects with no means to quantify, certainly have impact on estimation of the origin of power losses; and hence, on PV module degradation. This is a real challenge to the scientific community. 4) Resource constraints: The three main constraints to design and implement a PV monitoring system are a) energy efficiency; b) data storage; and c) data processing. As the sensors and data storage systems have limited battery energy supply [200], the communication protocols used in monitoring system should be tailored to provide high energy efficiency. 5) System calibration and economical challenges: Various sensing /measuring methods widely used in various fields such as environmental pollution monitoring, control of industrial processes and energy generation plants share the series of common

characteristics with PV monitoring systems which are. i) Heterogeneous nature, since different quantities have to be acquired;. ii) Comprise of several commercial products (sensors, controller, data acquisition boards, and data transfer mechanism etc.) integrated within it, iii) These systems require dedicated algorithm for data analysis that affects /controls the data acquisition, transmission and storage. Above mentioned five characteristics make these systems very flexible and suitable for different scenarios, but they may pose problems to quality assurance and traceability. Measuring devices have to be moved to laboratory for calibration purposes, but it is not so easy since these devices are deeply integrated with the monitoring systems [204]. Moreover, the results after calibration might not always represent the behaviour of measuring chains in operating conditions, and the effects of software components are not taken into account. These problems can be tackled by employing a conventional calibration procedure. However, this solution dramatically increases the overall calibration and operation cost since it requires transport of reference standards and expert technicians at the operating site. On the whole, selection of appropriate measuring devices whose calibration is remotely managed by the system under test connected through a good quality network is required for PV monitoring systems, to ensure reliable and energy efficient operation. 6) Other futuristic challenges and scope for improvements: In order to perceive the challenges of futuristic PV monitoring systems, it is important to relook at their expected properties, which include Real-time reporting, trend prediction, unmanned operation ability, accurate measurements, real-time data logging, proper storage facilities, triggers /alerts, off-site control and secure access. These properties can be realized by fulfilling the futuristic objectives /tasks as follows:

      

Ensuring accuracy of measurements, Efficient data logging, storage and transmission, System automation and real-time operation, Immediate fault detection and removal, Real time visualization of operational parameters, Prior estimation of system behaviour, System control for output maximization.

The above mentioned objectives/tasks pose several challenges; among which, some could be overcome to an extent through the following approaches:

 Use of measurement devices, sensors and transducers with  Fig. 36. Histogram of reported degradation rates.

proven efficiency and development of new ones with improved performance, Use of signal processing techniques such as wavelet transform for efficient data compression and reduced cost of data storage,

S.R. Madeti, S.N. Singh / Renewable and Sustainable Energy Reviews 67 (2017) 1180–1207

 Focusing towards new technologies such as OPAL-RT instead of   

microcontrollers for control and automation, Ensuring centralized or off-site control using high speed communication networks like satellites, wireless communication systems, Use of artificial intelligence techniques for future state prediction. Incorporating latest hardware /software technologies in monitoring systems for real-time and secure operation.

10. Discussions In order to sense the operational and metrological parameters of a PV system, different sensors are used as reported in the literature. But it is very difficult to identify, due to system size variation, working principles, structure of sensors, hardware used, and differences in installation sites. Therefore, based on these challenges, a summary is provided based on the working principles of different sensors, which are used to measure operational and metrological parameters such as current, voltage, solar radiation and temperature. For current measurement, the shunt resistance method is widely used because of its simplicity, low cost and reasonable accuracy. But for measurement of high current values, it becomes troublesome due to increase in size and high power loss. On the other hand, Hall-Effect sensor provides an alternative with good accuracy, low losses, high bandwidth and galvanic isolation, but its cost is very high. Closed-loop AMR-based current sensors provide almost similar performance to Hall-effect sensors. But they are only available for current ranges from one ampere up to over hundred amperes, and are costlier than the Hall-effect sensors. Flux gate current sensors provide high performance accuracy but are suitable only for particular applications, because they are very expensive. For high magnitude, direct currents fiber-optic current sensors are preferred. Alternating currents can be measured widely by using Rogowski coils and CTs. These techniques can measure from few amperes to mega amperes with low cost and high accuracy. Magnetic field sensors such as AMR and GMR allow indirect current sensing at reasonable price. However, the accuracy may be low due to skin-effect inside the primary conductor which can interfere with the magnetic field. These sensors are sensitive to external magnetic fields, which are difficult to shield. Potential dividers are generally used for measurement of low voltage applications due to its simple construction and reasonable accuracy. But due to high power loss, it is used in limited applications. Usually, PT and CCVT have been used to sense the medium and high level voltage with good accuracy. These instruments require oil to provide high insulation and cooling which makes them expensive. EOVTs are used in limited applications due to its high cost and limited life-span. A direct beam solar radiation can be measured by Pyrheliometer by converting solar heat to electrical signal. The horizontal beam and diffused solar radiations are measured by using Pyranometer, which works on thermo-mechanical, photovoltaic and thermoelectric principles. For temperature measurement, thermocouples based on Seebeck-effect can be used to measure wide range of temperature, but its measurement has less accuracy and varies non-linearly. Alternatively, RTD provides an accurate temperature measurement over wide range of resistance, which varies linearly with temperature. It is limited to fewer applications because increase in its size leads to increase in cost. A resistor type thermometer called thermometer can be used to measure the low range temperature with good accuracy compared to RTD with less cost. An effective data transmission system is very essential in order

1203

to transmit data. To facilitate communication between the sensors and DAQ system, various mediums are used such as wired, wireless and power line communications systems. In terms of coverage area and length, coaxial cables are not capable of operating over long distance as compared to fiber-optic cable. Usually, these coaxial cables are preferred by the researchers when distance between the sensors and data acquisition system is short. WLAN can cover only a small area approximately 20 km2 compared to GPRS-GSM, and can transmit data over distances of hundreds or thousands of kilometers through internet whereas PLC carries the information about hundreds or thousands of meters by using existing wired infrastructure without any additional installations which reduces the cost of the system. The maximum speed at which the data can be transmitted by using Coaxial cable is around 10 Gbps. Fiber-optical cable is only capable of transmitting data from 100 Mbps to 2000 Mbps. WLAN can reach maximum speed of data transmission compared to coaxial cable or fiber-optical but it is limited due to some factors, such as physical obstructions between the devices, radio interference, simultaneous communication of multiple devices in a network, and distance between the devices. PLC can reach the maximum data transmit capability of 200 Mbps. The average speed of GPRS-GSM is about 40– 50 Kbps, which is very slow speed as compared to other data transmission techniques. The installation cost and process of wired data transmission system is very high. As a result, wireless and PLC are the best candidates for communication since these do not need any additional cables between the sensor and data acquisition system for data transmission. Due to severe radio interference effect, wireless technology has limited applications. As a result, PLC technology can be considered as better choice for data transmission. Usually, data storage can be done on a device or online. The DSO and SD card are the devices used for data storage. MySQL is an online form of storage in which data are stored in a database. The capacity of SD card memory ranges from 4 GB to 32 GB. Unfortunately, the SD card has few disadvantages including it can be easily broken or lost due to its small size and also be exposed to virus, which corrupts all the useful information. Due to storage limitations, the permanent storage cannot be possible by DSO. It can erase previous data if the storage limit exceeds. MySQL is the best way to keep the data safe against any accidental change or loss. It is very difficult to find data from storage devices than the data, which is available online. The development of world-wide network has made it easier to acquire information online. Generally, data analysis is used to find out useful information in order to implement the successful computer-aided decisionmaking support system in PV monitoring systems. Few of these methods are complex, while the others are simple. LabVIEW constitutes a graphical programming language, which can analyse the data in short interval of time as compared to text-based programming environments. These LabVIEW programs are called VIs, which contains a Front panel (FP) and a Block diagram (BD). MATLAB and Microsoft Visual Basic can also be used to develop a GUI. A GUI can be created interactively or programmatically. C/C þ þ is the most common programming language used to develop software applications [205]. Neural network is complicated because it includes a multi-layer network that must be trained properly to obtain the desired results. On the other hand, fuzzy-logic is much simpler than neural network, but its performance is inferior. In order to improve the performance of fuzzy-logic, it can be combined with neural networks. The combination of both methods reduces the complexities involved. 11. Recommendations Most of the PV inverters have wide range of interface including

1204

S.R. Madeti, S.N. Singh / Renewable and Sustainable Energy Reviews 67 (2017) 1180–1207

sensor inputs and communication interfaces. These sensors inputs can be used to connect PV module temperature sensors and irradiance sensors as pyranometer. These inverters include monitoring capabilities for temperature as well as irradiance for MPPT evolution. Nevertheless, inverter integrated measurements are not sufficiently reliable and accurate. For an advanced PV monitoring system, it is suitable to measure the current or power at string level. The additional cost for advanced monitoring system depends on the capacity of PV plant. When more energy is produced from the installed PV plant, then economical benefit is higher. If PV plant produces less energy than expected, string level monitoring significantly reduces the cost and time taken to detect the failure. For these reasons, string level monitoring is strongly recommended. Some general recommendations for monitoring systems are as follows:

 Measurement Interval: The data should be sampled at a high

 



 



rate. Averaged values should be stored within 5–15 min. Larger sampling times cannot capture many system issues. One second data provides high resolution and fidelity capture of transient events such as grid failure and inverter shutdowns. Degradation rate can be easily calculated by large number of data points. Owing to the development of technology, cost of hardware implementation and data storage decreases. Outdoor cable connections: In outdoor applications, use protective boots for cable connectors to withstand harsh climatic working conditions. Limit corrosion effect by using all means. Equipment temperature rating: If the system is exposed to direct sun light, enclosure equipment temperature can dramatically increase. So equipment should be capable to withstand these extremes of about  55 °C to þ90 °C. Inspection interval: Quarterly visual check improves the reliability of monitoring system by avoiding following problems: i) Loose connections at sensors/ data acquisition systems, ii) Damage to the sensors outside (ambient temp sensors, pyranometer), iii) Vermin-proof enclosure or Moisture protection. Measuring instrument errors: Make sure that measuring instrument does not affect significantly the true value. Calibration interval: Bench calibrations are performed in indoor environment conditions which is an ideal situation. Reference standard and calibration equipment are affected by the change in outdoor environment, which may result in offset errors. To handle this, end to end calibration is recommended for which the limits may be ambient temperature above 15 °C and wind speed below 3 m/s. Cost: Rapid increase in different commercial products causes increase in overall cost of the monitoring systems. So, careful selection of the components (even the tiny resistor) is very important, it may be cost effective to replace it when it fails.

12. Conclusion In this paper, a comprehensive review of existing PV monitoring systems reported in the literature has been presented in terms of sensors being used as well as data acquisition systems. The sensors section deals with sensors used for important operational and metrological parameters (e.g., voltage, current, solar radiation, temperature) along with their working principles in different PV monitoring systems. It is very difficult to discuss each and every sensor used, because it becomes really complex, and every sensor has its own benchmarking. The section of data acquisition systems covers the controllers being used for data acquisition system, types of data transmission methods, data storage and data analyses. All

the necessary comparisons were discussed and tabulated wherever required. It is believed that the acquaintance of information provided in this review is crucial for the development of an effective, low cost PV monitoring system viable for even small and medium scale PV plants, without compromising on the desired performance.

References [1] Canale M, Fagiano I, Milanese M. KiteGen. A revolution in wind energy generation. Energy 2009;34:355–61. [2] Kalogirou SA. The potential of solar industrial process heat applications. Appl Energy 2003;76:337–61. [3] Tien W, Kuo K-C. An analysis of power generation from municipal solid waste (MSW) incineration plants in Taiwan. Energy 2010;35:4824–30. [4] Mercure J-F, Salas P. An assessment of global energy resource economic potential. Energy 2012;46:322–36. [5] Singh GK. Solar power generation by PV (photovoltaic) technology: a review. Energy 2013;53:1–13. [6] IEA PVPS, International Energy Agency – Photovoltaic Power Systems Programme, 2014. Trends in photovoltaic applications. report available at 〈http://www.iea-pvps.org/fileadmin/dam/public/report /national/IEAPVPS_-_Trends_2014_-_MedRes.pdf〉. [7] Benghanem M. Low cost management for photovoltaic system in isolated site with new IV characterization model proposed. Energy Convers Manag 2009;50:748–55. [8] Hanson AJ, Deline CA, MacAlpine SM, Stauth JT, Sullivan CR. Partial-shading assessment of photovoltaic installations via module-level monitoring. IEEE J Photovolt 2014;4(6):1618–24. [9] Sánchez-PachecoFJSotorrio-RuizPJHeredia-LarrubiaJRPérez-HidalgoFSidrachde-CardonaM. Low cost DC lines PLC based photovoltaic plants parameters smart monitoring communications and control module. In: Power Engineering, Energy and Electrical Drives (POWERENG) International Conference on 2011; pp.1–6; 2011. [10] DavarifarMRabhiAEl-HajjajiADahmaneM. Real-time model base fault diagnosis of PV panels using statistical signal processing. In: Renewable Energy Research and Applications (ICRERA) International Conference on 2013; pp. 599–604; 2013. [11] Sumathi S, Kumar LA, Surekha P. Solar PV and wind energy conversion systems: an introduction to theory, modeling with MATLAB/SIMULINK, and the role of soft computing techniques. Switzerland: Springer International Publishing; 2015. [12] HansenADSørensenPEHansenLHBindnerHW. Models for a stand-alone PV system; 2001. [13] Bechtel National I. Handbook for battery energy storage in photovoltaic power systems. J Power Sources 1980;5(4):340–1. [14] Borowy BS, Salameh ZM. Methodology for optimally sizing the combination of a battery bank and PV array in a wind/PV hybrid system. IEEE Trans Energy Convers 1996;11(2):367–75. [15] ISC Committee. IEEE standard for interconnecting distributed resources with electric power systems. New York, NY: Institute of Electrical and Electronics Engineers; 2003. [16] International Electrotechnical Commission. IEC 61724, photovoltaic system performance monitoring e guidelines for measurement, data exchange and analysis. 1st ed. Geneva, IEC: International Electrotechnical Commission; 1998. [17] Ziegler S, Woodward RC, Iu HH, Borle LJ. Current sensing techniques: a review. IEEE Sens J 2009;9(4):354–76. [18] Spaziani L. Using copper PCB etch for low value resistance. Unitrode Design Note DN-71; 1997. [19] RayWFHewsonCR. High performance Rogowski current transducers. In: Conference record of the IEEE Industry Applications Conference;Vol.5, pp.3083–3090; 2000. [20] RayWFHewsonCR. RCTrms Rogowski coil datasheet Power Electronic Measurements Ltd. [Online]. Available from: 〈http://www.pemuk.com〉; 2014, (accessed 12-12-14); 2005. [21] Ward DA, Exon JL. Using Rogowski coils for transient current measurements. Eng Sci Educ J 1993;2(3):105–13. [22] Radun A. An alternative low-cost current-sensing scheme for high-current power electronics circuits. IEEE Trans Ind Electron 1995;42(1):78–84. [23] DalessandroLKarrerNCiappaMCastellazziAFichtnerW. Online and offline isolated current monitoring of parallel switched high-voltage multi-chip IGBT modules. In: IEEE Power Electronics Specialists Conference;pp.2600– 2606; 2008. [24] Costa F, Poulichet P, Mazaleyrat F, Labouré E. The current sensors in power electronics, a review. EPE J 2001;11(1):7–18. [25] TiplerPA. An introduction to the Hall effect, Bell Technologies, Inc. [Online]. Available from: 〈http://www.fwbell.com〉; 2014; (accessed 18-01-15); 2005. [26] Tipler PA. Physics for scientists and engineers. New York: Worth Publishers, Inc; 1991. [27] Popovic RE, Randjelovic Z, Manic D. Integrated Hall-effect magnetic sensors. Sens Actuators A: Phys 2001;91(1):46–50.

S.R. Madeti, S.N. Singh / Renewable and Sustainable Energy Reviews 67 (2017) 1180–1207

[28] Blanchard H, De Montmollin F, Hubin J, Popovic RS. Highly sensitive Hall sensor in CMOS technology. Sens Actuators A: Phys 2000;82(1):144–8. [29] Ripka P. Review of fluxgate sensors. Sens Actuators A: Phys 1992;33(3):129– 41. [30] Lenz J, Edelstein S. Magnetic sensors and their applications. IEEE Sens J 2006;6(3):631–49. [31] Ripka P. Advances in fluxgate sensors. Sens Actuators A: Phys 2003;106(1):8– 14. [32] Primdahl F. The fluxgate magnetometer. J Phys E: Sci Instrum 1979;12 (4):241. [33] DezuariOBelloyEGilbertSEGijsMAM. Basic introduction to the use of magnetoresistive sensors, Appl. Rep. AN 37, Zetex Semiconductors, USA; 2003. [34] LaimerGKolarJW. Design and experimental analysis of a DC to 1 MHz closed loop magneto-resistive current sensor. In: Twentieth Annual IEEE Applied Power Electronics Conference and Exposition. APEC 2005. 2005;Vol.2: pp.1288–92; 2005. [35] Binasch G, Grünberg P, Saurenbach F, Zinn W. Enhanced magnetoresistance in layered magnetic structures with antiferromagnetic interlayer exchange. Phys Rev B 1989;39(7):4828. [36] Baibich MN, Broto JM, Fert A, Van Dau FN, Petroff F, Etienne P, Creuzet G, Friederich A, Chazelas J. Giant magnetoresistance of (001) Fe/(001) Cr magnetic superlattices. Phys Rev Lett 1988;61(21):2472. [37] Mosca DH, Petroff F, Fert A, Schroeder PA, Pratt WP, Laloee R. Oscillatory interlayer coupling and giant magnetoresistance in Co/Cu multilayers. J Magn Magn Mater 1991;94(1):L1–5. [38] E. Udd (Ed.), Fiber optic sensors: An Introduction for Engineers and Scientists, Wiley, New York; 1991. [39] Lee B. Review of the present status of optical fiber sensors. Opt Fiber Technol 2003;9(2):57–79. [40] Papp A, Harms H. Magnetooptical current transformer. 1: Principles. Appl Opt 1980;19(22):3729–34. [41] Cease TW, Johnston P. A magneto-optic current transducer. IEEE Trans Power Deliv 1990;5(2):548–55. [42] Bohnert K, Gabus P, Nehring J, Brändle H, Brunzel MG. Fiber-optic current sensor for electrowinning of metals. J Light Technol 2007;25(11):3602–9. [43] Blake J, Tantaswadi P, De Carvalho RT. In-line Sagnac interferometer current sensor. IEEE Trans Power Deliv 1996;11(1):116–21. [44] Nicati PA, Robert P. Stabilised current sensor using Sagnac interferometer. J Phys E: Sci Instrum 1988;21(8):791. [45] Bohnert K, Gabus P, Nehring J, Brandle H. Temperature and vibration insensitive fiber-optic current sensor. J Light Technol 2002;20(2):267–76. [46] Bohnert K, Brandle H, Brunzel MG, Gabus P, Guggenbach P. Highly accurate fiber-optic DC current sensor for the electrowinning industry. IEEE Trans Ind Appl 2007;43(1):180–7. [47] Frosio G, Dändliker R. Reciprocal reflection interferometer for a fiber-optic Faraday current sensor. Appl Opt 1994;33(25):6111–22. [48] AEO2013 Early Release Overview. [Online]. Available from: 〈http://www.eia. gov/forecasts/aeo/er/pdf/0383er(2013).pdf〉; 2015 (accessed 01-02-2015). [49] Moghe R, Lambert FC, Divan D. Smart “stick-on” sensors for the smart grid. IEEE Trans Smart Grid 2012;3(1):241–52. [50] MogheRYangYLambertFDivanD. Design of a low cost self powered “Stick-on” current and temperature wireless sensor for utility assets. In: 2010 IEEE Energy Conversion Congress and Exposition; pp. 4453–60; 2010. [51] ABB, Instrument Transformers and Sensors [Online]. Available from: 〈http:// www.abb.com/product/us/9AAC720011.aspx?country ¼US〉. 2015 (accessed 26-03-15). [52] Parker DM, McCollough ND. Medium-voltage sensors for the smart grid: Lessons learned. In: IEEE Power and Energy Society General Meeting 2011; pp.1-7. [53] Vaniz WR, Sieron RL. Apparatus for measuring the voltage of a transmission line conductor, U.S. patent 4,714,893, Dec. 22, 1987. [54] SorensenT. Voltage measuring device, U.S. patent 7,397,233, Jul. 8; 2008. [55] XuQRSendRPaprotnyIWhiteRMWrightPK. Miniature self-powered stick-on wireless sensor node for monitoring of overhead power lines. In: IEEE Energy Conversion Congress and Exposition; pp. 2672–2675; 2013. [56] HussinMFHaddadAHaridN. Voltage transducer for monitoring of high voltage overhead lines. In: Universities Power Engineering Conference (UPEC) 45th International 2010; pp. 1–5; 2010. [57] YangYDivanDHarleyRGHabetlerTG. Design and implementation of power line sensornet for overhead transmission lines. In: 2009 IEEE Power & Energy Society General Meeting; pp. 1–8; 2009. [58] Moghe R, Iyer AR, Lambert FC, Divan DM. A low-cost wireless voltage sensor for monitoring MV/HV utility assets. IEEE Trans Smart Grid 2014;5(4):2002– 9. [59] LiFMoorePJ. Determination of the phase sequence and voltage level of overhead conductors using non-contact sensors. In: Proceedings. IEEE PES Transmiss. Distribution Conference. Expo., Asia Pacific, pp.1–6; 2005. [60] Moghe R, Iyer AR, Lambert FC, Divan D. A low-cost electric field energy harvester for an MV/HV asset-monitoring smart sensor. IEEE Trans Ind Appl 2015;51(2):1828–36. [61] Iqbal Muhammad. An introduction to solar radiation. New York: Academic Press; 1983. [62] Meftah M, Dewitte S, Irbah A, Chevalier A, Conscience C, Crommelynck D, Janssen E, Mekaoui S. SOVAP/Picard, a space borne radiometer to measure the total solar irradiance. Sol Phys 2014;289(5):1885–99. [63] Ballestrín J, Estrada CA, Rodríguez-Alonso M, Pérez-Rábago C, Langley LW,

[64] [65] [66]

[67]

[68]

[69] [70] [71] [72] [73]

[74]

[75] [76] [77] [78]

[79] [80] [81]

[82] [83]

[84]

[85]

[86]

[87]

[88]

[89]

[90]

[91]

[92]

[93]

[94] [95] [96]

1205

Barnes A. Heat flux sensors: calorimeters or radiometers? Sol Energy 2006;80(10):1314–20. GargHP. Solar energy: fundamentals and applications. Tata McGraw-Hill Education; 2000. Coulson Kinsell. Solar and terrestrial radiation: methods and measurements. New York: Academic Press; 2012. Paulescu M, Paulescu E, Gravila P, Badescu V. Solar radiation measurements. Weather Modeling and Forecasting of PV Systems Operation London, U.K: Springer; 2013 (Chapter 2). Hukseflux (2012) Radiation measurement sensors. Available from: 〈http:// www.huksefluxusa.com/radiation-measurement.php〉; 2015 (accessed 1304-15). DeltaOHM. Pyranometers, albedometers, net irradiance meter. Manual. Available from: 〈http://www.Deltaohm.Com/Ver2008/Uk/Pyra02.htm〉; 2015 (accessed 18-04-15); 2012. Garg HP, Garg SN. Measurement of solar radiation. I. Radiation instruments. Renew Energy 1993;4(5):321–33. Moll WJH. A thermopile for measuring radiation. Proc Phys Soc Lond 1922;35:257–61. Paulescu M, Schlett Z. Performance assessment of global solar irradiation models under Romanian climate. Renew Energy 2004;29(5):767–77. Alados-Arboledas L, Batlles FJ, Olmo FJ. Solar radiation resource assessment by means of silicon cells. Sol Energy 1995;54(3):183–91. KingDLMyersDR. Silicon-photodiode pyranometers: Operational characteristics, historical experiences, and new calibration procedures. In: 26th IEEE Photovoltaic Specialists Conference, Anaheim, CA; pp.1285–1288; 1997. Alonso Garcia MC, Balenzategui JL. Estimation of photovoltaic module yearly temperature and performance based on nominal operation cell temperature calculations. Renew Energy 2004;29(12):1997–2010. Andreev VM, Grilikhes VA. Photovoltaic conversion of concentrated sunlight. London: Wiley; 1967. Bajzek Thomas J. Thermocouples: a sensor for measuring temperature. IEEE Instrum Meas Mag 2005;8(1):35–40. Potter David. Measuring temperature with thermocouples–a Tutorial. Natl Instrum 1996:340904B–01. Wakeham WA, Assael MJ. Measurement, instrumentation and sensors handbook: Spatial, Mechanical, Thermal and Radiation Measurement. Boca Raton, FL: CRC Press; 1999. Zhao Ran, et al. Temperature sensor made of polymer-derived ceramics for high-temperature applications. Sens Actuators A: Phys 2014;219:58–64. Temperature Sensor Handbook, "National Semiconductor’s. Japan; 1989. Wilshaw Alison R, Pearsall Nicola M, Hill Robert. Installation and operation of the first city centre PV monitoring station in the United Kingdom. Sol Energy 1997;59(1):19–26. Benghanem M, Hadj Arab A, Mukadam K. Data acquisition system for photovoltaic water pumps. Renew Energy 1999;17(3):385–96. Benghanem Mohamed, Maafi Abdelbaki. Data acquisition system for photovoltaic systems performance monitoring. IEEE Trans Instrum Meas 1998;47(1):30–3. Wichert B, Dymond M, Lawrance W, Friese T. Development of a test facility for photovoltaic–diesel hybrid energy systems. Renew Energy 2001;22 (1):311–9. Koutroulis Eftichios, Kalaitzakis Kostas. Development of an integrated dataacquisition system for renewable energy sources systems monitoring. Renew Energy 2003;28(1):139–52. XiaoChucheng. et al. An overview of integratable current sensor technologies. In: 38th IAS Annual Meeting Industry Applications IEEE Conference Record; 1251-1258; 2003. Wijeweera G, Shafai C, Rajapakse A. Measuring power system voltage remotely using micromachined electric field sensor. In Proc: IEEE Microsystems and Nanoelectronics Research Conference, MNRC, 2008; 209-12. Stoffel TL, Reda I, Myers DR, Renne D, Wilcox S, Treadwell J. Current issues in terrestrial solar radiation instrumentation for energy, climate, and space applications. Metrologia 2000;37(5):399. Barnabei AL, Grassi M, Dallago E, Malcovati P, Finarelli DG, Liberale A, Quaglia F. In Proc: IEEE A wireless irradiance-temperature-humidity sensor for photovoltaic plant monitoring applications. 2012; 1-4. KingDavid L. Photovoltaic module and array performance characterization methods for all system operating conditions. In: AIP Conference Proceedings; pp. 347–368; 1997. Mukaro R, Carelse XF, Olumekor L. First performance analysis of a silicon-cell microcontroller-based solar radiation monitoring system. Sol Energy 1998;63(5):313–21. Mukaro Raphael, Carelse Xavier Francis. A microcontroller-based data acquisition system for solar radiation and environmental monitoring. IEEE Trans Instrum Meas 1999;48(6):1232–8. DuryeaSIslamSLawranceW. A battery management system for standalone photovoltaic energy systems. In: Proceedings IEEE Industry Applications Conference; 4:2649-2654; 1999. Pietruszko SM, Gradzki M. Performance of a grid connected small PV system in Poland. Appl Energy 2003;74(1):177–84. Krauter Stefan CW, Depping Thomas. Remote PV-system monitored via satellite. Sol Energy Mater Sol Cells 2004;82(1):139–50. Papadakis Kostas, Koutroulis Eftichios, Kalaitzakis Kostas. A server database system for remote monitoring and operational evaluation of renewable energy sources plants. Renew Energy 2005;30(11):1649–69.

1206

S.R. Madeti, S.N. Singh / Renewable and Sustainable Energy Reviews 67 (2017) 1180–1207

[97] Forero N, Hernández J, Gordillo G. Development of a monitoring system for a PV solar plant. Energy Convers Manag 2006;47(15):2329–36. [98] Soler-Bientz R, Ricalde-Cab LO, Solis-Rodriguez LE. Developing a mobile stand alone photovoltaic generator. Energy Convers Manag 2006;47 (18):2948–60. [99] Mondol Jayanta Deb, et al. Long term performance analysis of a grid connected photovoltaic system in Northern Ireland. Energy Convers Manag 2006;47(18):2925–47. [100] So Jung Hun, et al. Performance results and analysis of 3kW grid-connected PV systems. Renew Energy 2007;32(11):1858–72. [101] Gagliarducci M, Lampasi DA, Podesta L. GSM-based monitoring and control of photovoltaic power generation. Measurement 2007;40(3):314–21. [102] Rosiek S, Batlles FJ. A microcontroller-based data-acquisition system for meteorological station monitoring. Energy Convers Manag 2008;49 (12):3746–54. [103] Boonmee Chaiyant, Plangklang Boonyang, Watjanatepin Napat. System performance of a three-phase PV-grid-connected system installed in Thailand: data monitored analysis. Renew Energy 2009;34(2):384–9. [104] Benghanem M. Measurement of meteorological data based on wireless data acquisition system monitoring. Appl Energy 2009;86(12):2651–60. [105] Carullo Alessio, et al. In situ calibration of heterogeneous acquisition systems: the monitoring system of a photovoltaic plant. IEEE Trans Instrum Meas 2010;59(5) 1098-03. [106] Chouder A, Silvestre S. Automatic supervision and fault detection of PV systems based on power losses analysis. Energy Convers Manag 2010;51 (10):1929–37. [107] Ayompe LM, et al. Measured performance of a 1.72 kW rooftop grid connected photovoltaic system in Ireland. Energy Convers Manag 2011;52 (2):816–25. [108] Ranhotigamage Chagitha, Mukhopadhyay Subhas Chandra. Field trials and performance monitoring of distributed solar panels using a low-cost wireless sensors network for domestic applications. Sens J IEEE 2011;11(10):2583–90. [109] Mahjoubi Ammar, Mechlouch Ridha Fethi, Brahim Ammar Ben. Data acquisition system for photovoltaic water pumping system in the desert of Tunisia. Proc Eng 2012;33:268–77. [110] Carullo Alessio, Vallan Alberto. Outdoor experimental laboratory for longterm estimation of photovoltaic-plant performance. Instrum Meas, IEEE Trans 2012;61(5):1307–14. [111] Wittkopf Stephen, et al. Analytical performance monitoring of a 142.5 kW p grid-connected rooftop BIPV system in Singapore. Renew Energy 2012;47:9– 20. [112] LópezMEAMantiñanFJGMolinaMG. Implementation of wireless remote monitoring and control of solar photovoltaic (PV) system. In: Transmission and distribution: Latin America conference and exposition (T&D-LA); pp. 1–6; 2012. [113] Chouder Aissa, et al. Monitoring, modelling and simulation of PV systems using LabVIEW. Sol Energy 2013;91:337–49. [114] JonkePeterEderChristophStöcklJohannesSchwarkMichael. Development of a module integrated photovoltaic monitoring system. In: Industrial Electronics Society, IECON 2013-39th Annual Conference of the IEEE, 2013; pp. 8080– 8084. [115] Tina Giuseppe, Marco, Grasso Alfio Dario. Remote monitoring system for stand-alone photovoltaic power plants: the case study of a PV-powered outdoor refrigerator. Energy Convers Manag 2014;78:862–71. [116] Sanchez-Pacheco Francisco J, et al. PLC-Based PV plants smart monitoring system: field measurements and uncertainty estimation. Instrum Meas IEEE Trans 2014;63(9):2215–22. [117] Schill Christian, Brachmann Stefan, Koehl Michael. Impact of soiling on IVcurves and efficiency of PV-modules. Sol Energy 2015;11(2):259–62. [118] Andò B, Baglio S, Pistorio A, Tina GM, Ventura C. Sentinella: smart monitoring of photovoltaic systems at panel level. IEEE Trans Instrum Meas 2015;64 (8):2188–99. [119] Moreno-GarciaIMPallares-LopezVGonzalez-RedondoMLopez-LopezJVaroMartinezMSantiagoI. In: Implementation of a real time monitoring system for a grid-connected PV park. Industrial Technology (ICIT) IEEE International Conference on 2015;pp. 2915–20; 2015. [120] Han J, Lee I, Kim SH. User-friendly monitoring system for residential PV system based on low-cost power line communication. IEEE Trans Consum Electron 2015;61(2):175–80. [121] Park J, Shen J, Kim S, Rho D. Implementation of monitoring and control systems for 50kW PV systems using the wire-wireless. Network 2012;5 (1):45–54. [122] Benghanem M. A low cost wireless data acquisition system for weather station monitoring. Renew Energy 2010;35(4):862–72. [123] Mukaro Raphael, Carelse Xavier Francis. A serial communication program for accessing a microcontroller-based data-acquisition system. Comput Geosci 1997;23(9):1027–32. [124] Othman NA, Damanhuri NS, Ibrahim IR, Radzali R, Mohd MN. Automated monitoring system for small scale dual-tariff solar PV plant in UiTM Pulau Pinang. Proc World Congr Eng 2010;2:945–7. [125] Anwari M, Dom MM, Rashid MI. Small scale PV monitoring system software design. Energy Proc 2011;12:586–92. [126] Hsueh YH, Lee JS, KuoSheng-An LC, Chiang CC, Huang CL, Huang YH, Chen CT. A vectorized data communication system for localization in Zigbee sensor networks. Adv Sci Lett 2012;9(1):487–92. [127] Wang W, Xu Y, Khanna M. A survey on the communication architectures in

smart grid. Comput Netw 2011;55(15):3604–29. [128] Eke R, Kavasoglu AS, Kavasoglu N. Design and implementation of a low-cost multi-channel temperature measurement system for photovoltaic modules. Measurement 2012;45(6):1499–509. [129] BendelCKirchhofJHenzeN. Application of photovoltaic solar cells in planar antenna structures. In: Photovoltaic Energy Conversion. Proceedings of 3rd World Conference on 2003;Vol. 1, pp. 220–223; 2003. [130] Kato D, Nakamura T. Application of optical fibers to the transmission of solar radiation. J Appl Phys 1976;47(10):4528–31. [131] Liang D, Monteiro LF, Teixeira MR, Monteiro MF, Collares-Pereira M. Fiberoptic solar energy transmission and concentration. Sol Energy Mater Sol Cells 1998;54(1):323–31. [132] Kandilli C, Ulgen K. Review and modelling the systems of transmission concentrated solar energy via optical fibres. Renew Sustain Energy Rev 2009;13(1):67–84. [133] Jaramillo OA, Del Rio JA, Huelsz G. A thermal study of optical fibres transmitting concentrated solar energy. J Phys D: Appl Phys 1999;32(9):1000. [134] Kribus A, Zik O, Karni J. Optical fibers and solar power generation. Sol Energy 2000;68(5):405–16. [135] Zidani C, Benyoucef B, Madini N. Exergetic assessment of transmission– concentrated solar energy systems via optical fibres for building applications. Int J Energy 2012;11(2):216–28. [136] GuerrieroPValloneGPrimatoMDi NapoliFDi NardoLd'AlessandroVDalientoS. A wireless sensor network for the monitoring of large PV plants. In: Power Electronics, Electrical Drives, Automation and Motion (SPEEDAM) International Symposium on 2014;pp. 960–65; 2014. [137] TsagarisATriantafyllidisDG. Data monitoring system for supervising the performance assessment of a photovoltaic park. In: Computational Intelligence and Informatics (CINTI) IEEE 13th International Symposium on 2012;pp. 385– 89; 2012. [138] Tejwani R, Kumar G, Solanki C. Remote monitoring for solar photovoltaic systems in rural application using GSM voice channel. Energy Proc 2014;57:1526–35. [139] Ho QD, Gao Y, Le-Ngoc T. Challenges and research opportunities in wireless communication networks for smart grid. IEEE Wirel Commun 2013;20 (3):89–95. [140] Di NapoliFGuerrieroPd'AlessandroVDalientoS. A Power Line Communication on DC bus with photovoltaic strings. In: Renewable Power Generation Conference RPG, 3rd 2014;pp. 1–6. IET; 2014. [141] JonkePEderCStöcklJSchwarkM. Development of a module integrated photovoltaic monitoring system. InIndustrial Electronics Society, IECON 2013-39th Annual Conference of the IEEE;pp. 8080–84; 2013. [142] Han J, Choi CS, Park WK, Lee I, Kim SH. PLC-based photovoltaic system management for smart home energy management system. IEEE Trans Consum Electron 2014;60(2):184–9. [143] Son YS, Pulkkinen T, Moon KD, Kim C. Home energy management system based on power line communication. IEEE Trans Consum Electron 2010;56 (3):1380–6. [144] GoleniewskiL. Telecommunications essentials: the complete global source for communications fundamentals, data networking and the Internet, and nextgeneration networks. Addison-Wesley Professional; 2002. [145] IndraA. Fiber Optics in Solar Energy Applications. Available from: 〈https:// scholar.google.co.in/scholar?q ¼Fiber þOpticsþ in þ Solar þ Energyþ Applications&btnG ¼ &hl ¼ en&as_sdt ¼0%2C5〉 (accessed 20-05-15); 2015. [146] Tseng CL, Jiang JA, Lee RG, Lu FM, Ouyang CS, Chen YS, Chang CH. Feasibility study on application of GSM–SMS technology to field data acquisition. Comput Electron Agric 2006;53(1):45–59. [147] Bucci G, Fiorucci E, Landi C, Ocera G. Architecture of a digital wireless data communication network for distributed sensor applications. Measurement 2004;35(1):33–45. [148] MitchellB. How fast is a Wi-Fi Network? 25/06/2015. Available from: 〈http:// compnetworking.about.com/cs/wireless/f/wirelessspeed.htm〉 (accessed 0104-15); 2015. [149] Natkaniec M, Kosek-Szott K, Szott S, Bianchi G. A survey of medium access mechanisms for providing QoS in ad-hoc networks. IEEE Commun Surv Tutor 2013;15(2):592–620. [150] O’Keeffe M., RF standards for short-range wireless connectivity. Electron Prod., Garden City, NewYork, 2007;49:11. [151] Halonen T, Romero J, Melero J, (Eds.), GSM, GPRS and EDGE performance: evolution towards 3G/UMTS. John Wiley & Sons, Chichester; 2004. [152] VoskaridesSCPattichisCSIstepanianRMichaelidesCSchizasCN. Practical evaluation of GPRS use in a telemedicine system in Cyprus. In: Information Technology Applications in Biomedicine. 4th International IEEE EMBS Special Topic Conference on 2003;pp. 39–42; 2003. [153] Peersman C, Cvetkovic S, Griffiths P, Spear H. The global system for mobile communications short message service. IEEE Pers Commun 2000;7(3):15–23. [154] BergerGLafonGSerrieGSigogneC. Recent progress in lightning studies at the Pic du Midi observatory. In: Lightning (APL) 7th Asia-Pacific International Conference on 2011;pp. 381–86; 2011. [155] Tejwani R, Kumar G, Solanki C. Remote monitoring for solar photovoltaic systems in rural application using GSM voice channel. Energy Proc 2014;57:1526–35. [156] Galli S, Scaglione A, Wang Z. For the grid and through the grid: the role of power line communications in the smart grid. Proc IEEE 2011;99(6):998– 1027. [157] LiuJZhaoBWangJZhuYHuJ. Application of power line communication in smart

S.R. Madeti, S.N. Singh / Renewable and Sustainable Energy Reviews 67 (2017) 1180–1207

[158]

[159] [160] [161] [162]

[163]

[164]

[165]

[166]

[167]

[168]

[169]

[170]

[171]

[172]

[173]

[174] [175]

[176] [177] [178]

[179] [180]

[181]

power consumption. In: Power Line Communications and Its Applications (ISPLC) IEEE International Symposium on 2010;pp.303–07; 2010. SintonRACuevasAStuckingsM. Quasi-steady-state photo-conductance, a new method for solar cell material and device characterization. In: Photovoltaic Specialists Conference Record of the Twenty Fifth IEEE;pp.457–460; 1996. Ding J, Zhao J, Ma B. Remote monitoring system of temperature and humidity based on GSM. Image Signal Process 2009;2009:1–4. Al-Ali AR, Zualkernan I, Aloul F. A mobile GPRS-sensors array for air pollution monitoring. IEEE Sens J 2010;10(10):1666–71. Luecke, G. Analog and digital circuits for electronic control system applications: using the TI MSP430 microcontroller. Newnes; 2005. Han J, Lee I, Kim SH. User-friendly monitoring system for residential PV system based on low-cost power line communication. IEEE Trans Consum Electron 2015;61(2):175–80. Roman E, Alonso R, Ibañez P, Elorduizapatarietxe S, Goitia D. Intelligent PV module for grid-connected PV systems. IEEE Trans Ind Electron 2006;53 (4):1066–73. Koizumi H, Mizuno T, Kaito T, Noda Y, Goshima N, Kawasaki M, Nagasaka K, Kurokawa K. A novel microcontroller for grid-connected photovoltaic systems. IEEE Trans Ind Electron 2006;53(6):1889–97. Viral R, Khatod DK. Optimal planning of distributed generation systems in distribution system: a review. Renew Sustain Energy Rev 2012;16(7):5146– 65. Kim IS, Kim MB, Youn MJ. New maximum power point tracker using slidingmode observer for estimation of solar array current in the grid-connected photovoltaic system. IEEE Trans Ind Electron 2006;53(4):1027–35. Xiao W, Lind MG, Dunford WG, Capel A. Real-time identification of optimal operating points in photovoltaic power systems. IEEE Trans Ind Electron 2006;53(4):1017–26. Park JH, Ahn JY, Cho BH, Yu GJ. Dual-module-based maximum power point tracking control of photovoltaic systems. IEEE Trans Ind Electron 2006;53 (4):1036–47. International Electrotechnical Commission. Photovoltaic system performance monitoring-guidelines for measurement, data exchange and analysis International Standard IEC 61724; 1998.1998-1. Vergura S, Acciani G, Amoruso V, Patrono GE, Vacca F. Descriptive and inferential statistics for supervising and monitoring the operation of PV plants. IEEE Trans Ind Electron 2009;56(11):4456–64. VerguraSAccianiGAmorusoVPatronoG. Inferential statistics for monitoring and fault forecasting of PV plants. In: IEEE International Symposium on Industrial Electronics;pp. 2414–2419; 2008. VerguraSVaccaF. Bootstrap technique for analyzing energy data from PV plant. In: Clean Electrical Power International Conference on 2009;pp. 268– 75; 2009. Boquete L, Ascariz JM, Cantos J, Barea R, Miguel JM, Ortega S, Peixoto N. A portable wireless biometric multi-channel system. Measurement 2012;45 (6):1587–98. Castaner L, Silvestre S. Modelling photovoltaic systems using PSpice. Chichester: John Wiley and Sons; 2002. SilvestreSGuaschDOrtegaPCalatayudR. Photovoltaic systems modelling using Matlab and Simulink. In: Proceedings of the 19th European photovoltaic solar energy conference; pp. 2207–10; 2004. Mellit A, Kalogirou SA. Artificial intelligence techniques for photovoltaic applications: a review. Prog Energy Combust Sci 2008;34(5):574–632. Mellit Adel, Soteris A Kalogirou. Artificial intelligence techniques for photovoltaic applications: a review. Prog Energy Combust Sci 2008;34(5):574–632. Sun C, Hu X, Moura SJ, Sun F. Velocity predictors for predictive energy management in hybrid electric vehicles. IEEE Trans Control Syst Technol 2015;23(3):1197–204. Chao KH, Ho SH, Wang MH. Modeling and fault diagnosis of a photovoltaic system. Electr Power Syst Res 2008;78(1):97–105. KaratepeEHiyamaT Controlling of artificial neural network for fault diagnosis of photovoltaic array. In: Intelligent System Application to Power Systems (ISAP), {C}{C}{C}{C}{C}{C}2011 16th International Conference on 2011;pp.1–6. StettlerSToggweilerPRemundJ. SPYCE: satellite photovoltaic yield control and evaluation. InProceedings of the 21st European photovoltaic solar energy conference; pp.2613–2616; 2006.

1207

[182] Drews A, De Keizer AC, Beyer HG, Lorenz E, Betcke J, Van Sark WG, Heydenreich W, Wiemken E, Stettler S, Toggweiler P, Bofinger S. Monitoring and remote failure detection of grid-connected PV systems based on satellite observations. Sol Energy 2007;81(4):548–64. [183] HousseinAHeraudNSouleimanIPelletG. Monitoring and fault diagnosis of photovoltaic panels. InEnergy Conference and Exhibition (EnergyCon), 2010 IEEE International; pp. 389–394; 2010. [184] Yagi Y, Kishi H, Hagihara R, Tanaka T, Kozuma S, Ishida T, Waki M, Tanaka M, Kiyama S. Diagnostic technology and an expert system for photovoltaic systems using the learning method. Sol Energy Mater Sol Cells 2003;75 (3):655–63. [185] Zadeh LA. Fuzzy sets. Inf Control 1965;8(3):338–53. [186] HuX. Sun F. Fuzzy clustering based multi-model support vector regression state of charge estimator for lithium-ion battery of electric vehicle.in:Proceedings: Intelligent Human-Machine Systems and Cybernetics. IHMSC'09. International Conference on 2009;vol. 1:pp 392–96; 2009. [187] Jain Lakhmi C, Martin NM, (Eds.), Fusion of Neural Networks, Fuzzy Systems and Genetic Algorithms: Industrial Applications. CRC press, Boca Raton, FL; 1998. [188] Holland JH. Adaptation in natural and artificial systems: an introductory analysis with applications to biology, control, and artificial intelligence. MIT Press, Cambridge, MA; 1992. [189] Zhang L, Wang Z, Hu X, Sun F, Dorrell DG. A comparative study of equivalent circuit models of ultracapacitors for electric vehicles. J Power Sources 2015;274:899–906. [190] Ishibuchi H, Fujioka R, Tanaka H. Neural networks that learn from fuzzy ifthen rules. IEEE Trans Fuzzy Syst 1993;1(2):85–97. [191] Mamdani EH. Application of fuzzy algorithms for control of simple dynamic plant. Electr Eng Proc Inst 1974;121(12):1585–8. [192] Sugeno M. An introductory survey of fuzzy control. Inf Sci 1985;36(1):59–83. [193] Pavlidou N, Vinck AH, Yazdani J, Honary B. Power line communications: state of the art and future trends. IEEE Commun Mag 2003;41(4):34–40. [194] Hu Xiaosong, Li Shengbo, Yang Yalian. Advanced machine learning approach for lithium-ion battery state estimation in electric vehicles. IEEE Trans Transp Electrification 2016;2(2):140–9. [195] Suganthi L, Samuel Anand A. Energy models for demand forecasting—A review. Renew Sustain Energy Rev 2012;16(2):1223–40. [196] UNION, PEAN. Directive 2009/65/EC of the European Parliament and of the council. Official Journal of the European Union L 302 2009;16(2):33-40. [197] Ren21, Renewables. Global status report. Renewable Energy Policy Network for the 21st Century, Paris, France; September 2010. [198] Bennett, Bonnie, Mark Boddy, Frank Doyle, Mo Jamshidi, Ogunnaike I. Assessment study on sensors and automation in the industries of the future: Reports on industrial controls, information processing, automation, and robotics. DOE Energy Efficiency and Renewable Energy Industrial Technologies Program, USA; 2004. [199] Gungor VC, Hancke GP. Industrial wireless sensor networks: challenges, design principles, and technical approaches. IEEE Trans Ind Electron 2009;56 (10):4258–65. [200] Gungor VC, Lambert FC. A survey on communication networks for electric system automation. Comput Netw 2006;50(7):877–97. [201] Gungor Vehbi C, Lu Bin, Hancke Gerhard P. Opportunities and challenges of wireless sensor networks in smart grid. IEEE Trans Ind Electron 2010;57 (10):3557–64. [202] WohlgemuthJHKurtzS. Using accelerated testing to predict module reliability. In: 37th IEEE Photovoltaic Specialists Conference (PVSC):pp.3601– 3605; 2011. [203] ChamberlinCERocheleauMAMarshallMW. Comparison of PV module performance before and after 11 and 20 years of field exposure. In: 37th IEEE Photovoltaic Specialists Conference (PVSC) 2011;pp.101–105; 2011. [204] CarulloAParvisMVallanA. Security issues for Internetbased calibration activities, In: Proceedings. IEEE Instrum. Meas. Technol. Conference., Anchorage, AK: pp. 817–822; 2002. [205] Shariff Farihah, Rahim Nasrudin Abd, Hew Wooi Ping. Zigbee-based data acquisition system for online monitoring of grid-connected photovoltaic system. Expert Syst Appl 2015;42(3):1730–42.