a smart monitoring infrastructure.pdf

11
A smart monitoring infrastructure design for distributed renewable energy systems Ersan Kabalci Department of Electrical and Electronics Engineering, Faculty of Engineering and Architecture, Nevsehir Haci Bektas Veli University, Nevsehir, Turkey article info Article history: Received 29 September 2014 Accepted 30 October 2014 Available online 5 December 2014 Keywords: Power line communication (PLC) Renewable energy sources Smart grid Quadrature phase shift keying (QPSK) Automatic meter reading (AMR) abstract The automatic meter reading is essentially required in renewable grids as in conventional grids. It is intended to propose a reliable measurement system that is validated in a photovoltaic power system to meet the requirement of a renewable grid. In the presented study, the photovoltaic plants are controlled by using a widely known maximum power point tracking algorithm that is named as ‘‘Perturb and Observe’’. The distribution line at the output of inverter is modelled according to realistic parameters of 25 km line. Besides carrying the generated line voltage, the grid is used as a transmission medium for the generated power measurements of photovoltaic plants and power consumptions of load plants separately. The modem constituting the power line communication manages the dual-channel transfer and transmits the consumed energy ratios of the load plants. One of the modems is located at the output of voltage source inverter and the other one of the load plants. The power consumption values of each load plants are individually measured and successfully transmitted to monitoring section in the modelled system. The obtained data that is only used for monitoring in this application can also be evaluated for automatic meter reading applications. Ó 2014 Elsevier Ltd. All rights reserved. 1. Introduction The renewable energy usage is rapidly increasing since each economic crisis and source deficiencies cause to increments on per kWh cost of fossil-fuel based energy [1]. Although the govern- ments and private sector leaders discuss on cheaper energy gener- ation methods, those all agree to increase the renewable energy source (RES) shares to decrease energy generation costs [2]. The RES used in electricity generation include wind energy, solar energy, geothermal energy, and tidal waves. The income of any electricity generation and distribution company in the market is as important as its outcomes [3]. Since the conventional grid is a passive model, the remote detection of the consumed energy is prevented by the losses occurred on the distribution line. Researchers also extensively study the microgrid and load classifi- cation issues. Zhou et al. presented an optimal load distribution model and load classifications [4,5]. The main components of a conventional grid can be classified into five topics that are electric- ity generation plant, transmission substations, distribution substa- tions, control centre, and end-users. Although the conventional grid has many deficiencies to be solved such as voltage sags, blackouts, and overloads, several novel technologies are being researched in order to improve the qualifications of the grid tech- nology [6]. The smart grid concept, which is expanded since early 2000s, implies for a data communication network based on conventional grid that collects and carries the measured and modulated data of transmission, distribution, and consumption units [3,7]. The devel- opments seen in a conventional grid led the researchers to power line communication (PLC) and the smart grid concepts. The smart grid is assumed as a conversion of conventional grid to a commu- nication medium that carries the data obtained from remote sens- ing, control, and monitoring processes. These communication issues are expected to be performed in a secure and sustainable way over wired and/or wireless communication infrastructures [8]. The wireless technologies used in smart grid are WI-Fi, WiMax, ZigBee, and Bluetooth while the wired communication technolo- gies cover PLC, fiber optics, and copper wires [9–11]. Even though the wireless smart grids are more flexible compared to PLC, the communication may fail because of the probable problems such as interference, shadowing, and/or fading [12,13]. Furthermore, weather conditions directly affect the wireless network and cause to unexpected attenuations and several problems in transmission length. Zhang et al. stated several problems related to smart home management systems that are based on communication methods [14,15]. The PLC systems are categorized as narrow band http://dx.doi.org/10.1016/j.enconman.2014.10.062 0196-8904/Ó 2014 Elsevier Ltd. All rights reserved. E-mail address: [email protected] Energy Conversion and Management 90 (2015) 336–346 Contents lists available at ScienceDirect Energy Conversion and Management journal homepage: www.elsevier.com/locate/enconman

Upload: cesar-gil

Post on 17-Nov-2015

14 views

Category:

Documents


3 download

TRANSCRIPT

  • Energy Conversion and Management 90 (2015) 336346Contents lists available at ScienceDirect

    Energy Conversion and Management

    journal homepage: www.elsevier .com/locate /enconmanA smart monitoring infrastructure design for distributed renewableenergy systemshttp://dx.doi.org/10.1016/j.enconman.2014.10.0620196-8904/ 2014 Elsevier Ltd. All rights reserved.

    E-mail address: [email protected] KabalciDepartment of Electrical and Electronics Engineering, Faculty of Engineering and Architecture, Nevsehir Haci Bektas Veli University, Nevsehir, Turkeya r t i c l e i n f oArticle history:Received 29 September 2014Accepted 30 October 2014Available online 5 December 2014

    Keywords:Power line communication (PLC)Renewable energy sourcesSmart gridQuadrature phase shift keying (QPSK)Automatic meter reading (AMR)a b s t r a c t

    The automatic meter reading is essentially required in renewable grids as in conventional grids. It isintended to propose a reliable measurement system that is validated in a photovoltaic power systemto meet the requirement of a renewable grid. In the presented study, the photovoltaic plants arecontrolled by using a widely known maximum power point tracking algorithm that is named as Perturband Observe. The distribution line at the output of inverter is modelled according to realistic parametersof 25 km line. Besides carrying the generated line voltage, the grid is used as a transmission medium forthe generated power measurements of photovoltaic plants and power consumptions of load plantsseparately. The modem constituting the power line communication manages the dual-channel transferand transmits the consumed energy ratios of the load plants. One of the modems is located at the outputof voltage source inverter and the other one of the load plants. The power consumption values of eachload plants are individually measured and successfully transmitted to monitoring section in the modelledsystem. The obtained data that is only used for monitoring in this application can also be evaluated forautomatic meter reading applications.

    2014 Elsevier Ltd. All rights reserved.1. Introduction

    The renewable energy usage is rapidly increasing since eacheconomic crisis and source deficiencies cause to increments onper kWh cost of fossil-fuel based energy [1]. Although the govern-ments and private sector leaders discuss on cheaper energy gener-ation methods, those all agree to increase the renewable energysource (RES) shares to decrease energy generation costs [2]. TheRES used in electricity generation include wind energy, solarenergy, geothermal energy, and tidal waves. The income of anyelectricity generation and distribution company in the market isas important as its outcomes [3]. Since the conventional grid is apassive model, the remote detection of the consumed energy isprevented by the losses occurred on the distribution line.Researchers also extensively study the microgrid and load classifi-cation issues. Zhou et al. presented an optimal load distributionmodel and load classifications [4,5]. The main components of aconventional grid can be classified into five topics that are electric-ity generation plant, transmission substations, distribution substa-tions, control centre, and end-users. Although the conventionalgrid has many deficiencies to be solved such as voltage sags,blackouts, and overloads, several novel technologies are beingresearched in order to improve the qualifications of the grid tech-nology [6].

    The smart grid concept, which is expanded since early 2000s,implies for a data communication network based on conventionalgrid that collects and carries the measured and modulated data oftransmission, distribution, and consumption units [3,7]. The devel-opments seen in a conventional grid led the researchers to powerline communication (PLC) and the smart grid concepts. The smartgrid is assumed as a conversion of conventional grid to a commu-nication medium that carries the data obtained from remote sens-ing, control, and monitoring processes. These communicationissues are expected to be performed in a secure and sustainableway over wired and/or wireless communication infrastructures[8]. The wireless technologies used in smart grid are WI-Fi, WiMax,ZigBee, and Bluetooth while the wired communication technolo-gies cover PLC, fiber optics, and copper wires [911]. Even thoughthe wireless smart grids are more flexible compared to PLC, thecommunication may fail because of the probable problems suchas interference, shadowing, and/or fading [12,13]. Furthermore,weather conditions directly affect the wireless network and causeto unexpected attenuations and several problems in transmissionlength. Zhang et al. stated several problems related to smarthome management systems that are based on communicationmethods [14,15]. The PLC systems are categorized as narrow band

    http://crossmark.crossref.org/dialog/?doi=10.1016/j.enconman.2014.10.062&domain=pdfhttp://dx.doi.org/10.1016/j.enconman.2014.10.062mailto:[email protected]://dx.doi.org/10.1016/j.enconman.2014.10.062http://www.sciencedirect.com/science/journal/01968904http://www.elsevier.com/locate/enconman

  • E. Kabalci / Energy Conversion and Management 90 (2015) 336346 337(NB-PLC) and broad band (BB-PLC) according to operating bands[9,10]. The proposed study in this paper deals with a renewableenergy generation and distribution system and monitoring thepower consumption rates of various consumers by operatingmulti-carrier based PLC infrastructure.

    The generation part of the modelled system consists of threephotovoltaic (PV) plants that are assumed to be located in separatefields. The converted PV energy is conducted to a DC busbar andsupplies the input voltage of the inverter. The converted energyis distributed to load models over a transmission line that ismodelled with 25 km length. The power consumptions of two loadplants are observed over transmission line constituting the auto-matic meter reading (AMR) process. The quadrature phase shiftkeying (QPSK) modem at the energy generation plant is capableof demodulating the multi-channel input data on the various car-rier frequencies. The carrier frequencies are set to 6 kHz and8 kHz in the designed QPSK modulator systems and the demodula-tors are arranged to recover the carrier frequencies. It is possible toreconfigure the QPSK demodulator in case of increasing the num-ber of load plants. The renewable energy generation system isintroduced in the second section with PV energy plants, energyconversion part and distribution line subsections. The PLC infra-structure is handled in the third section with modelled QPSK mod-ulator and demodulator structures. The analysis methods,optimization steps, and analysis results are discussed in the fourthsection.2. Renewable energy generation system

    Fig. 1 illustrates an example of smart grid applications, which isa microgrid structure consisting the main components of a conven-tional grid system [16]. All components of a smart grid, such assensing, monitoring, protection, and control units can be seen inthe figure that is originally shown in [16].

    The complete schematic diagram of the modelled smart grid isseen in Fig. 2. The modelled system can be handled in three sec-tions as illustrated in energy generation, energy conversion andmonitoring, and microgrid distribution and load sections. Theenergy generation part is constituted with PV plant models, boostconverters, and DC busbar.

    The maximum power point tracking (MPPT) algorithm isalso included to boost converter part. The energy conversion andFig. 1. An example structure for wired and wireless smart gridmonitoring section covers a three-phase full bridge inverterbesides QPSK modem and monitoring part. The distribution linethat follows the energy conversion stage carries the line voltagesto the loads. The modelled system is tested with two different loadplants to figure out QPSK communication over the same transmis-sion line. The drawn current and the consumed power by loads aremeasured and are modulated at each load plant. The modulateddata are overlapped to line via coupling interfaces over S and Tphases namely after the measurement and modulation processesare done. The parts of modelled system are analysed in the follow-ing subsections in detail.

    2.1. Energy generation plants

    The analytical model of a PV panel is built using the electricalequivalent circuit that is seen in Fig. 3. The specific parameters ofPV panels are defined according to a model of Sharp that provides170W maximum output power [1719]. The developed modeladjusts the main parameters of PV panel such as short circuit cur-rent (Isc), open circuit voltage (Voc), cell number, maximum powercurrent (Ipm), and maximum power voltage (Vpm) referring to anyPV panel. The current of PV panel is determined using Eq. (1)[2022];

    Io Ipv ID expqVo IoRs

    mkBT

    1

    Vo IoRs

    Rsh1

    where, Io is output current of panel, Ipv is the generated PV current,ID is the diode current, Vo is the output voltage, VT is thermal volt-age, Rsh is shunt resistance, Rs is series resistance. In addition, q isthe electron charge, m is the equivalent ideality factor, kB is theBoltzmann constant, T is the cell temperature of the junction.

    Table 1 shows the PV panel parameters such as short circuitcurrent, open circuit voltage, maximum power voltage, and maxi-mum power current that are used in Simulink design. The currentvoltage (IV) characteristic curve of the modelled PV module isillustrated according to variable irradiation levels that vary from200W/m2 to 1000W/m2 in Fig. 4(a).

    The power characteristics generated according to same testconditions to define the maximum power point of PV module areshown in Fig. 4(b). The simulated curves verify that the modelledPV module operates properly to reference module of Sharp accord-ing to the same irradiation values. The modelled PV panels areapplications in a generation and distribution scheme [16].

  • Fig. 2. The schematic diagram of renewable grid and QPSK modems.

    Fig. 3. The electrical equivalent circuit of a PV cell.

    338 E. Kabalci / Energy Conversion and Management 90 (2015) 336346arranged to construct three PV plants seen on the left-hand side ofFig. 2. The DC output voltages of PV plants are supplied to boostconverters, which are controlled by an MPPT algorithm to obtainmaximum output power.Table 1PV panel and PV plant parameters.

    PV panel parametersOpen circuit voltage (Voc) Maximum power voltage (Vpm) Short circui43.2 V 34.8 V 5.47 A

    Parameters of PV Plant #1PV number in a series string Parallel string number Maximum p16 16 556.8 V78.

    Parameters of PV plant #2PV number in a series string Parallel string number Maximum p16 16 556.8 V78.

    Parameters of PV plant #3PV number in a series string Parallel string number Maximum p16 16 556.8 V78.

    Total output powerAlthough several MPPT techniques are implemented, the hill-climbing and Perturb and Observe (P&O) algorithms are widelypreferred among in the literature. Both the hill climbing and P&Ogenerate a perturbation to pursue the maximum power curve.

    The P&O adjusts the operating voltage of PV array while thehill-climbing algorithm deals with the duty cycle of the DCDCconverter. These two methods control the PV voltage and currentindividually. The duty cycle control performed by hill-climbingalgorithm changes the current value of the PV array, which causesseveral changes at the output voltage of DCDC converter. How-ever, the P&O algorithm compares the output power to previouslyacquired value to track the reference output power. Since the P&Ocontrols the operating voltage, it decreases or increases outputvoltage of DCDC converter to obtain the reference voltage at theoutput of converter system [18,23,24]. The P&O algorithm usedin the Simulink simulation is given in the Fig. 5 where it is basedon comparing the actual output power to previously acquiredpower level.

    If acquired power level (P(t)) is less than previously acquiredthen algorithm increases the operating voltage (Vref) else decreases.t current (Isc) Maximum power current (Ipm)4.9 A

    ower voltage and current (VpmIpm) Maximum output power of plant (W)4 A 43,653 W

    ower voltage and current (VpmIpm) Maximum output power of plant (W)4 A 43,653 W

    ower voltage and current (VpmIpm) Maximum output power of plant (W)4 A 43,653 W

    130,959 W

  • Fig. 4. Characteristics of the modelled PV module, (a) IV characteristic at variable irradiation from 200 W/m2 to 1000 W/m2, (b) power characteristic under same conditions.

    Fig. 5. MPPT algorithm of the boost converter.

    E. Kabalci / Energy Conversion and Management 90 (2015) 336346 339The previous current and voltage levels are controlled in terms ofequivalency, if there is not any change occurred in the power level.The duty cycle ratio of MPPT algorithm is limited up to 49% in orderto prevent to exceed the maximum current value that is suppliedby the converter. The switching devices of boost converters arecontrolled at 50 kHz switching frequency. The outputs of eachboost converter are connected to the DC busbar by coupling thepositive and negative outputs.2.2. Energy conversion subsection

    The stabilized output voltage of DCDC converters is supplied toa full-bridge inverter that generates the AC line voltage. A full-bridge inverter performs the DCAC conversion process owing toits Insulated-gate bipolar transistor (IGBT) switches. The switchingsignals are generated with an enhanced Sinusoidal Pulse WidthModulation (SPWM) that is introduced in [25] by author. Althoughthe regular SPWM control eliminates the base-band harmonics, itdisregards the side band harmonics and its multiples that arecaused by carrier signal. However, it is theoretically assumed thatthe bandwidth of a modulated signal is infinite [26,27] thatexplains the main reason of harmonics measured in a modulatedsignal. The enhanced SPWM decreases total harmonic distortion(THD) of current and voltage by considering the side-band har-monics that are generated by carrier signal. Besides the regularbase band harmonics, the side band harmonics are also eliminatedowing to Bessel filtering steps that are performed in the modulator.In the regular SPWM scheme, a modulating reference waveformthat is in sinusoidal form is compared to a triangular carrier wave-form to produce the switching sequences. Eq. (2) shows the Fourierseries that is used to define each switching angle to eliminate har-monic contents in the side-bands;

    Sswt a02Xn1

    1an cosnxt bn sinnxt 2

  • Table 2Transmission line parameters.

    Properties Value

    Frequency 50 HzResistance 0.0481 O/kmInductance 6.4938 104 H/kmCapacitance 1.2628 109 F/kmLength 25 km

    Fig. 6. Coupling circuit model between grid and modem.

    340 E. Kabalci / Energy Conversion and Management 90 (2015) 336346where a0 is the average DC value of the switching signal. The Fouriercoefficients a0, an, and bn are obtained by using the followingequations;

    a0 1p

    Z pp

    Sswtdt 3

    an 1p

    Z pp

    Sswt cosnxtdt 4

    bn 1p

    Z pp

    Sswt sinnxtdt 5

    Eq. (6) shows the cn coefficient that defines the nth ordered har-monic of Ssw(t) signal,

    cn an jbn 6

    The amplitude distortion that should be considered in modula-tor design is caused by the ripples of DC voltage source and drawsthe most significant impact on the onoff spectral errors. In caseof amplitude distortion occurs in the SPWM waveforms, thisdeclines the amplitude of the fundamental component and causesto unexpected low ordered harmonic contents as shown in Eq. (7).The output voltage of the inverter is expressed in Eq. (7) dependingto the modulation index, and Bessel functions of modulating andcarrier frequencies that eliminates the amplitude distortion.

    VOt miVdc2 cosxrt 2Vdcp

    X1k1

    J0k mi p2 sin k p2

    cosk xct

    2Vdcp

    X1k1

    X1l1

    Jn kmip2 k sin k l p2

    cosk xct l xrt

    9>>>>=>>>>;7

    where mi is modulation index, Vdc is dc supply voltage, xr is mod-ulator frequency, xc is carrier frequency, J0, Jn is the Bessel function.The term of modulation index (mi) given in Eq. (8) is the ratio of themodulating amplitude, Vm, to the carrier signal, Vc.

    mi VmVc

    8

    The line frequency of the inverter is adjusted depending on themodulating signal frequency. The line voltage of the inverter isdetermined by the modulation index (mi) which define the operat-ing area of inverter as linear modulation when the mi is lower than1 or as over-modulation when the mi is higher than 1 [28,29]. TheSPWMmodulator performs as a voltage amplifier in the linear mod-ulation range and the gain (G) is defined as given in Eq. (9),

    G 0:5miVdcVp

    9

    When themi is set to 1 in the control process, the gain rate increasesup to 78.55% of the peak value of the square voltage. The SPWMbased switching signals are arranged to operate the inverter inthe linear region, where the inverter voltage is calculated as givenin Eq. (10) [29,30]

    VAB VBC VCA miffiffiffi3

    pVd2

    0 < mi 6 1 10

    The carrier signal frequency that is generated in the modulator isset to 5 kHz to switch the IGBTs of the inverter. The impedancesof the transmission line, which are seen in Table 2, are determinedaccording to values per kilometer, according to that are used inindustrial applications by considering the line losses.

    These values are used in a simulation environment. Fig. 6 showsthe coupling interface that is used to inject the modem data to grid.The coupling circuit is built with an isolation transformer and itsparallel RLC network.

    3. Power line communication and QPSK modulation

    The digital symbol sequences are used to adjust one or morefeatures such as amplitude, frequency and phase parameters ofhigh frequency sinusoidal signal, which is called carrier [28]. Thedigital modulation term defines the data transmission that is per-formed using digital symbol sequences over the transmission med-ium. The phase shift keying (PSK), amplitude shift keying (ASK),and frequency shift keying (FSK) are known as the three main dig-ital modulation types. These classifications are based on shift key-ing process where it is called ASK if the shifted parameter isamplitude, FSK if the shifted parameter is the frequency or PSK ifthe shifted parameter is the phase. The PSK is the most insensitivedigital modulation scheme to the noise and interferences amongothers [30].

    3.1. Theory of quadrature phase shift keying

    The M-ary PSK modulation stands for PSK modulation type thatuses m separate carrier phase to perform transmission space. Thetransmission space is constituted by dividing the 2p radian phasespace to equal M-parts. The most widely used PSK scheme inM-ary structure is the quadrature phase shift keying (QPSK, or4PSK) since it is not affected from Bit Error Rate (BER) corruptionwhen the efficiency is improved. The QPSK scheme is accepted asthe main algorithm in the systems like digital subscriber line(DSL) modems, code division multiple access (CDMA), 3G, Wi-Fi(IEEE 802.11) and worldwide interoperability for microwave access(WiMAX (IEEE 802.16)) [30]. The mathematical description of aQPSK signal is given in Eq. (11);

    SQPSKt A cos2pf ct hi; 0 6 t 6 T; i 1;2;3;4: 11

    where hi stands for the carrier phase angle as hi 2i1pM 2i1p

    4 ,Eq. (12) is obtained by re-arranging Eq. (11);

    sQPSKt A cos 2pf ct 2i 1p

    4

    ;0 6 t 6 T; i 1;2;3;4 12

    When the trigonometric equality is applied to Eq. (12), the sQPSK(t) isdefined as seen in Eqs. (13) and (14),

    sQPSKt A coshi cos2pf ct A sinhi sin2pf ct 13

    sQPSKt s1tu1t s2tu2t 14

  • Fig. 8. Block diagram of QPSK modulator.

    Fig. 9. Block diagram of QPSK demodulator.

    E. Kabalci / Energy Conversion and Management 90 (2015) 336346 341where u1 and u2 are orthonormal basis functions as given in Eqs.(15) and (16), s1(t) and s2(t) are given in Eqs. (17) and (18),

    u1t ffiffiffi2T

    rcos2pf ct; 0 6 t 6 T 15

    u2t ffiffiffi2T

    rsin2pf ct; 0 6 t 6 T 16

    s1t Z T0

    sQPSKtu1tdt ffiffiffiE

    pcoshi 17

    s2t Z T0

    sQPSKtu2tdt ffiffiffiE

    psinhi 18

    The symbol energy which is shown with E parameter supports theequation of E = (1/2)A2T. The phase relation of s1(t) and s2(t) is,

    hi tan1s2s1

    19

    When the equations analysed from Eqs. (13)(19), QPSK signal isdefined as the total equation as seen in Eq. (20) according to thetime axis,

    sQPSKtA It cos2pf ctA Qt sin2pf ct; 1< t

  • Fig. 10. QPSK model developed in Simulink: (a) modulator, (b) demodulator.

    342 E. Kabalci / Energy Conversion and Management 90 (2015) 336346determined from DC busbar since output voltages of independentenergy conversion plants are connected on. Each PV plant gener-ates equal DC voltages around 550 V that are consolidated overinductive couplers and are connected to busbar at the stabilizedvoltage levels as seen in Fig. 11a.

    The last curve of Fig. 11a shows the busbar voltage while theleft three curves illustrate the output voltages of each PV plant.The waveforms seen in the first three curves of Fig. 11a show theeffects of MPPT algorithm depending to the PV plant and varyaround 1% of the generated voltage. The three-phase line voltagesmeasured at the output of inverter are seen in Fig. 11b where thesettling times of the outputs are around a half cycle (0.01 s). Themonitoring scenario of the proposed study is based on observingthe power consumption of two separate load plants that areassumed to be located at 25 km faraway to the PV plants.The power consumptions of each load plants are remotely mea-sured and are separately carried to QPSK modem by quantizing atthe measurement points. The measured and quantized power con-sumptions are seen in the first and second curves of Fig. 12a and brespectively, where each figure belongs to a different load plant.The measurement results show that the first plant consumesaround 20 kVA (19.6 kW), while the second site consumes around10 kVA (9.8 kW). The measured and quantized data are injected toAC transmission line over the coupling circuit and transmitted tomonitoring centre in order to be filtered and calibrated. The fourthcurves of Fig. 12a and b show the demodulated data observed inthe monitoring centre. The demodulator filter is quite effectiveon recovering the received signal depending to its original struc-ture. The most accurate recovered signals are obtained at 100 Hzcut-off frequency of the modelled modem.

  • Fig. 11. Line voltages, (a) DC voltages of PV plants, (b) AC voltages of inverter.

    E. Kabalci / Energy Conversion and Management 90 (2015) 336346 343The last curves of Fig. 12a and b depict the calibrated outputthat illustrates the actual power consumption of any load plant.The first and last curves seen in Fig. 12 are examined to evaluatethe success of modem since the first curves show the measuredvalue of consumed power at the load plants and the last curveshows the recovered data in the monitoring centre. Another mea-surement carried out to analyze the system is THD ratios of three-phase AC transmission line, which expresses the quality of the gen-erated voltage. The line current THD (THDi) analyses are performedconsidering the spectrum to 50th (Fig. 13a) and to 200th (Fig. 13b)

  • Fig. 12. PLC and monitoring data, (a) first load plant, (b) second load plant.

    344 E. Kabalci / Energy Conversion and Management 90 (2015) 336346

  • Fig. 13. Line current THD analysis, (a) up to 50th order, (b) up to 200th order.

    E. Kabalci / Energy Conversion and Management 90 (2015) 336346 345orders to investigate the sideband harmonics. The THD analysesshowed that there was not any higher-ordered harmonics up tothe 200th order of harmonics and the THD ratio of line current ismeasured at 0.84% which is proper to IEEE-519 and IEC-61000standards [35,36].The side band harmonics, those are especially considered as 3rd,5th, 7th, 9th, 11th, and 13th are measured around 0.05% and 0.02%ratios due to thedistortion attenuation characteristic of the developedSPWMmodulation scheme. The analyses also show that the transmis-sion line provides high-quality ac line voltage to the load plants.

  • 346 E. Kabalci / Energy Conversion and Management 90 (2015) 3363465. Conclusion

    The proposed system in this study deals with design of anadvanced monitoring infrastructure that is employed in a distrib-uted generation system of PV plants. Each PV plant is constitutedwith a number of serial and parallel PV arrays where the PV energyis converted to DC output voltages and is consolidated over busbar.The energy conversion part of the proposed study is implementedwith DCDC boost converters and a three-phase full bridge inver-ter. The transmission line is simulated with 25 km length betweeninverter and load plants. The power consumption rates of loadplants are measured instantly and are delivered to a monitoringcentre that is located at the inverter side of the grid. This operationis realized by the designed QPSK modem that generates data bymodulating the measured power and injects the modulated datato transmission line to transmit to the monitoring centre. Thereceived signal detected from the grid is filtered to recover themeasured data, and then calibrated to its actual value that is mea-sured at the load plant.

    The proposed system is intended to observe distribution lineagainst leakages, losses and power consumptions instantly. Theproposed system may also be used to calculate energy demandsand consumptions to plan the performance of a distribution lineor to perform the billing procedures. The proposed PLC infrastruc-ture eliminates unreliability issues, especially at long distancesthat are caused by interference, shadowing, fading, and similardegradation of wireless systems. The experimental study will becarried out in the future studies to validate the obtained valuesand the reliability of the system.

    References

    [1] Liao CH, Ou HH, Lo SL, Chiueh PT, Yu Y. A challenging approach for renewableenergy market development. Renew Sustain Energy Rev 2011;15:78793.

    [2] Yajvender PV, Ashwani K. Potential impacts of emission concerned policies onpower system operation with renewable energy sources. Int J Electr Power2013;44:5209.

    [3] Peng L, Yan G. Clean energy grid-connected technology based on smart grid.Energy Proc 2011;12:2138.

    [4] Zhou K, Yang S, Chen Z, Ding S. Optimal load distribution model of microgrid inthe smart grid environment. Renew Sustain Energy Rev 2014;35:30410.

    [5] Zhou K, Yang S, Shen C. A review of electric load classification in smart gridenvironment. Renew Sustain Energy Rev 2013;24:10310.

    [6] Gao J, Xiao Y, Liu J, Liang W, Chen CL. A survey of communication/networkingin smart grids. Fut Gen Compr Sys 2012;28:391404.

    [7] European Technology Platform Smart grids. strategic deployment documentfor Europes electricity networks of the future. Draft for 3rd General Asm; 2008.

    [8] Ettore B, Tao H, Yingjun W, Mihai C. Classification and trend analysis of threatsorigins to the security of power systems. Int J Electr Power 2013;50:5064.

    [9] Yu FR, Zhang P, Xiao W, Choudhury P. Communication systems for gridintegration of renewable energy resources. IEEE Network 2011;25:229.[10] Galli S, Scaglione A, Wang Z. For the grid and through the grid: the role ofpower line communications in the smart grid. Proc IEEE 2011;99:9981027.

    [11] Bouhafs F, Mackay M, Merabti M. Links to the future: communicationrequirements and challenges in the smart grid. IEEE Power Energy Mag2012;10:2432.

    [12] Zhang P, Li F, Bhatt N. Next-generation monitoring, analysis, and control forthe future smart control center. IEEE Trans Smart Grid 2010;1:18692.

    [13] Kent CB, Wirth A, Halgamuge S. Decision tree ensembles for online operationof large smart grids. Energy Convers Manage 2012;59:918.

    [14] Zhang Di, Shah Nilay, Papageorgiou Lazaros G. Efficient energy consumptionand operation management in a smart building with microgrid. EnergyConvers Manage October 2013;74:20922.

    [15] Lidula NWA, Rajapakse AD. Microgrids research: a review of experimentalmicrogrids and test systems. Renew Sust Energy Rev 2011;15:186202.

    [16] Environmental Commissioner of Ontario. Annual energy conservation progressreport 2010 (volume one): managing a complex energy system.

    [17] Kabalc E, Kabalc Y, Develi I. Modelling and analysis of a power linecommunication system with QPSK modem for renewable smart grids. Int JElectr Power 2012;34:1928.

    [18] Kabalc E, Kabalc Y. BPSK modem based power line communication system forobserving photovoltaic panels. In: Proceedings of 2nd int. conf. on nuclear andrenewable energy resources July 47, 2010, Ankara, Turkey. p. 2849.

    [19] Sharp NE-170UC1 Multipurpose Module. Sharp: USA; 2008.[20] Mellit A, Pavan AM. Performance prediction of 20 kWp grid-connected

    photovoltaic plant at Trieste (Italy) using artificial neural network. EnergyConvers Manage 2010;51:243141.

    [21] Rey-Bou AB, Valverde R, Ruz-Vila F, Torrelo-Ponce JM. An integrativeapproach to the design methodology for 3-phase power conditioners inphotovoltaic grid-connected systems. Energy Convers Manage 2012;56:8095.

    [22] Liu G, Nguang SK, Partridge A. A general modeling method for IVcharacteristics of geometrically and electrically configured photovoltaicarrays. Energy Convers Manage 2011;52:343945.

    [23] Muthuramalingam M, Manoharan PS. Comparative analysis of distributedMPPT controllers for partially shaded stand alone photovoltaic systems.Energy Convers Manage 2014;86(October):28699.

    [24] Piegari L, Rizzo R. Adaptive perturb and observe algorithm for photovoltaicmaximum power point tracking. IET Ren Power Gen 2010;4:31728.

    [25] olak I, Kabalc E. Practical implementation of a DSP controlled multilevelinverter with low THD for motor drive applications. J Power Sources2011;196:758593.

    [26] Benvenuto N, Cherubini G. Algorithms for communications systems and theirapplications. England: Wiley & Sons Publishing; 2002.

    [27] Peebles PZ. Digital communication systems. USA: Prentice-Hall; 1987.[28] Hobraiche J, Vilain JP, Macret P, Patin N. A new PWM strategy to reduce the

    inverter input current ripples. Trans Power Electr 2009;24:17280.[29] Colak _I, Kabalci E, Bayindir R, Bal G. Modelling of a three phase SPWM

    multilevel VSI with low THD using Matlab/Simulink. In: Proceedings of EPE2009, 810 September, 2009, Barcelona, Spain. p. 110.

    [30] olak I, Kabalc E, Bal G. Parallel DCAC conversion system based on separatesolar farms with MPPT control. In: Proceedings of IEEE 8th int conf on powerelec, May 30June 3 2011, Jeju Korea. p. 146975.

    [31] Xiong F. Digital modulation techniques. 2nd ed. USA: Artech House; 2000.[32] Ho SW. Adaptive modulation (QPSK, QAM). Intel Corporation; 2004.[33] Glover I, Grant P. Digital communications. 2nd ed. UK: Prentice Hall; 2000.[34] Steber JM. PSK demodulation (Part 1). Watkins-Johnson Company; 1984. p. 1

    10.[35] IEEE Standard 519-1992. Recommended practices and requirements for

    harmonic control in electrical power systems. The IEEE; 1993.[36] Abidin MNZ. IEC 61000-3-2 Harmonics Standards Overview, Schaffner EMV

    AG; 2010.

    http://refhub.elsevier.com/S0196-8904(14)00957-1/h0005http://refhub.elsevier.com/S0196-8904(14)00957-1/h0005http://refhub.elsevier.com/S0196-8904(14)00957-1/h0010http://refhub.elsevier.com/S0196-8904(14)00957-1/h0010http://refhub.elsevier.com/S0196-8904(14)00957-1/h0010http://refhub.elsevier.com/S0196-8904(14)00957-1/h0015http://refhub.elsevier.com/S0196-8904(14)00957-1/h0015http://refhub.elsevier.com/S0196-8904(14)00957-1/h0020http://refhub.elsevier.com/S0196-8904(14)00957-1/h0020http://refhub.elsevier.com/S0196-8904(14)00957-1/h0025http://refhub.elsevier.com/S0196-8904(14)00957-1/h0025http://refhub.elsevier.com/S0196-8904(14)00957-1/h0030http://refhub.elsevier.com/S0196-8904(14)00957-1/h0030http://refhub.elsevier.com/S0196-8904(14)00957-1/h0040http://refhub.elsevier.com/S0196-8904(14)00957-1/h0040http://refhub.elsevier.com/S0196-8904(14)00957-1/h0045http://refhub.elsevier.com/S0196-8904(14)00957-1/h0045http://refhub.elsevier.com/S0196-8904(14)00957-1/h0050http://refhub.elsevier.com/S0196-8904(14)00957-1/h0050http://refhub.elsevier.com/S0196-8904(14)00957-1/h0055http://refhub.elsevier.com/S0196-8904(14)00957-1/h0055http://refhub.elsevier.com/S0196-8904(14)00957-1/h0055http://refhub.elsevier.com/S0196-8904(14)00957-1/h0060http://refhub.elsevier.com/S0196-8904(14)00957-1/h0060http://refhub.elsevier.com/S0196-8904(14)00957-1/h0065http://refhub.elsevier.com/S0196-8904(14)00957-1/h0065http://refhub.elsevier.com/S0196-8904(14)00957-1/h0070http://refhub.elsevier.com/S0196-8904(14)00957-1/h0070http://refhub.elsevier.com/S0196-8904(14)00957-1/h0070http://refhub.elsevier.com/S0196-8904(14)00957-1/h0075http://refhub.elsevier.com/S0196-8904(14)00957-1/h0075http://refhub.elsevier.com/S0196-8904(14)00957-1/h0085http://refhub.elsevier.com/S0196-8904(14)00957-1/h0085http://refhub.elsevier.com/S0196-8904(14)00957-1/h0085http://refhub.elsevier.com/S0196-8904(14)00957-1/h0085http://refhub.elsevier.com/S0196-8904(14)00957-1/h0085http://refhub.elsevier.com/S0196-8904(14)00957-1/h0100http://refhub.elsevier.com/S0196-8904(14)00957-1/h0100http://refhub.elsevier.com/S0196-8904(14)00957-1/h0100http://refhub.elsevier.com/S0196-8904(14)00957-1/h0100http://refhub.elsevier.com/S0196-8904(14)00957-1/h0105http://refhub.elsevier.com/S0196-8904(14)00957-1/h0105http://refhub.elsevier.com/S0196-8904(14)00957-1/h0105http://refhub.elsevier.com/S0196-8904(14)00957-1/h0105http://refhub.elsevier.com/S0196-8904(14)00957-1/h0110http://refhub.elsevier.com/S0196-8904(14)00957-1/h0110http://refhub.elsevier.com/S0196-8904(14)00957-1/h0110http://refhub.elsevier.com/S0196-8904(14)00957-1/h0115http://refhub.elsevier.com/S0196-8904(14)00957-1/h0115http://refhub.elsevier.com/S0196-8904(14)00957-1/h0115http://refhub.elsevier.com/S0196-8904(14)00957-1/h0120http://refhub.elsevier.com/S0196-8904(14)00957-1/h0120http://refhub.elsevier.com/S0196-8904(14)00957-1/h0125http://refhub.elsevier.com/S0196-8904(14)00957-1/h0125http://refhub.elsevier.com/S0196-8904(14)00957-1/h0125http://refhub.elsevier.com/S0196-8904(14)00957-1/h0125http://refhub.elsevier.com/S0196-8904(14)00957-1/h0130http://refhub.elsevier.com/S0196-8904(14)00957-1/h0130http://refhub.elsevier.com/S0196-8904(14)00957-1/h0135http://refhub.elsevier.com/S0196-8904(14)00957-1/h0140http://refhub.elsevier.com/S0196-8904(14)00957-1/h0140http://refhub.elsevier.com/S0196-8904(14)00957-1/h0155http://refhub.elsevier.com/S0196-8904(14)00957-1/h0165http://refhub.elsevier.com/S0196-8904(14)00957-1/h0185http://refhub.elsevier.com/S0196-8904(14)00957-1/h0185

    A smart monitoring infrastructure design for distributed renewable energy systems1 Introduction2 Renewable energy generation system2.1 Energy generation plants2.2 Energy conversion subsection

    3 Power line communication and QPSK modulation3.1 Theory of quadrature phase shift keying3.2 Simulink model of the QPSK modem

    4 Simulation and comprehensive analysis5 ConclusionReferences