Benchmarking of Grid Fault Modes in Single-Phase Grid-Connected Photovoltaic Systems

IEEE TRANSACTIONS ON INDUSTRY APPLICATIONS, VOL. 49, NO. 5, SEPTEMBER/OCTOBER 2013 2167

Benchmarking of Grid Fault Modes in Single-PhaseGrid-Connected Photovoltaic Systems

Yongheng Yang, Student Member, IEEE, Frede Blaabjerg, Fellow, IEEE, and Zhixiang Zou, Student Member, IEEE

Abstract—Pushed by the booming installations of single-phasephotovoltaic (PV) systems, the grid demands regarding the inte-gration of PV systems are expected to be modified. Hence, thefuture PV systems should become more active with functionalitiesof low-voltage ride through and grid support capability. Thecontrol methods, together with grid synchronization techniques,are responsible for the generation of appropriate reference signalsin order to handle ride-through grid faults. Thus, it is necessaryto evaluate the behaviors of grid synchronization methods andcontrol possibilities in single-phase systems under grid faults.The intent of this paper is to present a benchmarking of gridfault modes that might come in future single-phase PV systems.In order to map future challenges, the relevant synchronizationand control strategies are discussed. Some faulty modes are stud-ied experimentally and provided at the end of this paper. It isconcluded that there are extensive control possibilities in single-phase PV systems under grid faults. The second-order-general-integral-based phase-locked-loop technique might be the mostpromising candidate for future single-phase PV systems becauseof its fast adaptive-filtering characteristics and it is able to fulfillfuture standards.

Index Terms—Grid requirements, grid support, grid synchro-nization, low-voltage ride through (LVRT), phase-locked loop(PLL), single-phase photovoltaic (PV) systems.

I. INTRODUCTION

THE installation of single-phase photovoltaic (PV) systemshas been booming in recent years because of the matured

PV technology and the declined price of PV panels [1], [2].Pushed by the high penetration of renewable energy systems,many grid requirements have been released in order to regulateinterconnected renewable power generation [4]–[8]. Some ba-sic requirements are defined in the grid regulations, like powerquality, frequency stability, and voltage stability [3], [9], [16],and even more specific demands for wind turbines or high-voltage systems have been issued [6].

Traditionally, the grid-connected PV systems are small scaleat a residential level and designed to disconnect from the gridwithin a certain time tripping by a grid fault [3]. However, due

Manuscript received August 2, 2012; revised November 18, 2012; acceptedJanuary 1, 2013. Date of publication April 29, 2013; date of current versionSeptember 16, 2013. Paper 2012-IPCC-464.R1, presented at the 2012 IEEEEnergy Conversion Congress and Exposition, Raleigh, NC, USA, September15–20, and approved for publication in the IEEE TRANSACTIONS ON INDUS-TRY APPLICATIONS by the Industrial Power Converter Committee of the IEEEIndustry Applications Society.

Y. Yang and F. Blaabjerg are with Aalborg University, 9220 Aalborg East,Denmark (e-mail: [email protected]; [email protected]).

Z. Zou is with Southeast University, Nanjing 210096, China (e-mail:[email protected]).

Color versions of one or more of the figures in this paper are available onlineat http://ieeexplore.ieee.org.

Digital Object Identifier 10.1109/TIA.2013.2260512

to the thriving scenario of large-scale grid-connected single-phase PV systems in many distributed installations, the discon-nection could cause adverse conditions and negatively impactthe reliability, stability, and availability of the distributed grid[10]–[15]. For instance, the voltage fault may cause lightingflickers, low voltage, and power quality problems, leading to theloss in energy production and the necessity of PV integrationlimitation. However, if the grid-connected single-phase PV sys-tems can provide ancillary services, such as reactive power sup-port and low-voltage ride through (LVRT) capability [16]–[21],the customers will not experience many flickers and powerquality issues anymore, and the distributed system operatorswill not need to limit the PV integration into their grids. Itis expected that, in the near future, the grid-connected PVsystems should become more active and more “smart” withsuch functionalities because of the high penetration of PVsystems.

In that case, the control methods should be ready for single-phase PV applications because they are responsible for generat-ing appropriate reference signals in order to handle ride-throughgrid faults, which means that an evaluation and benchmarkingof possible control strategies for single-phase applications arenecessary. Practically, the single-phase PQ theory [22], [23]could be adopted in the control system. By regulating themaximum power point, the active power could be controlledwithin the boundaries in order to avoid overcurrent trippingunder grid voltage sags in such a way to enhance the LVRTcapability. Furthermore, the droop control methods could beused to adjust the active and reactive powers as reported in [24]and [25].

Moreover, as the prerequisite of a good control, the syn-chronization technique for single-phase PV systems has alsobecome of high interest. Since a voltage fault is normally ashort period, an accurate and fast synchronization method willensure a good performance of the whole PV system in the gridfaulty mode operation. Recent research demonstrates that thephase-locked loop (PLL)-based synchronization methods havemore attractiveness for such applications [6], [8], [26]–[30].Among these, the adaptive mechanism-based techniques gainmore attention because of their high robustness and fast re-sponse characteristics. Such kinds of methods may be the bestcandidates for single-phase PV systems operating in faulty-gridmodes. However, it may also cause undesired influences, whichhave been discussed in [31].

The objective of this paper is to study the performanceof single-phase grid-connected PV systems under grid faultsdefined by the basic grid codes of wind turbine systems con-nected to the grid. First, an overview of the existing grid

0093-9994 © 2013 IEEE

2168 IEEE TRANSACTIONS ON INDUSTRY APPLICATIONS, VOL. 49, NO. 5, SEPTEMBER/OCTOBER 2013

requirements is presented. Particular attention is paid on thepossible control strategies, which may help the single-phasePV systems to handle ride-through grid faults or operate underabnormal grid conditions. It is followed by an evaluation ofthe synchronization methods. Finally, faulty cases are simulatedand validated experimentally.

II. OVERVIEW OF SELECTED GRID REQUIREMENTS

One essential basis of the design and control for grid-connected PV inverters is the grid requirements. In some inter-national regulations [3], it is addressed that PV inverters shoulddisconnect from the grid in the presence of abnormal grid con-ditions in terms of voltage and frequency at the point of com-mon coupling (PCC). These requirements, including islandingprotection, are designed based on a low-level penetration of PVsystems and are set to ensure the safety of utility maintenancepersonnel and also the grid. Compared to the conventionalpower plants and wind power systems, typically PV systemsare connected to low-voltage and/or medium-voltage networks[2]. In this case, such grid requirements are valid and enough.

However, considering the impact of large-scale PV systemson a distributed grid to which they are connected, these gridrequirements are supposed to be revised or extended with somecombined standardized features as well as custom demands.Because the disconnections from the distributed grid can affectthe stability of the whole system and cause negative impactson customers’ equipment, several European countries haveupdated the grid requirements for medium- or high-voltagesystems. For instance, the German grid code requires that thesystems connected to the medium- or high-voltage networksshould have the capabilities of LVRT and grid support function-ality during grid faults [4], [5]. In the new Italian grid code, it isrequired that the generation units connected to the low-voltagegrid with the nominal power exceeding 6 kVA should have theability to ride through grid voltage faults [16].

Therefore, it is better for PV inverters to be equipped withLVRT capability in order to improve the operation of the powerconverters and the reliability of the whole system. It is expectedthat the aforementioned regulations will be extended for large-scale low-voltage PV applications [12]–[14], [17]–[21]. Similarto wind turbine power generations connected to the medium-and high-voltage levels, single-phase PV generation systemssupplying low-voltage networks in the future are supposed tomake a contribution to the network by means of also ridingthrough grid faults.

Different LVRT curves of a defined stay-connected time arepresented in Fig. 1. As it is noticed in Fig. 1, the generationsystems required in the German grid code should be capableof riding through a 0.15-s voltage fault when the grid voltageamplitude presents a drop to 0 V and inject some reactivecurrent IQ into the grid as well. The required reactive current IQto support the voltage in the German grid regulation is shownin Fig. 2, and it can be given as

IQ =

⎧⎨⎩

deadband, 0.9 p.u. ≤ U < 1.1 p.u.k · U−U0

UN· IN+IQ0, 0.5 p.u. ≤ U < 0.9 p.u.

−IN + IQ0, U < 0.5 p.u.(1)

Fig. 1. LVRT requirements of wind power systems of different countries [4].

Fig. 2. Voltage support requirements in the event of grid faults for windturbine systems [5].

where U , U0, and UN are the instantaneous voltage, the initialvoltage before grid faults, and the nominal voltage and IN andIQ0 are the nominal current and the reactive current before agrid failure.

III. CONTROL POSSIBILITIES UNDER GRID FAULTS

The traditional control strategy applied to the single-phaseconverter system includes two cascaded loops: an inner currentloop which is responsible for power quality issues and currentprotection [6], [8], [32] and an outer voltage control loop. In thiscase, it is possible to add control methods into the inner loop insingle-phase systems in grid faulty mode operations to supportthe grid. The overall structure of a single-phase grid-connectedPV system is given in Fig. 3.

In respect to the control of a three-phase system under gridfaults, four major methods are reported in the literature: unitypower factor control, positive sequence control, constant activepower control, and constant reactive power control [8], [32].These methods are not suitable for single-phase applicationssince it is difficult to employ directly a dq-rotating synchronousreference frame.

One possible solution to the single-phase case is inspiredby the orthogonal signal generator (OSG)-based PLL principle[6], [8], [32]. According to the single-phase active and reactivepower theory [22], [23], the components, vα and vβ , generated

YANG et al.: BENCHMARKING OF GRID FAULT MODES IN SINGLE-PHASE GRID-CONNECTED PV SYSTEMS 2169

Fig. 3. Overall control structure of a single-phase grid-connected PV system.

Fig. 4. Control diagram of single-phase systems under grid faults based onthe single-phase PQ theory and the OSG concept.

by the OSG system can be used to calculate the active andreactive powers as given by{

P = 12 (vαiα + vβiβ)

Q = 12 (vβiα − vαiβ)

(2)

where iαβ and vαβ are the grid current and voltage in theαβ system and P and Q are the active and reactive powers,respectively. Thus, by this mean, the current reference can begenerated as it is expressed as[

i∗αi∗β

]=

2

v2α + v2β

[vα vβvβ −vα

] [P ∗

Q∗

](3)

in which “∗” denotes the reference signal. Then, the detailedcontrol diagram based on the single-phase PQ theory and theOSG concept can be illustrated as it is shown in Fig. 4.

Notably, in this control system, the existing current controlmethods, such as proportional resonant (PR), resonant control,deadbeat control, repetitive controller, and hysteresis controlcan be adopted in the faulty-grid cases. Moreover, by employ-ing the Park transform (αβ−dq) to the grid current and thegrid voltage, the dc quantities of the current and voltage areobtained in the rotating synchronous reference frame, leading tothe possible use of the basic proportional–integral (PI) controlfor the current or power regulation [23], [33], [34]. The “QProfile” shown in Fig. 4 is in compliance with the grid codes asdescribed by (1) and in Fig. 2. The reference reactive power Q∗

is generated according to the voltage sag depth detected by thesynchronization units or a grid fault detection scheme, which

Fig. 5. Simplified single-phase grid-connected PV system.

means that the “Q Profile” is triggered by the detected voltageamplitude Vg .

Another control possibility is based on the concept of thefrequency and voltage droop control through active and reactivepowers, respectively. A droop control method could be adoptedto adjust the active and reactive powers in single-phase applica-tions under grid faults. It can be illustrated using the simplifiedgrid-connected PV system as shown in Fig. 5.

On the basis of the assumptions that the line impedanceis mainly inductive (XL � RL) and the power angle φ isvery small, the active power P and reactive power Q can beexpressed by [24], [25]{

P =VinVg

XLsinφ ≈ VinVg

XLφ

Q =VinVg cosφ−V 2

g

XL≈ (Vin−Vg)Vg

XL

(4)

where XL is the line reactance. Hence, the inverter voltagereference v∗in can be obtained, and it is controllable through theangle φ and the amplitude Vin by respectively regulating theactive and reactive powers with simple PI controllers. A droopcontroller can be given as{

φ = φ∗ −G1(s)(P − P ∗)Vin = V ∗

in −G2(s)(Q−Q∗)(5)

in which “∗” indicates the reference signal and G1(s) andG2(s) are the PI controllers that can control the active andreactive power sharing between the PV inverter and the grid.

This kind of control approach used to support the grid voltageunder a grid voltage sag is successfully tested in [24] and[25], where the PV inverter is working as a shunt device anddesigned to mitigate the grid voltage drops and the harmonicdistortions. However, since the single-phase PV systems arenormally connected to low-voltage distribution networks, theline is more resistive rather than inductive. Therefore, a largeinductor is required between the grid and the PV inverter inthis control strategy; otherwise, the voltage sag cannot be wellcompensated. This is the main weak point of such a control.Thus, in this paper, the single-phase PQ theory-based controlmethod is adopted.

It is also worth to know that the active power delivered tothe grid is limited by the inverter nominal current. Therefore, toavoid inverter shutdown because of the overcurrent protection,the PV panels should not operate in the maximum power pointtracking (MPPT) mode depending on the solar irradiation,which can be illustrated in Fig. 6. It is shown that, in the gridfaulty mode operation, the active power should be limited inorder to deliver the required reactive power without triggeringthe inverter overcurrent protection. Nevertheless, this aspect


Fig. 6. Limitation of active and reactive powers drawn from PV panels(dashed black: MPPT operation; dot-dashed red: grid faulty mode). (a) PVcharacteristics. (b) PQ characteristics.

could be used for reactive power support, e.g., during the nightwhen there is no solar irradiance [20].

Additionally, the double-frequency term presenting at the dcside (PV side) in single-phase systems will also have a negativeimpact on the control systems both under normal operation andin grid faulty mode operation [8]. Consequently, the designof the controllers, modulation techniques, and grid-interfacedcurrent filters (L, LC, or LCL) should be done in considerationof producing lower switching voltage stress and lower voltageripple at the dc link.

Anyway, there are extensive control possibilities in single-phase grid-connected PV systems, which are able to meet theupcoming requirements defined in the grid codes. Regardingsingle-phase PV systems with grid support and LVRT func-tionalities, the control method should be capable of providingaccurate and appropriate references without exceeding the dcnominal voltage, tripping the current protection due to constantactive power delivery, and failing to synchronize in compliancewith these demands in the near future.

IV. GRID SYNCHRONIZATION TECHNIQUES FOR

SINGLE-PHASE APPLICATIONS

The synchronization scheme plays a major role in the controlof single-phase systems under grid faults. A good synchroniza-tion system should respond to a voltage drop immediately whena phase-to-ground fault occurs at PCC as shown in Fig. 3. Manysynchronization methods are reported in the recent literature[6], [8], [26]–[30], which can be divided into two categories:mathematical analysis methods (e.g., Fourier analysis-basedsynchronization method) and PLL-based methods. Nowadays,the PLL-based synchronization methods have more attractive-ness. However, the main difference among various single-phase PLL methods is the configuration of the phase detector(PD), intuitively being a simple sinusoidal multiplier [26], [29].However, this process will produce a double-frequency term ina single-phase system.

Applying the Park transform to an OSG system is anotherway to extract the phase error for PLL-based methods. Hence,the task will be shifted to establish the OSG system. Suchkinds of PLL are reported in the literature, like T/4 delayPLL [6], [8], [27] and inverse Park transform-based PLL [6],[26], [27]. Another possibility is to use adaptive filters (AFs)which can self-adjust the output according to an error feedback

Fig. 7. Basic structure of a PLL.

loop. Two popular PLLs, the enhanced PLL (EPLL) [26],[34], [35] and the second-order generalized integrator (SOGI)-based PLL (SOGI-OSG) [6], [8], [27], [37], are based on thecombinations of AFs with a sinusoidal multiplier and an OSGsystem.

A basic PLL structure is given in Fig. 7, which consistsof a PD, a PI-based loop filter (LF), and a voltage-controlledoscillator (VCO). Thus, the small-signal model of this systemcan be given as

Θ̂(s)

Θ(s)=

Kps+Ki

s2 +Kps+Ki(6)

where Θ̂(s) and Θ(s) are the output and input phase, respec-tively, and Kp and Ki are the proportional and integral gains ofthe LF. The details of the PLL modeling can be found in [6].From (6), the settling time can be given by ts = 9.2/Kp, whichis adopted to evaluate the performance of different PLLs in thispaper. The following section will compare the selected PLLsand find the best one for the application in this paper.

A. T/4 Delay PLL

This PLL approach takes the input voltage vg as the “α”component in a “αβ” system, while the “β” component canbe obtained simply by introducing a phase shift of π/2 rad withrespect to the fundamental frequency of the input voltage. Thus,the Park transform can be employed to detect the phase error,which is expressed as the following:[

vdvq

]=

[cos θ̂ sin θ̂− sin θ̂ cos θ̂

] [vgvβ

]

=

[Vm sin(Δθ)−Vm cos(Δθ)

]≈

[VmΔθ−Vm

](7)

where vg = Vm sin(θ) = Vm sin(ωt+ φ), in which Vm, θ, ω,and φ are the amplitude, phase, frequency, and phase angle ofthe input signal vg , Δθ = θ − θ̂ is the detected phase error, andθ̂ is the locked phase.

The error Δθ is very small in steady state, and then, thelinearized equation shown in the very right side of (7) isobtained. The structure of the T/4 delay PLL is given in Fig. 8,where T and ω0 are the period and nominal frequency of theinput voltage vg .

B. EPLL

The EPLL introduced in [31] and [35] is based on a simpleAF, which can refine the transfer function according to afeedback algorithm driven by an error signal. It can be used totrack the input voltage in terms of amplitude Vm and phase θ.


Fig. 8. Structure of the T/4 delay PLL.

Fig. 9. AF-based PD of the EPLL.

The adaptive process is to minimize a so-called objectivefunction by modifying the filter parameters. Then, the ampli-tude is estimated. Define the objective function as

E(V̂m, θ̂) =1

2e2 =

1

2(vg − v̂g)

2 (8)

in which V̂m and θ̂ are the estimated amplitude and the lockedphase of the input voltage, respectively. Then, the desiredoutput of the filter can be expressed as v̂g = V̂m sin θ̂.

In order to minimize the objective function, the popularleast-mean-square adaptive algorithm is used [36]. Then, thefollowing differential equation is obtained [31]:

˙̂V m = −μ

∂E(V̂m, θ̂)

∂V̂m

= μe sin θ (9)

where μ is the control parameter. Subsequently, the PD imple-mentation of the EPLL can be given in Fig. 9.

One important feature of the EPLL concluded from theaforementioned discussion is that the output signal v̂g is lockedboth in phase and in amplitude compared to the conventionalPLL methods [35]. However, the performance, such as thespeed of the estimation process, is exclusively dependent on thecontrol parameter μ. By linearizing (9), this relationship can beobtained as [35]

V̂m(s)

Vm(s)=

1

τs+ 1(10)

where τ = 2/μ is the time constant.The response of such an AF in the EPLL system with

different time constants is shown in Fig. 10. It is noticed thata large value of μ will make the estimated output signal cometo steady state quickly, but it will have a high overshoot offrequency if μ is too large. The settling time of this system canapproximately be calculated as follows: 4τ = 8/μ.

C. SOGI-Based PLL

Another adaptive-filtering-based PLL solution is using SOGIto create the OSG system, commonly known as SOGI-OSG

Fig. 10. Response of the AF of an EPLL with different μ (different timeconstant τ ).

Fig. 11. PD of the SOGI PLL.

PLL [6], [32], [37]. The general OSG structure of SOGI-OSG PLL is depicted in Fig. 11, in which ω̂ is the estimatedfrequency of the input signal, qv̂g is the orthogonal signal withrespect to the input voltage vg, and ke is the control parameter.

The EPLL discussed previously is using only one-weight AF,which is the simplest one. If two-weight AFs are adopted insingle-phase applications, it will present a better performance,and it behaves like a “sinusoidal integrator” [6], [37], [38]. Thetransfer function of such kind of AF can be expressed as

GAF (s) =s

s2 + ω̂2. (11)

Multiplied by ω̂ which is defined previously, it shares thetransfer function of a SOGI in common [37], [39].

Thus, referring to Fig. 11, the closed-loop transfer functionsof the SOGI-OSG PLL can be obtained as{

v̂g(s)vg(s)

= keω̂ss2+keω̂s+ω̂2

qv̂g(s)vg(s)

= keω̂2

s2+keω̂s+ω̂2 .(12)

The detailed derivation of these transfer functions can be foundin [6] and [37]. In order to evaluate the performance of SOGI-OSG PLL, the settling time is given as

ts =9.2

keω̂.

D. Comparison of the PLLs

In order to find the most suitable solution for the single-phasegrid-connected PV system in LVRT operation, the aforemen-tioned synchronization methods are compared in faulty-gridcases by simulations and experiments with the parametersshown in Table I. The experimental setup is shown in Fig. 12.This system consists of two Delta dc sources connected inseries, forming the nominal dc voltage Vdc = 400 V, and a


TABLE ISIMULATION AND EXPERIMENTAL PARAMETERS

Fig. 12. Experimental setup.

three-phase Danfoss 5-kW VLT inverter which is configured asa single-phase system. An LC filter is used in this arrangement,and it is connected to the grid through a three-phase transformerwith the leakage inductance of LT = 4 mH. The nominal gridparameters and other parameters are shown in Table I.

The results shown in Figs. 13 and 14 are obtained whenthe grid has a 0.45 p.u. voltage sag by switching the resistorsRs and RL. More comparisons of these PLLs by simulationsin terms of the settling time and the overshoot of frequencyare provided in Table II where different changes are done likefrequency jump and phase jump, and it can be used to selectappropriate methods for different applications.

As it can be seen in the results, the performances of thesePLL methods are not very good during the voltage sag. TheT/4 delay method can follow the amplitude change quickly(a quarter of the grid nominal period approximately), whileit cannot be a good synchronization technique when the gridis subjected to frequency variations. Although the main meritof an EPLL is that it can estimate both the amplitude andthe frequency of the input voltage without doubling the inputfrequency oscillations, this kind of PLL method presents a slowtransient variation as shown in Figs. 13 and 14. This variationdemonstrates how the AF minimizes the objective function.With respect to the control of single-phase systems, the EPLLmethod is not suitable for calculating the active and reactivepowers because it is not based on the OSG concept, but it can beused to control the instantaneous power [34]. The SOGI-OSGPLL can track the input voltage with better performance com-pared to T/4 delay PLL and EPLL, particularly when the grid

presents a frequency variation/jump as shown in Table II. It canbe concluded that, together with a fast detection unit, the SOGI-OSG PLL is the best candidate for single-phase applications.Thus, in this paper, this synchronization method is selected.

V. SYSTEM RIDE-THROUGH OPERATION

A simple case is examined by simulation under the voltagesag in order to give a basic demonstration about single-phasesystems under grid faults and also validated experimentally.

Referring to Figs. 4 and 12, a PR controller with harmoniccompensation is used as the current controller, and based onthe comparison in Section IV, the SOGI-OSG PLL is adoptedto detect the grid fault and to synchronize with the grid. Theparameters of the PI controller for active power are Kpp = 1.5and Kpi = 52, while for reactive power, they are Kqp = 1and Kqi = 50. The rated power is set to be 1 kW. The otherparameters for the experiments are shown in Table I. The resultsare shown in Figs. 15 and 16.

A 0.45 p.u. voltage sag is generated by switching the resistorsRs and RL as shown in Fig. 12. During the fault, the systemis controlled to limit the active power output without trippingthe current protection. Thus, the reliability of the PV inverter isimproved. Practically, the active power could be controlled byregulating the maximum power point. In the LVRT operationmode, the reactive power is injected into the utility grid untilthe grid voltage recovers to 0.9 p.u., as it is required in thegrid codes. From Figs. 15 and 16, it is concluded that thesingle-phase system can provide reactive power according tothe depth of the voltage sag in such a way to support the gridand protect the customers’ equipment, and it can do it fast. Afterthe clearance of the voltage fault, the grid current and the outputactive power go back to their normal values.

Since the SOGI-OSG PLL is also used to detect the voltagesag, the transient behavior is not good as it is shown in Fig. 16.Thus, it is necessary to develop a specific fast sag detectionalgorithm in order to guarantee a better performance of single-phase PV systems under grid faults.

VI. CONCLUSION

This paper presents the future requirements for single-phasegrid-connected PV systems at a high penetration level undergrid faults. It can be concluded that the future grid-connectedPV systems will be more active and more “smart,” which meansthat the future grid-connected PV systems should have someancillary functionalities as the conventional power plants do inthe presence of an abnormal grid condition.

Different control strategies of such kind of single-phase PVsystems under grid faults are discussed, and it is concluded thatthe control possibilities play an important role in single-phaseapplications since they are responsible not only for the powerquality and protection issues but also for the upcoming ancillaryrequirements. It can also be concluded that the single-phasePV inverters are ready to provide grid support considering ahigh-level penetration. Selected detection and synchronizationtechniques are also compared in the case of grid fault condi-tions. The comparison demonstrates that the SOGI-OSG-based


Fig. 13. Comparison of the three selected PLLs under a grid voltage sag (0.45 p.u.) by simulations: (1) EPLL, (2) SOGI-based PLL, and (3) T/4 delay-basedPLL. (a) Grid voltage. (b) Amplitude of the grid voltage. (c) Grid frequency. (d) Phase of the grid voltage.

Fig. 14. Experimental results of the three selected PLLs under a grid voltage sag (0.45 p.u.): (1) EPLL, (2) SOGI-based PLL, and (3) T/4 delay-based PLL,[t = 40 ms/div]. (a) Grid voltage (100 V/div). (b) Amplitude (30 V/div). (c) Frequency (2 Hz/div). (d) Phase [2π rad/(5 divs)].

PLL technique might be the promising candidate for single-phase systems under grid faults. Furthermore, another adaptive-filtering-based PLL (EPLL) shows also a good performanceunder voltage sag, but it has transient variations. However, theconcept of EPLL leads to the possibility of direct instantaneouspower control for single-phase systems.

A single-phase case is studied and tested experimentallyat the end of this paper in order to demonstrate the overallsystem performance under grid faulty conditions, and it showssatisfactory performance.

TABLE IICOMPARISON OF THE PLLS


Fig. 15. Simulation results of a single-phase grid-connected system under voltage sag (0.45 p.u.) using PR controller and SOGI-OSG PLL synchronizationmethod. (a) Grid voltage and current. (b) Active and reactive powers. (c) Transient grid voltage and current before a voltage sag. (d) Transient grid voltage andcurrent after a voltage sag.

Fig. 16. Experimental results of a single-phase grid-connected system under voltage sag (0.45 p.u.) using PR controller and SOGI-OSG PLL synchronizationmethod, [t = 40 ms/div]. (a) Grid voltage [vg : 100 V/div] and current [ig : 5 A/div]. (b) Active power [P : 500 W/div] and reactive power [Q : 500 Var/div].(c) Transient response of grid voltage [vg : 100 V/div] and current [ig : 5 A/div] before a voltage sag. (d) Transient response of grid voltage [vg : 100 V/div] andcurrent [ig : 5 A/div] after a voltage sag.

ACKNOWLEDGMENT

The authors would like to thank Dr. H. Wang from theDepartment of Energy Technology, Aalborg University, for hisvaluable advice and the discussions.

REFERENCES

[1] EPIA-Market Report, 2011, EPIA, Brussels, Belgium, Tech. Rep. [On-line]. Available: http://www.epia.org/

[2] M. Braun, T. Stetz, R. Brundlinger, C. Mayr, K. Ogimoto, H. Hatta,H. Kobayashi, B. Kroposki, B. Mather, M. Coddington, K. Lynn,G. Graditi, A. Woyte, and I. MacGill, “Is the distribution grid ready toaccept large-scale photovoltaic deployment? State of the art, progress, andfuture prospects,” Progr. Photovoltaics, Res. Appl., vol. 20, no. 6, pp. 681–697, Sep. 2012.

[3] IEEE Application Guide for IEEE Std 1547, IEEE Standard for Inter-connecting Distributed Resources with Electric Power Systems, IEEE Std1547. 2-2008, 2009.

[4] F. Iov, A. D. Hansen, P. E. Soerensen, and N. A. Cutululis, “Mappingof grid faults and grid codes,” Risø Nat. Lab., Tech. Univ. Denmark,Roskilde, Denmark, Tech. Rep., 2007.


[5] Grid Code–High and Extra High Voltage, E. ON GmbH, Bayreuth,Germany, 2006.

[6] R. Teodorescu, M. Liserre, and P. Rodriguez, Grid Converters for Photo-voltaic and Wind Power Systems. Hoboken, NJ, USA: Wiley, 2011.

[7] F. Blaabjerg, M. Liserre, and K. Ma, “Power electronics converters forwind turbine,” IEEE Trans. Ind. Appl., vol. 48, no. 2, pp. 708–719,Mar./Apr. 2012.

[8] F. Blaabjerg, R. Teodorescu, M. Liserre, and A. V. Timbus, “Overviewof control and grid synchronization for distributed power generationsystems,” IEEE Trans. Ind. Electron., vol. 53, no. 5, pp. 1398–1409,Oct. 2006.

[9] S. B. Kjaer, J. K. Pedersen, and F. Blaabjerg, “A review of single-phasegrid-connected inverters for photovoltaic modules,” IEEE Trans. Ind.Appl., vol. 41, no. 5, pp. 1292–1306, Sept./Oct. 2005.

[10] E. J. Coster, J. M. A. Myrzik, B. Kruimer, and W. L. Kling, “Integrationissues of distributed generation in distribution grids,” Proc. IEEE, vol. 99,no. 1, pp. 28–39, Jan. 2011.

[11] G. M. S. Islam, A. Al-Durra, S. M. Muyeen, and J. Tamura, “Low voltageride through capability enhancement of grid connected large scale photo-voltaic system,” in Proc. IEEE IECON, Nov. 2011, pp. 884–889.

[12] M. Braun, G. Arnold, and H. Laukamp, “Plugging into the Zeitgeist,”IEEE Power Energy Mag., vol. 7, no. 3, pp. 63–76, May/Jun. 2009.

[13] H. Alatrash, R. A. Amarin, and L. Cheung, “Enabling large-scale PVintegration into the grid,” in Proc. IEEE Green Technol. Conf., Apr. 19/20,2012, pp. 1–6.

[14] R. Hudson and G. Heilscher, “PV grid integration—System managementissues and utility concerns,” Energy Procedia, vol. 25, pp. 82–92, 2012.

[15] M. H. Coddington, B. D. Kroposki, and T. S. Basso, “Evaluating fu-ture standards and codes with a focus on high penetration photovoltaic(HPPV) system deployment,” in Proc. IEEE PVSC, Jun. 20–25, 2010,pp. 544–549.

[16] CEI 0-21: Reference Technical Rules for Connecting Users to the Activeand Passive LV Distribution Companies of Electricity, Comitato Elet-trotecnico Italiano, 2012.

[17] C. H. Benz, W. Franke, and F. W. Fuchs, “Low voltage ride throughcapability of a 5 kW grid-tied solar inverter,” in Proc. EPE/PEMC,Sep. 2010, pp. T12-13–T12-20.

[18] K. Fujii, N. Kanao, T. Yamada, and Y. Okuma, “Fault ride through capa-bility for solar inverters,” in Proc. EPE Appl., 2011, pp. 1–9.

[19] H. Kobayashi, “Fault ride through requirements and measures of dis-tributed PV systems in Japan,” in Proc. IEEE-PES Gen. Meeting,Jul. 22–26, 2012, pp. 1–6.

[20] A. Marinopoulos, F. Papandrea, M. Reza, S. Norrga, F. Spertino, andR. Napoli, “Grid integration aspects of large solar PV installations: LVRTcapability and reactive power/voltage support requirements,” in Proc.IEEE Trondheim PowerTech, Jun. 2011, pp. 1–8.

[21] C. Photong, C. Klumpner, and P. Wheeler, “A current source inverter withseries connected ac capacitors for photovoltaic application with grid faultride through capability,” in Proc. IEEE IECON, Nov. 2009, pp. 390–396.

[22] M. Saitou and T. Shimizu, “Generalized theory of instantaneous activeand reactive powers in single-phase circuits based on Hilbert transform,”in Proc. IEEE PESC, 2002, vol. 3, pp. 1419–1424.

[23] M. Prodanovic, K. De Brabandere, J. Van Den Keybus, T. Green, andJ. Driesen, “Harmonic and reactive power compensation as ancillaryservices in inverter-based distributed generation,” IET Gen. Transmiss.Distrib., vol. 1, no. 3, pp. 432–438, May 2007.

[24] R. A. Mastromauro, M. Liserre, T. Kerekes, and A. Dell’Aquila, “Asingle-phase voltage-controlled grid-connected photovoltaic system withpower quality conditioner functionality,” IEEE Trans. Ind. Electron.,vol. 56, no. 11, pp. 4436–4444, Nov. 2009.

[25] J. C. Vasquez, R. A. Mastromauro, J. M. Guerrero, and M. Liserre, “Volt-age support provided by a droop-controlled multifunctional inverter,”IEEE Trans. Ind. Electron., vol. 56, no. 11, pp. 4510–4519, Nov. 2009.

[26] R. M. Santos Filho, P. F. Seixas, P. C. Cortizo, L. A. B. Torres, andA. F. Souza, “Comparison of three single-phase PLL algorithms for UPSapplications,” IEEE Trans. Ind. Electron., vol. 55, no. 8, pp. 2923–2932,Aug. 2008.

[27] A. Nicastri and A. Nagliero, “Comparison and evaluation of the PLLtechniques for the design of the grid-connected inverter systems,” in Proc.IEEE ISIE, Jul. 2010, pp. 3865–3870.

[28] A. Nagliero, R. A. Mastromauro, M. Liserre, and A. Dell’Aquila, “Mon-itoring and synchronization techniques for single-phase PV systems,” inProc. SPEEDAM, Jun. 2010, pp. 1404–1409.

[29] S. Golestan, M. Monfared, F. D. Freijedo, and J. M. Guerrero, “Design andtuning of a modified power-based PLL for single-phase grid-connectedpower conditioning systems,” IEEE Trans. Power Electron., vol. 27, no. 8,pp. 3639–3650, Aug. 2012.

[30] L. N. Arruda, S. M. Silva, and B. J. C. Filho, “PLL structures for utilityconnected systems,” in Conf. Rec. 36th IEEE IAS Annu. Meeting, 2001,vol. 4, pp. 2655–2660.

[31] M. K. Ghartemani, S. A. Khajehoddin, P. K. Jain, and A. Bakhshai,“Problems of startup and phase jumps in PLL systems,” IEEE Trans.Power Electron., vol. 27, no. 4, pp. 1830–1838, Apr. 2012.

[32] P. Rodriguez, A. Luna, R. S. Munoz-Aguilar, I. Etxeberria-Otadui,R. Teodorescu, and F. Blaabjerg, “A stationary reference frame gridsynchronization system for three-phase grid-connected power convertersunder adverse grid conditions,” IEEE Trans. Power Electron., vol. 27,no. 1, pp. 99–112, Jan. 2012.

[33] B. Bahrani, A. Rufer, S. Kenzelmann, and L. A. C. Lopes, “Vec-tor control of single-phase voltage-source converters based on fictive-axis emulation,” IEEE Trans. Ind. Appl., vol. 47, no. 2, pp. 831–840,Mar./Apr. 2011.

[34] S. A. Khajehoddin, M. Karimi Ghartemani, A. Bakhshai, and P. Jain, “Apower control method with simple structure and fast dynamic response forsingle-phase grid-connected DG systems,” IEEE Trans. Power Electron.,vol. 28, no. 1, pp. 221–233, Jan. 2013.

[35] M. Karimi-Ghartemani and M. R. Iravani, “A new phase-locked loop(PLL) system,” in Proc. IEEE MWSCAS, 2001, vol. 1, pp. 421–424.

[36] P. S. R. Diniz, Adaptive Filtering: Algorithms and Practical Implementa-tion. New York, NY, USA: Springer-Verlag, 2010.

[37] M. Ciobotaru, R. Teodorescu, and F. Blaabjerg, “A new single-phase PLLstructure based on second order generalized integrator,” in Proc. IEEEPESC, Jun. 2006, pp. 1–6.

[38] S. Douglas and T.-Y. Meng, “A nonlinear error adaptive notch filterfor separating two sinusoidal signals,” in Proc. ASILOMAR, Nov. 1991,vol. 2, pp. 673–677.

[39] X. Yuan, W. Merk, H. Stemmler, and J. Allmeling, “Stationary-framegeneralized integrators for current control of active power filters with zerosteady-state error for current harmonics of concern under unbalanced anddistorted operating conditions,” IEEE Trans. Ind. Appl., vol. 38, no. 2,pp. 523–532, Mar./Apr. 2002.

[40] A. Camacho, M. Castilla, J. Miret, J. C Vasquez, and E. Alarcon-Gallo,“Flexible voltage support control for three phase distributed generationinverters under grid fault,” IEEE Trans. Ind. Electron., vol. 60, no. 4,pp. 1429–1441, Apr. 2013.

[41] S. M. Silva, B. M. Lopes, B. J. C. Filho, R. P. Campana, andW. C. Bosventura, “Performance evaluation of PLL algorithms for single-phase grid-connected systems,” in Conf. Rec. IEEE 39th IAS Annu. Meet-ing, Oct. 2004, vol. 4, pp. 2259–2263.

[42] M. Z. C. Wanik, A. Mohamed, A. F. A. Kadir, and I. Erlich, “Low voltageride through capability of fuel cell generation system connected to lowvoltage network,” in Proc. IEEE CET , Jun. 2011, pp. 369–373.

[43] M. Bollen, Understanding Power Quality Problems: Voltage Sags andInterruptions. Piscataway, NJ, USA: IEEE Press, 2000.

[44] Y. Yang and F. Blaabjerg, “Synchronization in single-phase grid-connected photovoltaic systems under grid faults,” in Proc. IEEE PEDG,Aalborg, Denmark, Jun. 25–28, 2012, pp. 476–482.

[45] J. Miret, M. Castilla, A. Camacho, L. Garcia de Vicuña, and J. Matas,“Control scheme for photovoltaic three-phase inverters to minimize peakcurrents during unbalanced grid-voltage sags,” IEEE Trans. Power Elec-tron., vol. 27, no. 10, pp. 4262–4271, Oct. 2012.

[46] R. Yan and T. K. Saha, “Investigation of voltage stability for residentialcustomers due to high photovoltaic penetrations,” IEEE Trans. PowerSyst., vol. 27, no. 2, pp. 651–662, May 2012.

Yongheng Yang (S’12) received the B.Eng. in elec-trical engineering and automation from Northwest-ern Polytechnical University, Xi’an, China, in 2009.During 2009–2011, he was enrolled in a master-doctoral program in the School of Electrical En-gineering at Southeast University, Nanjing, China.During that period, he was involved in the modelingand control of single-phase grid-connected photo-voltaic (PV) systems. From March to May 2013,he was a Visiting Scholar in the Department ofElectrical and Computer Engineering at Texas A&M

University, College Station, TX, USA. He is currently working toward the Ph.D.degree in the Department of Energy Technology at Aalborg University, AalborgEast, Denmark.

His research interests include grid detection, synchronization, and control ofsingle-phase photovoltaic systems in different operation modes, and reliabilityfor next-generation PV inverters.


Frede Blaabjerg (S’86–M’88–SM’97–F’03) re-ceived the Ph.D. degree from Aalborg University,Aalborg East, Denmark, in 1995.

He was employed at ABB-Scandia, Randers,Denmark, from 1987 to 1988. During 1988–1992, hewas a Ph.D. student at Aalborg University, Aalborg,Denmark, and became an Assistant Professor in1992, an Associate Professor in 1996, and a FullProfessor of power electronics and drives in 1998.He was a part-time Research Leader at the ResearchCenter Risoe for wind turbines. In 2006–2010, he

was the Dean of the Faculty of Engineering, Science and Medicine and becamea Visiting Professor at Zhejiang University, China, in 2009. His research areasare in power electronics and applications such as wind turbines, photovoltaicsystems, and adjustable-speed drives.

Dr. Blaabjerg has been the Editor-in-Chief of the IEEE TRANSACTIONS

ON POWER ELECTRONICS since 2006. He was a Distinguished Lecturer forthe IEEE Power Electronics Society in 2005–2007 and for the IEEE IndustryApplications Society from 2010 to 2011. He was the recipient of the 1995Angelos Award for his contributions to modulation technique and the AnnualTeacher Prize from Aalborg University. In 1998, he received the OutstandingYoung Power Electronics Engineer Award from the IEEE Power ElectronicsSociety. He has received ten IEEE prize paper awards and another prize paperaward at PELINCEC Poland 2005. He received the IEEE PELS DistinguishedService Award in 2009 and the EPE-PEMC 2010 Council award.

Zhixiang Zou (S’12) received the B.Eng. degree inelectrical engineering and automation from South-east University, Nanjing, China, in 2007, where heis currently working toward the Ph.D. degree.

From 2007 to 2008, he was employed at the StateGrid Electric Power Research Institute, Nanjing. Hisresearch interests include modeling and control ofgrid converters, distributed generation, and powerquality.

Documents

Benchmarking of Grid Fault Modes in Single-Phase Grid-Connected Photovoltaic Systems