Islanding detection in smart grids

Adrian Timbus, Alexandre Oudalov and Carl N.M. Ho

ABB Switzerland Ltd., Corporate Research Segelhofstrasse 1K, P.O. Box, 5405 Baden-D ttwil, Switzerland, a Email: adrian.timbus@ch.abb.com, alexandre.oudalov@ch.abb.com

AbstractAll distributed generators (DG), especially those connected to low voltage distribution grids are required to detect islanding conditions. The methods for detecting islanding are classied in three main categories: passive, active and communication based. Passive methods are based on grid monitoring, are easy to implement but have a large non-detection zone in the case local generation meets the load demand. Active methods, which are commonly used today, may reduce the non-detection zone but in the case of large amount of DGs installed, power quality problems are foreseen. The communication based methods are seldom used today mainly because of high cost of communication. However, if the islanding detection schemes can use the communication infrastructure to be deployed for smart grids, e.g. for metering, feeder automation, etc. the communication based methods will become cost competitive with the active methods without their weaknesses. This paper focuses on communication based methods and analyzes the inuence of bidirectional power ow on the existing methods. The paper also proposes a new method to be used in a smart grid infrastructure. Index Termsislanding detection, smart grid, distributed generation

I. I NTRODUCTION Lately, distributed generation based on renewable energy resources registered an exponential growth facilitated by the policy makers, global concerns about CO2 emissions, energy shortage, interest in clean energy production, etc. Energy suppliers using power plants based on conventional fuels (coal, gas, etc.) are also investing in the renewable sector such as wind turbines and photovoltaic systems. This leads to a fast and continuous development of technology in this specic area, bringing these systems at a point where their reliability and competence is not anymore discussed. However, for connecting such systems to the utility grid, several conditions have to be met. These conditions are normally published by standardizing institutions such as IEC and IEEE but also by local (country) regulating authorities. A very important requirement which is mandatory to distributed generators is their ability to detect islanding conditions. Islanding refers to the condition of a distributed generator (DG) continuing to power a part of the grid even though power from electric utility is no longer present. Unintentional islanding can be dangerous to utility workers, who may not realize that the particular part of the network is still powered even though there is no power from the main grid. Also, unintentional and non-controllable islanding can damage

customer equipment, especially in situations of re-closing into an island. For that reason, distributed generators must detect islanding and immediately disconnect. The probability of having islanding conditions is very small because, when the grid power is lost, the generated power (both active and reactive) of distributed generators has to match almost perfectly the power consumed by the loads including losses, otherwise the under/over voltage and under/over frequency relays of generators would cease power generation. Lately there has been a lot of interests in micro-grids, i.e. distribution grids that can operate in controllable, intentional islanding conditions, decoupled from the main grid. In case of such grids, islanding detection is still important in order to switch the control modes of distributed generators from power injection to voltage and frequency control during disconnection and opposite during reconnection to the main grid. This paper gives an overview of the available methods for detecting islanding, without going to much into technicalities for each individual method. The contribution of the paper is an assessment of smart grids inuence on islanding detection methods. The overview of existing methods is detailed in Section II followed by a brief description of the requirements from the standardization organizations and local (country level) demands for islanding detection in Section III. Section IV introduces the smart grid concept and assesses some possible changes to islanding detection methodologies given the availability of a communication infrastructure in smart grids. II. OVERVIEW OF METHODS Booming of distributed requirements for islanding research and development islanding conditions. These main categories:

generators industry combined to detection has led to an intensive of methods for identication of methods can be classied in three

passive methods active methods communication based methods

A comprehensive report on methods and their classication has been realized in [1] and [2]. A brief description of the methods and their classication is given below.

978-1-4244-5287-3/10/$26.00 2010 IEEE

3631

A. Passive methods Passive methods are based on monitoring of grid variables by a dedicated algorithm residing in the control of distributed generator or outside in a dedicated device. Most passive methods are looking for abnormal changes in frequency, voltage or phase angle but also in some particular harmonics or the total harmonic distortion (THD). If the monitoring algorithm detects large or sudden changes of these variables at the point of connection with the utility grid, the inverter will trip. The most common passive methods are [1][3]: over/under voltage monitors whether or not the grid voltage goes out of the limits established by the relevant standards (see Fig. 1) over/under frequency monitors whether or not the grid frequency goes out of the limits imposed by the relevant standards (see Fig. 1) monitoring rate of change of frequency (ROCOF) and voltage (ROCOV) phase monitoring monitors fast jumps of grid voltage phase voltage harmonic monitors selective (3rd, 5th, etc.) or total harmonic distortion (THD) of grid voltage [4], [5] Passive methods are quite effective in majority of situations that may occur in the grid; however their non-detection zone (NDZ) does not cover situations when the power absorbed by the load matches almost perfectly the power generated by DG. In such case, the variations in voltage, frequency or phase angle are lower than specied in the standard because the network remains balanced even though the connection with the main grid has been lost. Therefore, the DG would not trip even though an island has been formed. One can use a combination of passive methods and use multi-criteria decision making. In [6] the islanding detection algorithm uses passive methods including frequency and voltage monitoring and the most widely recognized methods like the Rate of Change of Frequency (ROCOF), Voltage Vector Shift (VVS), and Voltage Unbalance. However, passive methods are generally considered as insufcient anti-islanding protection [1]. B. Active methods Active methods appeared as a necessity to minimize the non-detection zone of islanding detection methods in conditions when generation matches load. Normally, active methods inject a small disturbance to the utility grid and based on the grid response decide whether or not an island has formed. Disturbances in terms of shifts from normal values to grid voltage magnitude, frequency or phase angle can be added by the DG and, in case of grid connected situation, these disturbances should be corrected by the grid through the voltage and frequency control. However, if the voltage magnitude, frequency or phase angle follow the shift introduced by the DG, it is most likely that the grid has been disconnected, hence an island has been formed. The most common active methods are using:

positive feedback inside the DG control the controller tries to alter grid variables such as frequency, phase or voltage magnitude in order to perform: a frequency jump or phase jump the DG deliberately alters the frequency or phase of the injected current in order to change the frequency or phase of the voltage. If grid frequency follows the inverter current, an island has formed and the PV inverter should disconnect [1], [7], [8] a frequency bias similar to above [1] Sandia Frequency Shift similar to frequency jump but elaborated by Sandia National Laboratories [9] [11] voltage positive feedback, Sandia Voltage Shift tries to alter the voltage magnitude at the point of common coupling. If the grid follows the changes generated by the DG, the grid voltage will go out of imposed operating ranges and consequently the inverter would trip [1], [10][12] injection of harmonics via DG non-characteristic harmonics are injected by the DG and the grid response is registered [13], [14]. The method is also called detection of impedance at certain frequency. In [13][17] injection of non-characteristic inter-harmonic current is used to derive the grid impedance at that particular frequency. The value of inter-harmonic has to be chosen close to the fundamental grid frequency in order to assume that the identied value of grid impedance can be approximated for fundamental frequency.

impedance detection active method which has been promoted by the requirements in the German standard [18]. In [19][21] a current spike is periodically injected at the point of common coupling by a grid tied power converter. Based on the voltage response to this disturbance, the grid impedance value is determined using Fourier transform. The inuence of non-linear loads connected close to the point of common coupling (PCC) is also addressed and as a consequence additional signal processing method is necessary in order to obtain accurate results. In [22], the phase angle signal used to generate the reference for current controller is slightly altered to be able to estimate the grid impedance based on grid reaction to the generated current. A high frequency signal (600 Hz) is injected at zero crossing in [23] for determining the value of grid impedance. Active and reactive power oscillations are used in [24][26] to identify the value of grid impedance. Although, active methods give better islanding identication, they also distort the delivered power in order to detect islanding conditions. Disturbances in power grid are not suitable when a signicant number of DGs are connected on the same feeder. As discussed in [17], because the injected inter-harmonic generated by the DG is unique in the utility network, this particular method

3632

TABLE I: Comparison of islanding detection methods

Method type Passive Active Strengths - grid friendly - easy and cheap to implement - low NDZ - some easy to implement Weaknesses - NDZ larger compared to others - can create power quality problems - can lead to nuisance trip - some difcult to implement - possible interaction between converters in the same grid - expensive to implement - need communication infrastructure - need involvement of utility Grid friendly yes - suitable for a nite number of generators

Communication based

- reliable - some easy to implement - theoretically no NDZ

yes

would allow the connection of more than one similar generation units on the same grid because synchronization in respect to the inter-harmonic injection is possible [15], [17]. However, only a nite number of inverters can be connected on the same feeder and this number is directly depending on the standard demands regarding disconnection time and the number of injections necessary to obtain accurate impedance identication. Large penetration level of inverters using active methods not only decreases the quality of power in the grid but also increases the nondetection zone of all active methods as discussed in [1]. C. Communication based methods Another category of methods for detecting islanding is based on communication between DG and the utility grid. In [1], three main methods using communication are detailed:

D. Comparison of islading methods Table I draws on main features of main categories of islanding methods. Passive methods are the basic protection package of every distributed generator connected to utility grid. As of today, active methods are preferred due to their low non-detection zone. However, one of the main drawbacks which may contribute to a shift from using active methods is their negative contribution to the quality of power in the grid. Future availability of a communication network for the power grid combined with more interest from utilities to monitor the assets in the grid may facilitate a move towards the use of communication based methods for islanding detection. III. I SLANDING REGULATIONS Global standardization authorities such as IEC and IEEE are the main divers for standardizing requirements and testing conditions for islanding detection. However, there are also country specic regulations which may differ from both IEC or IEEE approaches and which complicate the development of a general solution for the global market. Japan, Germany, Austria and more recently Spain and Italy are known for applying different requirements for connecting DGs to the utility grid. A. IEC standards The IEC 61727 standard [33] denes the characteristics of utility interface for photovoltaic systems. The standard addresses power quality issues (harmonics, icker), normal voltage and frequency operation range, and safety related matters. Regarding frequency deviations, IEC 61727 suggests that for frequencies outside the range of 1Hz, the PV system should cease energizing the network within 0.2 s. With regard to islanding detection, the only mention of this standard is that the PV system needs to disconnect from utility network within 2 seconds after islanding conditions occurred. IEC 62116 [34] denes the test procedure for islanding prevention measures for PV systems. The test system, as well as testing conditions, is well detailed and the PV system is tested in three cases: at maximum output power having the input voltage range set larger than 90% at 5060% of the maximum power of the inverter with the input voltage range of 50% 10%

Power line used as carrier for communication between the PV inverter and utility grid [27][29]. A continuous signal is transmitted by utility network via the power line. A receiver is necessary to be connected to the DG for detecting the loss of this signal and hence determining islanding conditions. Signal produced by disconnect. This method assumes that the utility recloser is equipped with a transmitter which communicates with DG when opens [1] SCADA based method uses placement of voltage sensors at the location where DG is connected and integration of those sensors in the SCADA system for monitoring and alarming the PV system to disconnect in case of islanding [1], [30]. With an increasing number of DGs connected to the grid, real time monitoring of voltage for each generator in distribution grid can be a cumbersome process.

Communication based methods necessitate the involvement of utility in implementation of islanding detection schemes, making them less favorite for practical implementation. Because communication adds also costs to both DG and the grid infrastructure, these methods are not commonly used today. Using communication standards such as IEC 61850 [31] and its extensions for distributed generators [32], a SCADA/DMS application can be implemented to communicate to all DGs in a distribution grid and inform them about islanding conditions.

3633

at 2533% (or minimum allowed) of maximum power with less than 10% input voltage range All testing steps are described in detail in IEC 62116.

B. IEEE standards Through IEEE 1547 standard [35], IEEE denes the requirements for connection of distributed resources and utility grid. The specied time for ceasing energy production in case of islanding conditions are the same as dened in IEC standard, i.e. 2 seconds. IEEE 929 [36] denes islanding requirements for PV inverters. The standard describes the procedure for testing the capability of PV inverters to detect islanding situations. An RLC circuit having a quality factor Q=2.5 is used to balance the generated power of PV system. Once the power is balanced, the utility circuit breaker is open and the test is initiated and time to trip is recoded. C. UL standards Many PV inverter manufacturers advertise their products as UL certied. The Underwriters Laboratories Inc. (UL) species the requirements for inverters, converters, controllers and interconnection system equipment for use with distributed energy resources in the UL 1741 [37] standard. Investigating the standard, it has been found that the part related to islanding detection and operating ranges for voltage and frequency has been superseded from the latest version of the standard (2005) and reference to IEEE 1547 standard [35] is made. D. German standard The DIN VDE 0126 standard applied in Germany was one of the most demanding standards for islanding detection. The DIN VDE 0126 imposed the methodology for detecting islanding conditions to impedance measurement. Consequently, all PV systems installed in Germany needed to have an impedance detection algorithm in order to be certied. With the new version of the standard (DIN VDE 0126-1-1) [18], the detection of islanding conditions could be done using one the following three methods:

impedance measurement detection of 1 grid impedance jump within 5 seconds resonant circuit detection of balanced islanding conditions (using a resonant circuit with a quality factor Q2) within 5 seconds (similar to [36]) three phase voltage monitoring passive monitoring of external conductor voltage causes disconnection in the event of voltage or phase failures

E. Spanish standard The Spanish requirements on the connection of photovoltaic installation onto low voltage network are published in the royal decree 1663/2000 [38]. The protection issues are addressed in Article 11 and over/under voltage/frequency protection only are mentioned. The values of over/under voltage are 1.1 and 0.85 the nominal grid voltage, while 1 Hz is allowed for frequency variations. There is no specication for islanding detection in this decree. F. Italian standards The rules for connecting distributed generators in low voltage grids are published by Italian grid operator ENEL [39] and therefore apply to interconnections to the low voltage grids owned by ENEL and considered best practice for other grids in Italy. Similar to Spain, the requirements from ENEL address over/under frequency/voltage protections and do not mention balanced islanding conditions. IV. S MART GRIDS AND THEIR INFLUENCE ON ISLANDING A. What are smart grids? The electrical power system was built up over more than 100 years. It is now one of the most effective components of the infrastructure on which modern society depends. It delivers electrical energy to industry, commercial and residential consumers, meeting ever-growing demand. Most of todays generation capacity relies on fossil fuels and contributes signicantly to the increase of carbon dioxide in the worlds atmosphere, with negative consequences for the climate and society in general. To satisfy both the increasing demand for power and the need to reduce carbon dioxide emissions, we need an electric system that can handle these challenges in a sustainable,

1.35 Vn 0.05 s 1.1 Vn 2s Vn

1.1 Vn 1.2 Vn 1s 0.16 s

0.2 s

Continuous Operation

Vn

0.2 s

0.16 s

Continuous Operation

0.16 s

0.85 Vn 2s 0.5 Vn 0.1 s

0.88 Vn 2s 0.5 Vn 0.16 s

f n - 1Hz

fn

f n + 1Hz

f n - 0.7Hz

fn

f n + 0.5Hz

(a) IEC requirements

(b) IEEE 1547 requirement (a) Todays grid (b) Smart grid

Fig. 1: Illustration of requirements for voltage and frequency operation limits by IEC and IEEE.

Fig. 2: Illustration of topologies for todays and smart grid.

3634

TABLE II: Differences between todays grid and the smart grid.
Area Communications Metering Operation Generation Power ow control Restoration following disturbance System topology Reliability Todays grid None or unidirectional Electromechanical Manual equipment checks, time based maintenance Centralized Limited Manual Radial; generally top-down power ow Prone to failures and cascading outages; essentially reactive Smart grid Bidirectional Digital (enabling real-time pricing and net metering) Remote monitoring, predictive, event driven maintenance Centralized and distributed Comprehensive, automated Self-healing Meshed; multiple power ow pathways Automated, pro-active protection;prevents outages before they start

reliable and economic way. This network will utilize the same basic electric infrastructure we know today, but will also draw on advanced monitoring, control and communications technology that is presently only beginning to be applied. The result will be a grid that is largely automated, applying greater intelligence to operate, monitor and even heal itself. This smart grid will be more exible, more reliable and better able to serve the needs of a digital economy. Table II points out the main differences between todays grid and the smart grid. The topology area in the table, hints at what is perhaps the most fundamental shift that a fully realized smart grid will require. Todays power systems are designed to support large generation plants that serve faraway consumers via a transmission and distribution system that is essentially one-way. But the grid of the future will necessarily be a two-way system where power generated by a multitude of small, distributed sources in addition to large plants ows across a grid based on a network rather than a hierarchical structure. B. Smart grid inuences on islanding With regard to islanding detection, the availability of bidirectional communication foreseen in the smart grid infrastructure combined with the power quality issues brought by active methods, especially when a large number of DGs are installed in the grid, will favor the communication based methods for islanding detection. One issue which could hold down the growth of these methods is the conservative behavior of utilities and not adopt a centralized protection scheme in case of islanding. However, another particular propriety of smart grids, i.e. multiple ow paths could have a negative inuence on the communication based methods listed above. This paper provides some insights on how islanding is affected by this. 1) Inuence of bidirectional power ow on power line carrier: In this case, a dedicated communication device periodically injects a high frequency signal on the secondary side of the transformer station 1 and 2, in front of switching equipments SE1 and SE2. In case the breakers SE3, SE4 and SE5 are closed and substation 1 (SS1) and SS2 operate in loop conguration, two signals may interfere and cause nuisance tripping of some generators. The two signals can be set with different frequencies but then it has to be ensured that all DGs located on the feeder listen to both frequencies.

2) Inuence of bidirectional power ow on signal produced by disconnect: The principle of this method is fairly simple, once a switching equipment (SE) opens, it sends out a trip signal to all generators in the network. Providing the existence of a communication infrastructure between the switch and all generators (radio, GPRS, wired), this could be easily realized for todays radial distribution networks. However, when multiple paths for power ow exist, the switch needs to send out the trip signal in either direction. This can cause undesired disconnection of generators from a properly working part of the network. As an example, let us consider the situation illustrated in Fig. 3 where two substations (SS) are interconnected through the switches SE4 and SE5. If SE1 is open and SE2 is closed, the power is supplied from SS2 and the local generators. In case SE5 opens, it should send the trip signal only to generators located in the direction of SE4 and downstream. Unless SE5 communicates with SE1 and SE2 and employs an algorithm for establishing the network topology, it will be very difcult for it to decide to which generators to send the trip signal (towards SE4 or towards SE2?). When having only a few switches involved and a fairly simple network topology, the decision algorithm could be implemented at the switch level, however in case of more complex network topology, computational communication time may become an issue. As a solution to this, the paper proposes an implementation of the islanding detection algorithm at the substation level. Here, a substation computer (which may have also other functions within a smart grid framework) can monitor the status of the switches along the feeder and the interconnecting lines with the neighboring substations. Also, the substation computer can get the network topology from the upper level e.g. distribution control center and hence can perform the logic for deciding to which DGs shall send the disconnect signal in case of islanding. This method distinguishes itself from the SCADA based method presented in the following section through the fact that is localized to a small part of the network and hence can operate in real time, meeting the requirements for disconnect within 2 seconds imposed by standards. Another novelty of this method is the fact that anti-islanding groups (AIG) could be dened. In case of radial feeders, these groups shall comprise all the distributed generators located between 2 consecutive switching devices. In case of meshed grid, AIG could be dened between the switches with direct

3635

SCADA/ DMS Communication line Power line Distribution grid

Substation Computer

Substation Computer

SE 1
Substation 1

SE 2
Substation 2

AIG 1 AIG 4

AIG 2 AIG 3

SE 3

SE 4

SE5

Fig. 3: Schematic of possible implementation of a hierarchical islanding detection method based on communication between DGs and a substation computer.

connection (without the possibility of interrupting between any combination of two switches, e.g. SE1-SE3, SE1-SE4, SE4-SE5). Fig. 3 illustrates the concept of AIG within a distribution grid. The main goal of the AIG is to speed up the processing time to detect islanding and this is due to the fact that distributed generators are structured in groups. Once a part of the feeder is disconnected due to opening of switches, the group of DGs connected to that part of the network (modeled as one AIG) receives the disconnect signal from the substation computer. 3) Inuence of bidirectional power ow on SCADA based methods: Along with the drawbacks of SCADA based methods mentioned in [2] another issue of this method is the time necessary to assess islanding. Although is expected that smart grids will provide bidirectional real-time communication, as of today the pulling rate of data into SCADA system is around 5 to 10 seconds. Even considering a very small processing time for islanding assessment, this is far behind the requested disconnection time imposed by regulations, i.e. 2 seconds. Consequently, in order for this method to become efcient and compliant with the standards, a much higher sampling rate will be necessary. Due to large size of distribution networks, this will stress the need for more space in the SCADA database and faster data processing. Moreover, due to large number of

customers of Distribution System Operators, the data sent to SCADA system is generally aggregated at secondary substation level, thus no concrete information exists about a particular customer, DG in this case. This situation my change with the installation of smart meters all over the distribution grid, however this will challenge the communication bandwidth and speed and data storage if all data is stored at control center level. An alternative is to decentralize the functionality of the control center among primary and secondary substations. V. C ONCLUSIONS This paper addresses the challenges imposed by smart grids to existing islanding detection methods. Since it is envisioned that bidirectional real time communication will be a key enabler for smart grids, particular focus is set on the islanding detection methods based on communication. The ndings of this paper shows that multi-path power ow imposes the most demanding challenges for such methods. The paper proposes an islanding detection method based on local monitoring of switching devices (per substation level) and use of antiislanding groups to select the right generators to trip when switching equipment opens. R EFERENCES
[1] W. Bower and tection methods M. for Ropp, Evaluation photovoltaic utility of islanding deinteractive power

3636

[2] [3]

[4]

[5] [6] [7]

[8] [9]

[10]

[11]

[12]

[13] [14] [15]

[16]

[17] [18] [19]

systems, IEA-PVPS, http://www.iea-pvps.org, Online report T509, March 2002. [Online]. Available: http://www.oja-services.nl/ieapvps/products/download/rep5 09.pdf , Evaluation of islanding detection methods for utility-interactive inverters in photovoltaic systems, Sandia National Laboratories, Tech. Rep., 2002. F. De Mango, M. Liserre, A. DellAquila, and A. Pigazo, Overview of anti-islanding algorithms for PV systems. Part I: Passive methods, in Proc. of Power Electronics and Motion Control Conference, 2006, pp. 18781883. S.-I. Jang and K.-H. Kim, An islanding detection method for distributed generations using voltage unbalance and total harmonic distortion of current, IEEE Trans. on Power Delivery, vol. 19, no. 2, pp. 745752, 2004. H. Zeineldin, M. I. Marei, E. F. El-Saadany, and M. M. A. Salama, Safe controlled islanding of inverter based distributed generation, in Proc. of PESC04, vol. 4, 2004, pp. 25152520 Vol.4. A. Burek and E. Rosolowski, Fuzzy logic based loss of grid protection algorithm, in International Universities Power Engineering Conference, 2005. C. Jeraputra and P. Enjeti, Development of a robust anti-islanding algorithm for utility interconnection of distributed fuel cell powered generation, IEEE Trans. on Power Electronics, vol. 19, no. 5, pp. 1163 1170, 2004. J. B. Jeong, H. J. Kim, S. H. Back, and K. S. Ahn, An improved method for anti-islanding by reactive power control, in Proc. of ICEMS05, vol. 2, 2005, pp. 965970. V. John, Z. Ye, and A. Kolwalkar, Investigation of anti-islanding protection of power converter based distributed generators using frequency domain analysis, IEEE Trans. on Power Electronics, vol. 19, no. 5, pp. 11771183, 2004. J. Stevens, R. Bonn, J. Ginn, and S. Gonzalez, Development and testing of an approach to anti-islanding in utility-interconnected photovoltaic systems, Sandia National Laboratories, Tech. Rep. SAND 2000-1939, 2000. G. A. Kern, R. H. Bonn, J. Ginn, and S. Gonzalez, Results of Sandia National Laboratories grid-tied inverter testing, in 2nd World Conference and Exhibition on Photovoltaic Solar Energy Conversion, 1998. Z. Ye, L. Li, L. Garces, C. Wang, R. Zhang, M. Dame, R. Walling, and N. Miller, A new family of active anti-islanding schemes based on DQ implementation for grid-connected inverters, in Proc. of PESC04, vol. 1, Aachen, Germany, 2004, pp. 235241. L. Asiminoaei, R. Teodorescu, F. Blaabjerg, and U. Borup, A new method of on-line grid impedance estimation for PV inverter, in Proc. of APEC 04, vol. 3, 2004, pp. 15271533 Vol.3. , A digital controlled PV-inverter with grid impedance estimation for ENS detection, IEEE Trans. on Power Electronics, vol. 20, no. 6, pp. 14801490, 2005. F. Bertling and S. Soter, A novel converter integrable impedance measuring method for islanding detection in grids with widespread use of decentral generation, in Proc. of Power Electronics, Electrical Drives, Automation and Motion, 2006, pp. 503507. A. Timbus, R. Teodorescu, F. Blaabjerg, and U. Borup, Online grid measurement and ENS detection for PV inverter running on highly inductive grid, IEEE Power Electronics Letters, vol. 2, no. 3, pp. 7782, 2004. , Online grid impedance measurement suitable for multiple PV inverters running in parallel, in Proc. of APEC 06, 2006, pp. 907911. DIN-VDE, Automatic disconnection device between a generator and the public low-voltage grid, DIN-VDE Std. 0126-1-1, 2005. B. Palethorpe, M. Sumner, and D. Thomas, Power system impedance measurement using a power electronic converter, in Proc. of Harmonics and Quality of Power, vol. 1, 2000, pp. 208213.

[20] , System impedance measurement for use with active lter control, in Proc. of Power Electronics and Variable Speed Drives, 2000, pp. 2428. [21] M. Sumner, B. Palethorpe, and D. Thomas, Impedance measurement for improved power quality-Part 2: a new technique for stand-alone active shunt lter control, IEEE Trans. on Power Delivery, vol. 19, no. 3, pp. 14571463, 2004. [22] M. Ciobotaru, V. Agelidis, and R. Teodorescu, Accurate and lessdisturbing active anti-islanding method based on PLL for grid-connected PV Inverters, in Proc. IEEE Power Electronics Specialists Conference PESC, 1519 June 2008, pp. 45694576. [23] M. Ciobotaru, R. Teodorescu, and F. Blaabjerg, On-line grid impedance estimation based on harmonic injection for grid-connected PV inverter, in Proc. IEEE International Symposium on Industrial Electronics ISIE, 47 June 2007, pp. 24372442. [24] M. Ciobotaru, R. Teodorescu, P. Rodriguez, A. Timbus, and F. Blaabjerg, Online grid impedance estimation for single-phase grid-connected systems using pq variations, in Proc. IEEE Power Electronics Specialists Conference, 2007, pp. 23062312. [25] A. Timbus, P. Rodriguez, R. Teodorescu, and M. Ciobotaru, Line impedance estimation using active and reactive power variations, in Proc. IEEE Power Electronics Specialists Conference, 2007, pp. 1273 1279. [26] A. Timbus, R. Teodorescu, and P. Rodriguez, Grid impedance identication based on active power variations and grid voltage control, in Conference Record of the 2007 IEEE 42nd IAS Annual Meeting Industry Applications Conference, 2007, pp. 949954. [27] M. E. Ropp, K. Aaker, J. Haigh, and N. Sabbah, Using power line carrier communications to prevent islanding [of pv power systems], in Conference Record of the Twenty-Eighth IEEE Photovoltaic Specialists Conference, 2000, pp. 16751678. [28] R. Benato, R. Caldon, and F. Cesena, Carrier signal-based protection to prevent dispersed generation islanding on MV systems, in CIRED, 2003. [29] O. Abarrategui, I. Zamora, D. Larruskain, and A. Iturregi, Power line carrier communications and its interest in the current power grid scenario, in ICREPQ, 2008. [Online]. Available: http://www.icrepq.com/icrepq-08/347-abarrategui.pdf [30] N. Strath, Islanding detection in power systems, Ph.D. dissertation, Lund University, 2005. [Online]. Available: http://www.iea.lth.se/publications/Theses/LTH-IEA-1051.pdf [31] IEC, IEC 61850 Communication networks and systems in substations, IEC Std., 2003. [32] , IEC 61850-7-420 Communications Systems for Distributed Energy Resources (DER), IEC Std., 2006. [33] , Photovoltaic (PV) systems - Characteristics of the utility interface, IEC Std. 61 727, 2004. [34] , Test procedure of islanding prevention measures for utilityinterconnected photovoltaic inverters, IEC Std. 62 116, 2008. [35] IEEE, IEEE standard for interconnecting distributed resources with electric power systems, IEEE Std. 1547, 2003. [36] , IEEE recommended practice for utility interface of photovoltaic (PV) systems, IEEE Std. 929, 2000. [37] UL, Inverters, Converters, Controllers and Interconnection System Equipment for Use With Distributed Energy Resources, UL Std. UL 1741, 1999. [38] CNE, Legislation development of the Spanish electric power act, 2002, vol. 4, Royal decree 1663/2000. [39] ENEL, DK 5940 - Criteria for connection of generation systems to ENEL low voltage distribution networks, 2006.

3637