Development of a Passive Islanding Detection Method based Fuzzy Controller for a Grid Connected Distributed Generator.

DOI : 10.17577/IJERTV5IS120242

Download Full-Text PDF Cite this Publication

Text Only Version

Development of a Passive Islanding Detection Method based Fuzzy Controller for a Grid Connected Distributed Generator.

Himanshu Jyoti Saikia

Dept. of Electrical Engineering National Institute of Technology Rourkela

Odisha, India

Abstract Over recent years the demand for power has increased and hence there has been a growing interest in Renewable Energy Generating units such as wind, water, photovoltaic, fuel cell units. Connection of DG to the power system network effectively reduces the losses and decrease of transmission network capacity with the consumers being provided with higher reliability. Islanding happens when a part of the distribution system turns out to be electrically disconnected from the rest of the power system, yet keeps on being energized by DG associated with the subsystem. It is important when using DG in an interconnected system that the power distributed system is capable of detecting an unintentional islanding condition.

The IEEE 1547 Standard for Interconnecting Distributed Resources with Electric Power Systems directs that for avoiding islands in power systems the protection equipments and measures must be in order. Even if islanding occurs the duration of the situation must be decreased as much as possible. In order to achieve this, various methodologies have been proposed. Intelligent detection methods such as Fuzzy Logic and Wavelet Transformation have been effectively used to create such detection algorithms.

Keywords- Islanding detection, Fuzzy ,ROCOP, ROCOF .

  1. INTRODUCTION

    With the introduction of DGs there has been a significant improvement in electricity service at a higher level of reliability and power quality in comparison to the typical power so availed from the grid. It is more resilient since it serves low power demands continuously. While the central power system remain crucial to the load utility, their flexibility is limited. Large power generation facilities require immense transmission and distribution network to transmit power. DG provides a relative low capital cost in response to incremental changes in power while avoiding transmission and distribution losses by placing the power source within the already existing network and sending power back to grid when needed.

    The recent interest for installation of DG into the consumer end has put forward unique engineering challenges to DG interconnection. Thus the traditional protection systems need to be reconfigured and redesigned for islanding in DG systems. The island formed is an unregulated power system and is considered as undesirable as it poses a danger to the

    utility equipments and deterioration in the power quality. Islanding operation can be permitted owing to the consumer power reliability by means of providing power to the facility where DG is located. But however such arrangements require considerable engineering efforts to enable such island operation which includes a portion of primary systems.

  2. PASSIVE ISLANDING DETECTION TECHNIQUES

    It is based on the monitoring of system parameters such as voltages, current and frequency and do not introduce perturbations unlike active islanding detection methods. During islanding these parameters vary and a certain threshold is set for normal operation conditions. Exceeding this threshold the relay trips and islanding is detected. The precision of setting the threshold is maintained so that differentiation from other disturbances and islanding is made. The detection methods are fast and do not introduce disturbances in the system but however have a large NDZ.

    1. Classification of Passive Islanding detection techniques.

      The passive islanding detection techniques are classified

      as-

      1. Rate of change of output power (ROCOP)

        The rate of change of output power, at the DG side, is greater than the rate of change of output power before the DG is islanded. This method is effective for load and generation mismatch. [8]

      2. Rate of change of frequency (ROCOF)

        During islanding the rate of change of frequency is very high. The rate of change of frequency is given by[5]-

        where

        P= power mismatch at DG side. H = moment of inertia.

        G= rated generation capacity.

        The ROCOF relay performs operation by continuous monitoring of the frequency signal. However the relay fails to operate during equal load condition.

      3. Rate of change of frequency over power

        The in a small generation system is larger than that of the power system with larger capacity. This concept is utilized for islanding detection. [8]

      4. Voltage unbalance

        Voltage unbalance is one of the methods of islanding as disturbances in voltage waveforms are observed. However this method is not completely reliable owing to the fact that distribution loads are single phase and disturbances are very small.

      5. Harmonic distortion

    If the power system has inverter based DG, harmonics are introduced in the system parameters. Monitoring the THD of the voltage waveform after and before islanding can be done to identify the changes. When DG is islanded, monitoring of third harmonic of voltages of DG provides an insight for islanding detection.

  3. SYSTEM FOR THE CONTROLLER APPLICATION.

    The system used for islanding detection comprises of a 9MW wind farm.The wind farm consists of a total of six wind turbines each 1.5MW connected to a 120kV grid through a 30km, 25kV feeder. A 6MW load is connected to the 575V generation bus of the DG. The circuit breaker 1 is responsible for Major Island formation. Islanding is initiated at 0.5 seconds by opening CB1 and CB2 and CB3 remaining closed.

    Fig.1- Power system network studied.

    1. Parameters of the system components-

      1. Generator : Rated kV=120, Vbase=120kV, f= 60Hz, Rated short circuit MVA= 1000

      2. DG : Wind Farm (9 MW) comprising of six 1.5 MW turbine each, exporting power to a 120kV grid through a 30 km 25kV feeder.

      3. Transformer (TR 1): rated MVA= 47, f = 60Hz, rated kV= 120/25, Vbase=25kV, R1 = 0.00267 pu, L1 = .08 pu, R2 = 0.00267 pu,L2

        = .08 pu, Rm= 500 pu, Xm= 500 pu.

      4. Transformer (TR 2): rated MVA= 10, f= 60 Hz, rated kV= 25kV/575V, R1 = 0.0008 pu, L1 = 0.025pu, R2 = 0.0008 pu, L2 = 0.025pu, Rm= 500 pu, Xm= 500 pu.

      5. Distribution line: Three phase PI section,

        30 km, Rated kV=25, Rated MVA=20,Vbase =25kV, R1=0.1153 ohms/km, R0 = 0.413 ohms/km, L1=1.05e-3 H/km,L0=3.32e-3 H/km, C1= 11.33 e-9 F/km, C0 =5.01e-9F/km

      6. Load: 6MW resistive load.

  4. PROPOSED FUZZY LOGIC CONTROLLER The parameters such as voltage and current are extracted

    from the DG location and the calculation of frequency deviation, rate of change of power, rate of change of frequency, positive sequence impedance calculation are done.

    The study includes the loss of mains in which the main circuit breaker CB1 isolates the DG and load at a specified time and creates an islanding condition. For each parameter, three sets of operating ranges, low, medium and high have been used. For instance since any one criterion of the parameter is not sufficient to determine islanding, hence the set of three ranges of individual parameter can be made as a rule base for islanding detection.

    The ranges of the values of each individual parameter are transformed to Membership Functions (MF) of the fuzzy controller. Mamdani rule base has been used owing to its linguistic simplicity. The MF for the parameters are-

    Table 1- Membership Functionsfor frequency deviation.

    Name

    Type

    Range

    Low

    Trapezoidal

    [-3 -2.9 -0.37 -0.3]

    Normal

    Trapezoidal

    [-0.28 -0.24 0.38 0.43]

    High

    Trapezoidal

    [0.47 0.49 2.85 3 ]

    Table 2- Membership Functions for rate of change of frequency.

    Name

    Type

    Range

    Low

    Trapezoidal

    [-34 -33 -3.6 -3]

    Normal

    Trapezoidal

    [-2.65 -2.44 2.84 2.95]

    High

    Trapezoidal

    [3.49 4.53 33 34]

    Table 3- Membership Functions for rate of change of output power.

    Name

    Type

    Range

    Low

    Trapezoidal

    [-2000 -1980 -1335 -1300]

    Normal

    Trapezoidal

    [-1290 -1270 1290 1300]

    High

    Trapezoidal

    [1323 1352 1980 2000]
  5. FLOW CHART OF THE PROPOSED CONTROLLER

    Start

    Calculate deviation in frequency

    Calculate deviation in frequency

    Calculate Rate of change of frequency

    Calculate Rate of change of frequency

    Calculate Rate of Change of Power

    Calculate Rate of Change of Power

    Measure V and I from DG terminal

    Fig.4- Rate of change of frequency during islanding at 0.5 seconds.

    Fig.5- Rate of change of output power during islanding at 0.5 seconds.

    Rule Base for islanding detection

    Rule Base for islanding detection

    Fuzzy Inference System output

    0.75

    0.25

    Islanding Detected

    No islanding

    Islanding Detected

    No islanding

    Fig.2- Flow Chart of the proposed controller

  6. SIMULATION RESULTS

    The system is subjected to islanding at 0.5 seconds by the disconnection of CB1 producing the Major Islanding Condition. The parameters were measured from the DG terminal.

    Fig.3 -Voltage and current at DG terminal with islanding at 0.5 seconds.

    Fig.6 Frequency deviation during islanding at 0.5 seconds.

    The controller processes the input signals and the FIS output signal during islanding at 0.5 seconds is as follows-

    Fig.7- The FIS output during islanding at 0.5 seconds.

    It is observed that the FIS output is 0.25 during normal and

    0.75 during islanding. The FIS output resorts to 0.75 during the starting owning to initial transients during machine startups. The detection speed is recorded at 6.4ms.

  7. CONCLUSION

The proposed controller takes into account the three input parameters for islanding detection and analyses them in the developed rule base. The controller detects islanding during Loss of Mains or Major Island formation.

The controller can be further modified for detection of minor island formation during islanding of multiple DG interconnections. The rule base can be reconfigured for distinction between load change and islanding condition.

REFERENCES

  1. P. Mahat, Z.Chen and B. Bak-Jensen,Review of islanding detection methods for distributed generation, Proc. 3rd Int.Conf. Elect. Utility Dereg. Restruct. Power Technol., pp.2743 -2748, 2008.

  2. Aziah Khamis, Hussain Shareef, Erdal Bizkevelci, Tamer Khatib, A Review of islanding detection techniques for renewable distributed generation systems, ELSEVIER, Renewable and Sustainable Energy Re views 28 (2013) p 483-493.

  3. IEEE 1547 Standard for Interconnecting Distributed Resources with Electric Power Systems, 2003.

  4. C.Y. Chan, T.K.Y. Lau, S.K.K. Ng . An impact study of ROCOF relays for islanding detection 10th International Conference on Advances in Power System Control, Operation & Management, 2015 page 35 (6 .)

  5. Ten, C.F. And Crossley, P.A.; "Evaluation of Rocof Relay Performances on Networks with Distributed Generation", Developments in Power System Protection, 2008. DPSP 2008. IET 9th International Conference on, vol., no., pp.523,528, 17-20 March 2008

  6. Dube, Anirudh, and Aditya Sindhu. "Comparative analysis of passive islanding detection methods for grid-connected Distributed Generators." 2015 Annual IEEE India Conference (INDICON). IEEE, 2015.

  7. Redfern, M. A., O. Usta, and G. Fielding. "Protection against loss of utility grid supply for a dispersed storage and generation unit." IEEE Transactions on Power Delivery 8.3 (1993): 948-954.

  8. Redfern, M. A., J. I. Barrett, and O. Usta. "A new loss of grid protection based on power measurements." Developments in Power System Protection, Sixth International Conference on (Conf. Publ. No. 434). IET, 1997.

  9. Hagh, M. Trafdar, et al. "New islanding detection algorithm for wind turbine."Environment and Electrical Engineering (EEEIC), 2011 10th International Conference on. IEEE, 2011

  10. Rosolowski, Eugeniusz, Arkadiusz Burek, and Leszek Jedut. "A new method for islanding detection in distributed generation. Wroclaw University of Technology, Poljska (2007).

  11. Samantaray, S. R., Khalil El-Arroudi, Geza Joos, and Innocent Kamwa. "A fuzzy rule-based approach for islanding detection in distributed generation."IEEE transactions on power delivery 25, no. 3 (2010): 1427-1433.

  12. Samantaray, S. R., Trupti Mayee Pujhari, and B. D. Subudhi. "A new approach to islanding detection in distributed generations." Power Systems, 2009. ICPS'09. International Conference on. IEEE, 2009.

Leave a Reply