Protection of a Low-Voltage DC Ring

Microgrid System

Aswani J

Dept. of Electrical and Electronics Engineering

Amrita Vishwa Vidyapeetham

Kollam, Kerala

aswanijraj@gmail.com

P Kanakasabapathy

Dept. of Electrical and Electronics Engineering

Amrita Vishwa Vidyapeetham

Kollam, Kerala

kanakasabapathy_p@yahoo.com

Abstract— The main challenge for DC distribution system is

the lack of a good protection system. In this paper a ring type DC

microgrid is considered. Upon fault in the grid, the existing

system shut down the entire grid. A new protection algorithm is

proposed which isolates only the faulted section, without the

entire grid shut down. The algorithm senses the fault using

differential current and then current derivative is taken for

segment isolation. The proposed algorithm is verified using

Simulink/MATLAB.

Keywords—DC Microgrids, Fault protection, Segment

isolation, Fault location.

I. INTRODUCTION

Microgrid is a part of electrical distribution system with distributed resources (DRs) and loads which can operate as an isolated power system [1]. The distributed resources are mainly renewable energy based and some of them are PV arrays, Wind turbines, fuel cells, microturbines etc. The microgrid is connected to the main grid at the point of common coupling (PCC). Under normal condition, the loads can be met by the DRs; if necessary it will take power from the main grid also. In islanded mode, microgrid is disconnected from the main grid.

Based on the bus system to which the components are connected the microgrids can be divided into AC microgrid and DC microgrid. For the systems where most of the sources and loads are DC, power system can be solely based on DC. This would eliminate the need for multiple converters to convert DC to AC and then AC to DC. There by reducing the cost, complexity and possibly increasing the efficiency [2].

Some of the advantages of DC microgrids are increase in system efficiency due to the reduction in conversion losses between the DC output sources and loads, absence of grid synchronization with utility grid and reactive power etc. Due to presence of stored energy of DC capacitors and the voltage control of AC/DC converter, the DGs in the microgrid are not easy to trip against the blackouts or voltage sags occurring in the utility grid, thus DC microgrid has its own fault ride through capability. On the other hand, the main challenge of a DC microgrid is the lack of an efficient protection system [3].

Some of the specific applications of DC distribution system are telecommunications, navy shipboard, aircraft and automotive power systems. DC grids are promising solution for environment friendly rural areas and industrial power parks. As a futuristic vision application of DC technology extends to the distribution level for offshore wind farms and also the city centers. The protection in DC system is more difficult than that of AC system because of the absence of zero cross point of current in DC system. DC rated protective systems require quenching chamber to divide the arc into smaller arcs, this makes the protective device bulkier and costlier [4].

Commercially available protection devices for low voltage DC (LVDC) systems are fuses, molded-case circuit breakers, LV power circuit breakers and isolated-case circuit breakers. In a case study presented in [5], power electronic switches and ultra-fast hybrid DC CBs or fuses are proposed as possible choices for fast operation. Due to the fast discharge of converter capacitor, instead of taking the amplitude of converter current, current derivative should be taken for fault detection [5].

In [6] the protection of LVDC ring bus microgrid is proposed. The microgrid is divided into segments. For every segment there will be master and slave controller which detects and isolates the faulted segment without de-energizing the entire grid. The system uses the differential relaying principle for fault identification. In [7], the ring microgrid is divided into nodes and links where each node have controller which can control the circuit breakers associated with the node. Upon fault on any link, the faulted link alone gets isolated. [8] Proposes a protection scheme for DC microgrid which uses line current derivative. The line current derivative depends on the line inductance and thus it is used for locating the fault.

In this paper a new protection algorithm is proposed for the DC ring microgrid protection. The ring microgrid is divided as segments between the nodes. The algorithm utilizes the differential current in the ends of the segment for sensing the fault and then line current derivative is taken for segment isolation. For every node there will be controller which can control the circuit breakers nearer to it. The line current derivative can also be used for fault location.

978-1-4673-9925-8/16/$31.00 ©2016 IEEE 17

Fig. 1. DC ring microgrid structure

The rest of the paper is organized as follows; section II gives the idea about microgrid configuration and proposed algorithm. Fault location identification is described in section III; section IV gives the simulation details of the proposed system and V describes the results and observations of the simulations and fault location estimation.

II. PROPOSED PROTECTION SYSTEM

A. LVDC Microgrid Configuration

A ring type DC microgrid is considered in this paper. Two sources and a load are connected to the DC bus to form a ring like structure. The proposed system configuration is shown in Fig. 1. The sources and the loads can be either AC or DC.

The point at which the components are connected at the bus is called a node. Between the nodes there will be segments, since a bipolar system is considering, each segment will have a positive and a negative pole. In each pole there will be solid state circuit breakers at both ends. Likewise there are current sensors at the ends of the segment. Thus there will be four circuit breakers and current sensors in every segment.

For every node there will be a controller which can operate the circuit breakers associated with it. The controller senses current from the adjacent current sensors and can read data from the controllers on either side. The proposed system with the controllers and protection algorithm is shown in Fig. 2

B. Possible faults in LVDC microgrids

Different fault types in DC microgrids are pole-pole fault and pole-ground fault. Pole-pole fault occurs when the positive and negative pole of the DC system is short circuited, whereas pole-ground fault occurs when either positive or negative pole comes in contact with the ground [9].

C. Proposed Algorithm

The proposed relay algorithm is shown in Fig. 3. Each controller has to perform four algorithms, for two segments on either side having two poles on each. For a particular segment, i1 be the current entering the segment and i1

ꞌ the current leaving

the segment.

Fig. 2. DC ring microgrid with protection.

Fig. 3. Flow chart of the proposed relay algorithm.

Here i1 is sensed by the controller from the nearer sensor and i1

ꞌ is read by the controller from the adjacent controller. If current entering the segment is taken as positive then current leaving will be negative. Under normal condition the current entering and leaving the segment will be equal; therefore their sum will be zero. But under faulted condition there will be a difference in current, so that a possibility of fault is identified. Then the controller checks for the current derivative. If the current derivative is not within the threshold value, a fault is detected. Then controller sends commands to the solid state circuit breakers at the ends of the segment to keep open. Segment isolation is done by taking the current derivative alone, so that a minor fault in the segment does not affect the system.

18

Fig. 4. Simulation circuit of Pole-Pole fault.

Fig. 5. Simulation circuit of Pole-Ground fault.

Since the faulted segment alone is isolated, the microgrid will not get completely de-energized. The load will be powered through the healthy segments.

III. IDENTIFICATION OF FAULT LOCATION

The response of the DC system during a fault is depicted using a small section of the DC system, which can be considered as an RLC equivalent circuit [10]. The current response of the equivalent RLC circuit can be expressed as,

(1)

Where vc(0) is the capacitor voltage and ic(0) is current

through the inductor, just before the fault occurrence. R is the

equivalent resistance of the fault resistance and cable

resistance, L is the cable inductance upto fault point.

Converting (1) to time domain then taking the derivative, and

considering negligible fault resistance, it is observed that only

real part contribute to the fault current [8]. Therefore the

equation is obtained as,

(2)

LCs

L

Rs

siLvsi Lc

1

)0(/)0()(

2

L

v

dt

di c )0(||

19

From the above equation it is clear that as the line current

derivative depends on the line inductance. As the inductance

depends on the line length, the fault distance can be calculated

using the line current derivative.

While considering the fault protection algorithm, since it

finds the current derivative that data can be used for the

identification of fault location. Fuzzy logic can be used for

locating the fault by taking current derivative data as input

which is obtained from the execution of relay algorithm and

then the output will be the distance from the source end. For a

defined fault resistance the fault can be located using Fuzzy

Logic with single input and single output.

IV. SIMULATIONS

A. Pole-Pole Fault

The proposed ring bus is simulated using MATLAB/Simulink as shown in Fig. 4. The ring structure is obtained by considering DC source and DC load. Source voltage is taken as 48V DC and it is boosted to 230V to get 230V DC bus voltage. A pole-pole fault is considered at the middle of the segment with a very small value of fault resistance. There are three segments in the system, for convenience bidirectional circuit breakers and protection algorithm is given only in one segment. The system is simulated for 10sec and the fault is introduced at 3 sec.

B. Pole-Ground Fault

The MATLAB/Simulink diagram of pole-ground fault is shown in Fig. 5. There will be separate loads in positive and negative pole. Either positive or negative pole to ground voltage is taken as 115V. A pole-ground fault is given at the midpoint of the positive pole of a segment. The system is simulated for 10sec and the fault is introduced at 3sec.

V. RESULTS AND DISCUSSIONS

A. Pole-Pole Fault

The Fig. 6 shows the simulation results of the proposed

system with pole-pole fault without any protection. After the

fault introduced at 3sec the current entering the segment

increases and settles at a high value, current leaving the

segment will become reversed due to the fault. Fig. 7 shows

the results obtained after introducing the protection algorithm.

At 3sec, when the fault occurs, the algorithm senses the fault

and the segment is isolated, thus the current entering and

leaving the segment is reduced to zero. The load current and

the voltage are retained to its nominal values.

B. Pole-Ground Fault

The Fig. 8 shows the simulation results of the proposed

system with pole-ground fault without any protection. It is

observed that, current entering the segment is rising and

current leaving is reversed. The load voltage and current gets

reduced to zero. After keeping the protection system, the fault

current entering the segment is reduced to zero, also the load

current and voltage are maintained at nominal value, this is

shown in Fig. 9.

0 1 2 3 4 5 6 7 8 9 10

0

200

400

600

800

1000

Time (Sec)

Curr

ent

(A)

Current Entering the Segment

0 1 2 3 4 5 6 7 8 9 10-500

-400

-300

-200

-100

0

100

Time (Sec)

Curr

ent

(A)

Current Leaving the Segment

Fig. 6. Current entering and leaving the segment during pole-pole

fault (without protection)

0 1 2 3 4 5 6 7 8 9 10-5

0

5

10

15

20

Time (Sec)C

urr

ent

(A)

Current Entering the Faulted Segment

0 1 2 3 4 5 6 7 8 9 10-5

0

5

10

15

20

Time (Sec)

Curr

ent

(A)

Current Leaving the Faulted Segment

0 1 2 3 4 5 6 7 8 9 100

5

10

15

20

25

Time (Sec)

Cu

rren

t (A

)

Load Current

0 1 2 3 4 5 6 7 8 9 100

100

200

300

400

500

Time (Sec)

Volta

ge

(V

)

Load Voltage

Fig. 7. Current entering and leaving the segment,load current and

load voltage during pole-pole fault (with protection)

C. Fault Location using Fuzzy Logic

For locating the fault the system is simulated under the

following conditions,

Segment length = 400m

Cable inductance = 0.97mH/km

Cable resistance = 121mΩ/km

20

0 1 2 3 4 5 6 7 8 9 10-100

-50

0

50

100

Time (Sec)

Cu

rren

t (A

)Current Entering the Faulted Segment

0 1 2 3 4 5 6 7 8 9 10-100

-50

0

50

100

Time (Sec)

Cu

rren

t (A

)

Currrent Leaving the Faulted Segment

0 1 2 3 4 5 6 7 8 9 10-5

0

5

10

15

Time (Sec)

Curr

ent

(A)

Load Current

0 1 2 3 4 5 6 7 8 9 10-50

0

50

100

150

200

250

Time (Sec)

Voltage (

V)

Load Voltage

Fig. 8. Current entering and leaving the segment,load current and

load voltage during pole-ground fault (without protection)

Fault is introduced at various points in the faulted segment

then the current derivative obtained is tabulated in Table I.

For implementing the Fuzzy logic, the input and output

membership functions has to be defined. The input-output

membership function for the Fuzzy Logic can be defined

using the range of values of current derivative and the fault

distance. The range of values can be selected from the data

in the Table I.

Thus 10 membership functions can be defined for current

derivative as inputs from 1700 to 900, and the

corresponding output membership functions will be there

for the fault distance from 0 to 400m. The membership

functions defined for input and output are shown in Fig.10.

The rule base for the Fuzzy Logic can be defined based on

the input-output range of values itself. This is shown in

fig.11. A current derivative input in a specific range

corresponds to the distance range in the output.

0 1 2 3 4 5 6 7 8 9 10-5

0

5

10

15

Time (Sec)

Cu

rren

t (A

)

Current Entering the Segment

0 1 2 3 4 5 6 7 8 9 10-5

0

5

10

15

Time (Sec)

Cu

rren

t (A

)

Current Leaving the Segment

0 1 2 3 4 5 6 7 8 9 100

2

4

6

8

10

12

Time (Sec)

Curr

ent

(A)

Load Current

0 1 2 3 4 5 6 7 8 9 100

50

100

150

200

250

Time (Sec)

Voltage

(V

)

Load Voltage

Fig. 9. Current entering and leaving the segment, load current and

load voltage during pole-ground fault (with protection)

TABLE I. RELATIONSHIP BETWEEN FAULT DISTANCE AND CURRENT DERIVATIVE

R (mΩ)

L(mH)

Distance from

source side(m)

Current Derivative

- - At source

side 1666.66

4.84 0.0388 40 1546.66

9.68 0.0776 80 1446.66

14.52 0.1164 120 1333.33

19.36 0.1552 160 1250

24.2 0.194 200 1186.66

29.04 0.2328 240 1113.33

33.88 0.2716 280 1046.66

38.88 0.3104 320 983.33

43.56 0.3492 360 916.66

48.4 0.388 400

(At load end) 860

21

Fig.10. Fuzzy membership functions for input and output

Fig.11 Rule base for the Fuzzy Logic

The results obtained after implementing fuzzy is

given in Table II. Introducing faults at different locations,

estimating the values of slope using the algorithm, an

approximate fault location is obtained using the fuzzy

logic. The relation between the current derivative and the

fault distance is plotted for convenience, as shown in

Fig.12.

TABLE II Fault Location Results

Current Derivative Fault Distance(m)

1570 25.1

1500 61.75

1410 101.9

1290 137

1255 168.4

1215 172.7

1121 235

1080 260

1011 298

955 327

888 323.7

880 323.9

0 50 100 150 200 250 300 350 400800

1000

1200

1400

1600

1800

Distance (m)

Cu

rren

t D

eriva

tive

Current Derivative vs Distance

Fig. 12 . Relationship between current derivative and fault distance

VI. CONCLUSIONS

A new protection system for the DC ring type microgrid without de-energizing of the entire grid has been proposed in this paper. Differential current and current derivative is used for fault sensing and isolation of segment. The current derivative thus obtained is used for locating the fault with the help of Fuzzy Logic. The proposed scheme tested and verified using is MATLAB/Simulink.

REFERENCES

[1] R. H. Lasseter, “Microgrids,” in Proc. IEEE Power Engineering Society

Winter Meeting, vol. 1, no. 2, 2002, pp. 305–308.

[2] M. Starke, L. Tolbert, and B. Ozpineci, “Ac vs. dc distribution: A loss

comparison,” in Proc. IEEE/PES Transmission Distribustion Conference

Exposition, 2008, pp. 1–7.

[3] Kakigano.H, Miura.Y, Isle T,“Low-Voltage Bipolar Type DC MIcrogrid for Super High Quality Distribution”, IEEE Trans.Power Electron.,vol. 25,no. 12, pp. 3066-3055, Dec 2010.

[4] D. M. Saeedifard, M. Graovac, R. Dias, and R. Iravani, “DC power systems: Challenges and opportunities,” in Proc. IEEE Power Energy Soc.Gen. Meeting, July 2010, pp. 1–7.

[5] D. Salomonsson, L. Soder, and A. Sannino, “Protection of low-voltage

dc microgrids,” IEEE Trans. Power Del., vol. 24, no. 3, pp. 1045–1053,

July 2009.

[6] J. Park, J. Candelaria, “ Fault detection and isolation in low voltage DC Bus microgrid system”, IEEE Trans. Power Del., Apr. 2013, vol. 28, no.

2, pp. 779-787.

[7] J. Park, J. Candaleria, Liuyan Ma, Dunn K, “DC Ring-Bus Microgrid Fault Protection and Identification of Fault Location”,IEEE Trans.Power Del.,Oct 2013,vol. 28, no. 4, pp. 2574-2584.

[8] Meghwani.A, Srivastava S.C,Chakrabarti S, “A New Protection Scheme for DC Microgrid using Line Current Derivative”,IEEE Power and Energy Soc. Gen. Meeting,Oct 2015, pp. 1-5.

[9] J. Candelaria and J.-D. Park, “VSC-HVDC system protection: A review

of current methods,” in Proc. IEEE/Power Energy Soc. Power

Syst. Conf. Expo., Mar. 2011, pp. 1–7.

[10] S. D. A. Fletcher, P. J. Norman, S. J. Galloway, and G. M. Burt, “Determination of protection system requirements for dc unmanned aerial vehicle electrical power network for enhance capability and

survivability,” in Proc. IET Electrical Systems in Transportation, vol. 1,

July 2011, pp. 137–147.

[11] P. Jarventausta, P. Verho and J. Partanen, “Using Fuzzy sets to Model the Uncertainity in the Fault Location Process of Distribution Networks,”IEEE Trans. Power Delivery, Apr. 1994 vol. 9, no. 2, pp. 954–960.

[12] P. Kanakasabapathy, Meenakshi Mohan, “Digital protection scheme for microgrids using wavelet transform”, Electron Devices and Solid-State Circuits (EDSSC), 2015 IEEE International Conference on 1-4 June 2015, pp .664 – 667.

22