First Version

1

Abstract—This paper presents an open loop synchronization

strategy for connecting an islanded Micro-grid to system. This

strategy adopts the impedance pre-insertion method to reduce

the synchronizing transients to be within an acceptable level, as

long as both the Micro-grid side and system side are operating

within the power quality requirements. To verify the feasibility of

the method, it is applied to a sample case study. The results have

shown that the size of impedance can be designed properly to

effectively mitigate the transient phenomenon, and at the same

time, the impacts to the stability of the generating units are

within the normal limit.

Index Terms—microgrids, islands, open loop synchronization,

impedance Pre-insertion

I. INTRODUCTION

UE to electricity market deregulations and the need to

reduce environmental footprints from centralized electric

power plants, distributed generations (DG) and

renewable resources have been gaining enormous support in

both research and industrial applications. A Micro-grid is

formed by an aggregation of DG units and local loads that is

in complete autonomous operation (as an island) from the rest

of the power grid [1]-[2]. A fundamental component of the

Micro-grid is the existence of a synchronizer, which allows

the connection of the Micro-grid to the rest of the bulk grid.

This process of connecting two operating electric systems is

known as synchronization. Usually, there are 3 common

synchronization scenarios to consider:

- A single synchronous machine synchronizes to grid

- Micro-grid (islanded system) synchronizes to grid

- A grid synchronizes with another grid (During system

restoration)

Similar to a single synchronous machine to system

synchronization, Micro-grids synchronization must be done

carefully. If the synchronization criteria are not met at the

moment when breaker closes, a high transient may occur,

which will result in equipment damage and power quality

concerns [9]-[10]. Faulty synchronization can:

- Damage transformer and synchronous generator

windings primarily due to the peak inrush current,

which may also lead to misoperation of feeder relays

[3].

Y. Zhou is with the Electrical Engineering Department in University of

Alberta, Edmonton, T6G 1H9 Canada (e-mail: yaxiang@ualberta.ca).

- Damage the generator due to the sudden mechanical

stresses experienced by the generator shaft and prime

mover [11].

- Cause generators to become unstable or unable to

maintain in synchronism with the system primarily due

to an initial out-of-step closing of breaker, which may

lead to the relay tripping of generators [13].

From the synchronization concerns addressed above,

the problem of synchronization is to reduce switching

transients and ensure stability of all generators.

The standard synchronization practice is to adjust the

voltage difference ∆V, angle difference ∆θ, and frequency

difference ∆f into an acceptable range before the synchronism

check relay sends the signal to close the breaker [5]. If the

acceptable range is not met, the generators within one system

are tuned based on remote sensing and feedback control of the

frequency and voltage signals from the breaker location. Case

studies of Micro-grid to system synchronizations based on

feedback control strategies have been presented in literature

[6], [7]. Active synchronization strategy that takes into

account of voltage unbalances and harmonic distortions is

presented in [4].

In order to implement such feedback control systems in

reality, infrastructure support such as communication links

must be built to carry the signals from remote sensing to local

machines. The cost of this infrastructure support can be quite

high especially when the Micro-grids are in remote areas,

where it is located far away from the point of interconnection

to the main grid. In addition, synchronizing a Micro-grid to

system is more difficult than synchronizing a single generating

unit to system due to the fact that the Micro-grid’s voltage and

frequency are determined by multiple DG units and local

loads. In other words, for Micro-grid to system

synchronization, multiple communication links would have to

be built to allow the tuning of individual DG units. Therefore,

synchronizing a Micro-grid to system by feedback control can

not only drive up the cost of infrastructure support, but also

may increase the time and effort to synchronize the two

islands. Especially during system restoration process, the

coordinated control of multiple generators within one system

through feedback may delay the time required to synchronize

any surviving islands [8].

Various transient mitigation methods have been

implemented in practice in different industrial applications.

The feedback control method used in generator

synchronization to reduce the voltage difference across the

Open Loop Synchronization of Micro-grid to

System Based on Impedance Pre-insertion

First A. Author, Fellow, IEEE, Second B. Author, and Third C. Author, Jr., Member, IEEE

D

First Version

2

breaker is only one of them. Other methods include:

impedance pre-insertion has been widely used in industry for

reducing the capacitor switching transients [14]; Point-on-

wave switching used in capacitor or transformer energization

[15]; Silicon-Controlled Rectifier (SCR) based soft starter

used in motor starting to reduce inrush current [16]; and

Sequential phase energization for transformer [17].

This paper proposes an impedance pre-insertion

based synchronization strategy that can connect the Micro-grid

to system in an open loop manner. The main advantage of this

strategy is that the synchronization can be performed without

any feedback controls, i.e. reduces the cost and time required

for Micro-grids synchronization.

The rest of the paper is organized as follows. Section

II presents the proposed method for open loop synchronization

in detail. In Section III, the issues to consider for the proposed

open loop scheme are addressed. Simulation results are shown

in Section IV and the paper concludes in Section V.

II. PROPOSED METHOD – OPEN LOOP SYNCHRONIZATION

The open loop synchronization scheme is to design the

impedance pre-insertion value such that the peak transients are

acceptable, when the two parties are operating within the

power quality limits (PQ limits) as shown in Fig. 1. As

shown, impedance is inserted in series with breaker 1 (BRK1).

This impedance is purely an inductor. After breaker 1 is closed

and system has reached a new steady state, the impedance is

bypassed by closing breaker 2 (BRK2).

It is important to realize that both parties typically operate

within their power quality limits (PQ limits) before the

synchronization process. They are also the pre-established

operating regions. The operating regions [19] (i.e. the power

quality requirements) are established by the utility to ensure

the grid operates safely and reliably during acceptable system

disturbances. If both sides of the breaker are not operating

within their normal operating range, the synchronization

process is not permitted.

Based on Fig. 1, once BRK1 is closed, a transient current

will flow through the impedance, which will propagate to the

DG units within the Micro-grid. This transient current is

composed of an AC and a DC component similar to the case

of a three-phase fault at the generator terminal. The influence

of the DC component on the peak transient current can be

taken into account by applying a factor K to the AC

component [18]. The peak current is reduced as the size of

𝑋𝑖𝑛𝑠𝑒𝑟𝑡 increases as shown by (1). The peak current 𝐼𝑝𝑒𝑎𝑘 is

proportional to:

"

| | | |peak

total insert d eq

V VI

X X X X

(1)

where ∆𝑉 is the rms voltage across breaker before it closes.

𝑋𝑖𝑛𝑠𝑒𝑟𝑡 is an inductor inserted at the breaker location as shown

in Fig. 1. 𝑋𝑑′′ and 𝑋𝑒𝑞 are the synchronous machine’s direct

axis subtransient reactance and equivalent circuit impedance

respectively. 𝑋𝑒𝑞 includes any transformer leakage reactances,

series line impedance (positive sequence) and system

impedance.

Pre-established

Operating Region

Bulk Grid Generator,

Microgrid

( )gV pu

( )gf Hz

60.2

59.7

1.10.9 Pre-established Operating Region ( )sV pu

( )sf Hz

60.05

59.95

1.10.9

1 2

BRK 1

BRK 2

Circuit Breaker

Z insert

Fig. 1. Power quality requirements for open loop synchronization

The open loop synchronization procedure is

developed in Fig. 2. As shown, all the measurements including

voltage, angle and slip are performed at the circuit breaker

location. In addition, if both parties are not operating within

their respective PQ limit, then it implies that the power system

potentially has a power quality issue, which should be

resolved first before synchronization is performed. When both

sides are naturally within PQ limits under the normal system

condition, breaker can be closed at a small angle difference.

This angle difference is explained in more detail in the Section

III. Once the circuit breaker is closed, the impedance could be

bypassed shortly by closing BRK2 after system stabilizes and

reaches a new steady state operation.

Fig. 2. Procedure for performing open loop synchronization

In order to implement the open loop synchronization

scheme, there are some fundamental issues, which have to be

considered. These issues will be addressed in the next section.

III. ISSUES TO SOLVE

The issues that need to be solved in this section include:

- What is the acceptable level of transients?

- How to determine the proper impedance value?

- Can the Micro-grid maintain stability with the designed

impedance?

- Are the impedance bypass transients acceptable?

First Version

3

A. Acceptable Transient Level

Acceptable transient level is calculated based on the

synchronization criteria from the standard IEEE C50.12 for

salient-pole synchronous generators [20]. The standard

specifies, “Generators shall be designed to be fit for service

without inspection or repair after synchronizing that is within

the limits listed…”. The limits are:

- Angle ±10° - Voltage 0 to 5% (of nominal)

- Slip ±0.067𝐻𝑧

In other words, the current and torque transients

experienced by the generator under the above conditions are

considered acceptable. The IEEE limits established above can

be illustrated by an acceptable transient region shown in Fig.

3.

The worst acceptable transient is obtained based on

examining the peak current incurred due to the conditions

specified by the four corners of the region in Fig. 3. Based on

the short circuit theory, higher voltage difference across

breaker leads to higher transients, thus, only the right two

corners are remained to examine. The worst case lies in the

case ∆𝑉 = +0.05𝑝. 𝑢. , ∆𝑓 = +0.067𝐻𝑧 due to the fact that

higher machine frequency than system immediately after the

closing of breaker tends to advance the rotor angle and the

transient torque.

Fig. 3. Determine the worst acceptable transient level

B. Impedance Value Design

The method to effectively design the appropriate size of

impedance to limit the worst transients under the open loop

scheme is developed in this section. Firstly, possible worst

case transients have been identified based on the PQ limits.

Secondly, the peak current as a function of impedance is

determined.

Prior to synchronization, each party is expected to operate

within their established power quality requirements, i.e.

voltage and frequency are within their normal operating range

as specified in Fig. 4. Although the angle requirement is 10

degrees based on IEEE standard, but an IEEE industrial

survey reported that the maximum closing angle can reach up

to 20 degrees when synchronized manually by an operator

[11]. Therefore, for open loop synchronization, a worst case

angle difference of 20 degrees is used throughout this paper.

1V

1f

60.2Hz

59.7Hz

1.1pu0.9pu

2V

2f

60.05Hz

59.95Hz

1.1pu0.9pu

1 2

43

43

21

Micro-grid Side Operating Range System Side Operating Range

Fig. 4. Worst case transients based on open loop synchronization

It is intuitively clear based on Fig. 4 that the largest voltage

and frequency mismatch will result in the largest

synchronizing transients. As a result, there are four

combinational cases to consider, which are case (2, 3), case (4,

1), case (3, 2) and case (1, 4). The first number refers to the

operating point in the Micro-grid side, and the second number

refers to the system side.

Although the peak current due to closing of breaker 1 can

be obtained by repeated dynamic simulations, but the

analytical expression has the advantage of simplifying the

design process. Therefore, analytical expression for

calculating the peak transient current as a function of the

impedance value is derived.

Analytically, the Micro-grid is modelled as a synchronous

generator with a load connected to its terminal. Before the

breaker is closed, the current through the impedance is zero.

However, the stator current of the synchronous machine

consists of a steady state flow of load current as shown in Fig.

5(a). At the instant of breaker switching, this switch can be

represented by two opposing voltage sources in series, as

shown in Fig. 5(b). By using the superposition theorem, the

circuit shown in Fig. 5(b) can be split into two equivalent

circuits, as shown in Fig. 5(c).

SGLI

jXsys

V Vgrid

insertjX

SG

V Vgrid

V

LI

SG

V

VgridLI

V

''

djX

(a) (b)

(c)

jXsys

jXsys jXsys

insertjX

insertjX insertjX

Fig. 5. Equivalent circuit representation of closing breaker 1. (a) Circuit

before breaker 1 is closed. (b) Circuit after breaker 1 is closed. (c) Equivalent circuits after breaker 1 is closed.

The circuit on the left side of Fig. 5(c) represents the

same circuit before breaker is closed. The current in this

circuit represents the steady state component. The right side

circuit can be used to calculate the transient resulting from the

switching. The steady state component of the stator current

equals to the load current. Assuming ∆𝑓 = 0, the analytical

expression for the transient component based on Kirchhoff’s

Voltage Law (KVL) for the circuit is:

Acceptable TransientRegion

f (Hz)

V (pu)

(+0.05pu, +0.067Hz)

Worst Acceptable Transient

(+0.05pu, -0.067Hz)(-0.05pu, -0.067Hz)

(-0.05pu, +0.067Hz)

First Version

4

( )( ) 2 sin( )

di tL Ri t V wt

dt

(2)

where, L and R are the equivalent inductance and resistance

seen by the breaker, respectively. 𝑤 is the angular frequency

at 60𝐻𝑧. is the voltage difference phase across breaker. The

solution to the differential equation (2) is:

/

2 2

2( ) sin( ) sin( ) Rt L

eq

Vi t wt e

X R

(3)

where, eqX includes the series impedances 𝑋𝑑

′′, 𝑋𝑖𝑛𝑠𝑒𝑟𝑡

and𝑋𝑠𝑦𝑠. equals 𝑡𝑎𝑛−1(𝑋𝑒𝑞/𝑅) is the impedance angle.

The peak transient stator current becomes the sum of

load current and the synchronizing current computed

analytically in time domain by (3). It is worth mentioning that

this analytical expression is derived on the basis ∆𝑓 = 0 .

However, the effects due to ∆f from the PQ limits can be

neglected in calculating the peak current. An intuitive

explanation is realized by considering two voltage sources

connected together with different frequencies. If one voltage

source is taken as a reference running at angular frequency 𝑤1,

the instantaneous voltage difference when the breaker is

closed at t=0 is dictated by:

1 1( )

1 2| | | |j w t jw tj wtV V e e V e

(4)

∆𝑤 represents the angular frequency difference between the

two voltage phasors having magnitudes 1V and

2V ,

respectively.

Based on the time window of sub-transient, the peak

inrush current typically occurs within t < 2 cycles from the

initial start of synchronization. Therefore, the impact of ∆𝑓 on

∆𝑉 across the breaker is negligible since ∆𝑤𝑡 ≈ 0.

C. Transient Stability Evaluation

The section is to investigate the transient stability of

the generator due to the impact of open loop synchronization.

It is well known that the power transfer capability or the

stability margin between a generator and the system decreases

due to a fault on one of the parallel transmission lines. As the

line trips, the equivalent impedance between the generators to

the system are increased, thus the stability margin is

decreased. A similar phenomenon occurs when the impedance

is inserted at the synchronizing breaker location. Therefore,

the goal is to ensure that the worst case maximum rotor angle

reached does not go beyond the stability limit of a generator.

Furthermore, the effect of high Micro-grid loading levels may

aggravate the impacts on stability, so it should also be

considered in the analysis.

1) Generator to system synchronization

The synchronization phenomenon from a transient

stability point of view can be explained by a generator to

system synchronization example in Fig. 6. The power angle

curve (P-δ curve) is widely used for conducting transient

stability analysis, e.g. application of equal area criterion based

on power angle curves to find the critical fault clearing time to

ensure generators remain in synchronism with the system.

Similarly, power angle curves and equal area criterion can be

applied accordingly to analyze the synchronization

phenomenon after breaker is closed as shown in Fig. 7.

'E 'jXdtjX

sysZ

Infinite Bus

1 2

1 2

1 2

V V V

f f f

Synchronizing

Breaker

Fig. 6. Generator to system synchronization example

Before the closing of synchronizing breaker, the

electric power transfer between the generator and system is

zero. As the breaker is closed, the generator experiences a

sudden loading condition primarily due to the relative phase

difference to the grid. If the Micro-grid frequency is higher

than the system prior to synchronization, the rotor angle is

going to increase further until the synchronous frequency (i.e.

the system frequency) is reached after breaker closes. This

point where maximum rotor angle occurs is denoted by 𝛿𝑚𝑎𝑥

in Fig. 7. The rotor during this time period is dissipating the

kinetic energy offset stored initially. After 𝛿𝑚𝑎𝑥 is reached,

the rotor begins to swings in the opposite direction.

Eventually, the electromechanical oscillation will stabilize due

to the effects of damping.

-90 0 90 180Power Angle (deg)

Pe

o

max o

min o

Kinetic energy dissipation

due to f1>f2Pm

maxPe

o Initial angle difference

across breaker

Fig. 7. Power angle curve for generator to system synchronization

Based on the classical definition of the swing

equation with damping ignored, the motion of the rotor

dynamics are [21],

2

2

2 ( )

( )( )

m e

syn

syn

H d tP P

w dt

d tw t w

dt

(5)

The electrical power transfer from the internal voltage of

generator to the system after breaker is closed can be

expressed as: '

'sin( )s

e

d t sys

E VP

X X X

(6)

where, 𝐸′ represents the internal voltage behind direct axis

transient reactance, sV is the voltage magnitude of the system,

First Version

5

and 𝛿 is the power angle between E’ and sV . The first

equation of (5) by taking integral can be expressed as:

maxmax

00

2

( )

m e

syn

H dP P d

w dt (7)

For generator synchronization, since the mechanical power

𝑃𝑚 ≈ 0 , equation (7) can be integrated and expressed in a

closed form to obtain the value 𝛿𝑚𝑎𝑥 in the first swing:

1 2

max 0 1 2'cos cos( ) [2 ( )]

T

s syn

HXf f

EV w (8)

where, 𝑋𝑇 is the total series reactance, H is the machine inertia

constant, and 𝛿0 = ∆𝜃 is the initial angle difference when

breaker is closed.

2) Micro-grid to system synchronization

In the case of Micro-grid to system synchronization,

an equivalent diagram for stability analysis is shown in Fig.

8(a). Before the breaker closes, the power is mainly consumed

by the load as shown in Fig. 8(b). After the breaker closes in

Fig. 8(c), the power produced by the generator is exchanged

with the grid.

'E 'jXdLjX sysjX

Infinite Bus

LZ

Z insert

1 2

1 2

1 2

V V V

f f f

'E 'jXd

LZLI

0Vt 'E 'jXd eqjX

Infinite Bus

LZ

(a)

(b) (c)

sV

sV

Fig. 8. Circuits for transient stability analysis (a) Micro-grid to system

synchronization schematic diagram. (b) Before closing breaker (c) After breaker is closed.

The load is modelled as constant impedances for transient

stability studies. Y-∆ transformation is used to eliminate the

node Vt of Fig. 8(c) such that the nodes retained are shown in

Fig. 9, which are only the internal voltage 𝐸′, the infinite bus,

and the reference node to ground.

'E 12Z

10Z20Z

Y 1 2

Fig. 9. Equivalent circuit for one machine with load to an infinite bus

From network theory, the real power at node 1 of Fig.

9 is given by *

1Re EI or expressed as:

'2 '

_ 11 11 12 12cos( ) cos( )e after sP E Y E V Y

(9)

where,11 10 12Y Y Y ,

10 101/Y Z , and 12 121/Y Z .

12Z is the

series impedance of the transmission network, including

transformers, lines and the value of impedance pre-insertion.

10Z is the equivalent shunt impedance connected to the

machine terminal, which includes any local loads.11 and

12

are the impedance angle corresponding to 11Y and

12Y ,

respectively.

Similar to synchronization of generators, the

theoretical 𝛿𝑚𝑎𝑥 for the first swing when Micro-grid

synchronizes to system can be obtained by substituting (9) into

(7), which gives a non-linear equation:

12 12 max 12

2

1 2 max

' [sin( ) sin( )]

/ [2 ( )] ( )( ) 0

s o

syn m c o

E V Y

H W f f P P

(10)

where,'2

11 11cos( )cP E Y represents the power dissipation in

the network from first part of (9). Due to the impact of

loading, the initial power angle 𝛿0 in this case is the sum of

the initial angle difference across breaker and the loading

angle before synchronization.

The impacts of load and impedance pre-insertion on

the power-angle characteristic are shown in Fig. 10 and Fig.

11, respectively. Both figures examine the case where Micro-

grid frequency is higher than the synchronous frequency.

-50 0 50 100 150-1

0

1

2

3

4

5

Load Angle (deg)

Ele

ctr

ic P

ow

er

(pu)

Before Breaker Close

After Breaker Close

Pm

max o

min o o

Fig. 10. Impact of load to Micro-grid synchronization

-50 0 50 100 150-1

0

1

2

3

4

5

Load Angle (deg)

Ele

ctr

ic P

ow

er

(pu)

Before Breaker Close

After (X Insert = 0)

After (X Insert = 0.1 pu)

After (X Insert = 0.3 pu)

Pmax

max 0

min 0

0

Pm

Fig. 11. Impact of impedance pre-insertion to Micro-grid synchronization

According to Fig. 10, increasing the loading level of

Micro-grid reduces the stability margin, where the margin is

defined as the ratio of the difference (𝑃𝑚𝑎𝑥 − 𝑃𝑚 ) to 𝑃𝑚𝑎𝑥 .

Furthermore, by increasing the impedance size, the stability

margin is decreased significantly. Therefore, it is true that both

factors negatively impact the power transfer capability

between the Micro-grid to the system. However, the maximum

First Version

6

rotor angle reach is mainly affected by the loading level of the

Micro-grid and the phase difference across the breaker.

Based on the open loop synchronization requirements

and transient stability analysis, the worst case that should be

examined in terms of maximum rotor angle advancement is

based on the peak loading level and the designed impedance

value. In addition, the voltage levels (Micro-grid and system

sides) should be at the lowest of the PQ limits according to

(9). The frequency difference should be positive (i.e. higher

Micro-grid frequency than system) to result in higher rotor

advancement within the first swing. As a result, the worst case

open loop synchronization scenario from a transient stability’s

point of view is case (1, 3) from Fig. 4.

D. Impedance bypass consideration

Similar to the first breaker switching, impedance

bypass also produces a second transient effect on the system.

Therefore, the peak current from the second transient must be

evaluated to ensure an acceptable disturbance level.

According to the principal of superposition, bypass

switching can be represented by two opposing voltage sources

as shown in Fig. 12, which is equivalent to a single voltage

source in the transient circuit. This circuit is quite similar to

synchronization but with ∆𝑓 = 0 . The voltage difference

across the impedance before the bypass depends on both the

impedance value and the amount of current flow through the

impedance.

Impedance Insertion

Micro-grid,Generators Bulk Grid

Superposition

Impedance InsertionV1 V2 V1 V2

V

Fig. 12. Representing impedance bypass by equivalent voltage source

In the impedance design section, an acceptable

transient level is determined based on the IEEE C50.12.

Similar approach can be taken for impedance bypass. If the

voltage across the impedance before bypass is lower than the

acceptable bypass voltage, the bypass peak current will not be

a concern. The acceptable voltage level for bypass according

to IEEE C50.12 and Fig. 3 is computed to be 0.1855 p.u.

The actual voltage across the impedance in steady

state can be computed through load flow studies. The

initializations to load flow study should consider the

Automatic Voltage Regulation (AVR) and Governor Control

settings because it will affect both the reactive and active

power output from the synchronous machine. As a

consequence, the machine controls can affect directly the

power flow through the impedance, which may result in

higher bypass voltage.

Based on the open loop power quality limits in Fig. 1,

the voltage range for all busses in the system is typically

between 0.9-1.1 p.u. The AVR of synchronous generator

generally maintains the terminal voltage of the machine at a

preset level. After the impedance switching, since the grid

voltage is stiff, the busses near the point of interconnection

tend to follow the system voltage. However, due to the nature

of open loop synchronization, the voltage set-point of AVR is

not adjusted during the process of synchronization, which

means it is regulating the terminal voltage to the level prior to

synchronization. Therefore, there exists a reactive power (Q)

flow at the impedance in steady state. In addition, Q can flow

in either direction through the impedance.

Normally in a Micro-grid, when there is more than

one generator, the speed governors are operating in droop

control mode to assure proper load sharing between the DG

units. From synchronization point of view, before the

impedance bypass in steady state, the Micro-grid frequency

becomes the system frequency. The governor responds to the

change in frequency by a percentage change in the mechanical

torque or output power [21] as in (11). Therefore, real power

(P) flow also exists through the impedance.

(%) 1

( )microgrid sys

m

nom

f ffP pu

R f R

(11)

where, R is the droop constant andnomf is the rated machine

frequency.

The larger the frequency and voltage differences

prior to synchronization, the larger the power flowing through

the impedance due to AVR and governor controls. Based on

the open loop power quality limits, possible worst case bypass

voltages can be identified as case (2, 3), case (4, 1), case (3, 2)

and case (1, 4).

IV. SIMULATION RESULTS

The system under study consists of one 6.6 MVA

synchronous generator connected in a 30 mile (25kV) feeder

to the main substation. A peak loading of 6MW is evenly

distributed along the feeder as shown in Fig. 13. The three-

phase short circuit level at the point of interconnection is 346

MVA. The system and synchronous generator data can be

found in Appendix. The impedance pre-inserted at the circuit

breaker location is purely an inductor.

15 miles 15 miles

25/4.16kVYg/Yg

SGSGSUB

25kV

346MVA

CB

6.6MVA

+

AVR

2.0MW

0.65MVAr

V1,f1V2,f2

Impedance

Insertion

2.0MW

0.65MVAr

2.0MW

0.65MVAr Fig. 13. Single line diagram of case study

Determination of the acceptable transient level is

illustrated through the case study presented in Fig. 13. Based

on the EMTP simulations in MATLAB/Simulink, current and

torque peaks under the operating point specified in Fig. 3 are

found to be 1.5 and 1.7 p.u., respectively. Their time-domain

simulation results are shown in Fig. 14.

First Version

7

Fig. 14. Time domain simulation to obtain acceptable transients

Possible worst case synchronizing transients based on

Fig. 4 will take place in the following 4 cases shown in Fig.

15, where the operating point of both parties deviate the most.

The worst transient of all occurs in case (2, 3). This is because

prior to synchronization, the Micro-grid side voltage is

maintained at 1.1 p.u. and the Micro-grid frequency is

+0.25Hz higher than the system frequency.

Fig. 15. Worst case transients based on power quality limits

Since the worst transients are identified in case (2, 3), then

the size of impedance can be chosen either by repetitive

EMTP simulations or by the analytical expression shown in

(3). A comparison of the two approaches is shown in Fig. 16.

The analytical method provides an upper bound for the peak

stator current. For this case study, the impedance value

determined analytically is 0.6 p.u. or 56.82 Ω.

Fig. 16. Peak currents versus sizes of impedance pre-insertion for case (2, 3)

For transient stability analysis, the worst case occurs

in case (1, 3). The theoretical prediction of the maximum rotor

angle reached is obtained by simplifying the circuit in Fig. 13

to Fig. 9 using repetitive Y-∆ transformations. Therefore, the

maximum swing angle can be calculated by using (10). From

Fig. 17, it can be seen that increasing the inserted impedance

does not have a significant effect on the maximum rotor angle.

Based on the acceptable stability limit and the designed

impedance value, a feasible impedance range is within 0.6p.u.-

1.1p.u.

Fig. 17. Feasible impedance range based on short circuit and stability analysis

In order to evaluate the severity of bypass transients,

load flow study is used to find the worst bypass voltage among

the four possible cases. Case (4, 1) is the worst case as shown

in Fig. 18 because it results in the largest complex power |S|.

This |S| through the designed impedance (0.6p.u.) incurs a

voltage difference of 0.178p.u. across the bypass breaker,

which is lower than the acceptable level of 0.1855p.u.

Therefore, the bypass transient is acceptable.

-0.1 -0.05 0 0.05 0.1 0.15-0.4

-0.3

-0.2

-0.1

0

0.1

0.2

0.3

0.4

Real Power Flow (p.u.)

Reactive P

ow

er

Flo

w (

p.u

.)

maxS

Case (4, 1)

Case (2, 3)

Case (3, 2)

Case (1, 4)

Fig. 18. Worst possible power flows through the impedance computed by

load flow studies

V. CONCLUSION

The purpose of this work is to propose an open loop

synchronization strategy for a Micro-grid to system, which can

be applied to practical Micro-grid implementations. An

impedance pre-insertion method has been proposed to reduce

synchronizing transients. Based on this idea, an open loop

scheme to synchronize Micro-grids to grid has been

developed. Methods to design the open loop scheme are

established. Study results have shown that the proposed

scheme is highly feasible. The potential uses of the proposed

scheme include reducing cost of Micro-grid synchronization

and support fast restoration of power systems.

APPENDIX

TABLE I THEVENIN EQUIVALENT (SUB) DATA

Short circuit power (MVA) 346

Nominal voltage (kV) 25

0 0.5 1 1.5-2

-1

0

1

2

time (s)

Sta

tor

Curr

ent

(pu)

0 0.5 1 1.50.5

1

1.5

2

time (s)

Torq

ue T

e (

pu)

Breaker closes to begin synchronization

Acceptable torque transient

Acceptable current transient

0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.81

1.5

2

2.5

Impedance Size (pu)

Peak C

urr

ent

(pu)

EMTP

Analytical

Acceptable

Z Insertion = 0.6 pu

Acceptable transient level

First Version

8

X/R Ratio 7

TABLE II SYNCHRONOUS GENERATOR DATA

Nominal Power (MVA) 6.6

Nominal Voltage (kV) 4.16

Pair of poles 2

Inertia constant (s) 2.52

Stator resistance (pu) 0.00

𝑋𝑑(pu) 1.26

𝑋𝑑′ (pu) 0.26

𝑋𝑑′′ (pu) 0.18

𝑋𝑞(pu) 1.24

𝑋𝑞′′ (pu) 0.21

𝑇𝑑𝑜′ (pu) 1.4446

𝑇𝑑𝑜′′ (pu) 0.0218

𝑇𝑞𝑜′′ (pu) 0.0690

TABLE III EXCITATION SYSTEM DATA

AVR IEEE Type 1

𝑇𝑟 5ms

𝐾𝑎 300

𝑇𝑎 50ms

𝐾𝐸 1.0

𝑇𝐸 0.65s

𝐾𝑓 0.048

𝑇𝑓 0.95s

𝐸𝑓𝑚𝑖𝑛 -5

𝐸𝑓𝑚𝑎𝑥 8

REFERENCES

[1] F. Katiraei, M. R. Iravani, and P. W. Lehn, “Micro-grid Autonomous

Operation During and Subsequent to Islanding Process,” IEEE Trans. Power Del., vol. 20, no. 1, Jan. 2005.

[2] W. Kohn, Z. B. Zabinsky, and A. Nerode, “A Micro-grid Distributed

Intelligent Control and Management System,” IEEE Trans. Smart Grid, Aug. 2015.

[3] H. Laaksonen, D. Ishchenko, and A. Oudalov, “Adaptive Protection and

Micro-grid Control Design for Hailuoto Island,” IEEE Trans. Smart Grid, vol. 5, no. 3, May. 2014

[4] F. Tang, J. M. Guerrero, J. C. Vasquez, D. Wu, and L. Meng,

“Distributed Active Synchronization Strategy for Micro-grid Seamless Reconnection to the Grid Under Unbalance and Harmonic Distortion,”

IEEE Trans. Smart Grid, Mar. 2015

[5] M. J. Thompson, “Fundamentals and Advancements in Generator Synchronizing Systems,” 65th Annual Conference for Protective Relay

Engineers, pp. 203-214, Apr. 2012.

[6] T. M. L. Assis, G. N. Taranto, “Automatic Reconnection From Intentional Islanding Based on Remote Sensing of Voltage and

Frequency Signals,” IEEE Trans. Smart Grid, vol. 3, no. 4, Dec. 2012.

[7] C. Cho, J. Jeon, J. Kim, S. Kwon, K. Park, and S. Kim, “Active

Synchronizing Control of a Microgrid,” IEEE Trans. Power Electron,

vol. 26, no. 12, Dec. 2011.

[8] Ontario Power System Restoration Plan, IESO, Ontario, 2015 [9] L. C. Gross, L. S. Anderson, and R. C. Young, “Avoid Generator and

System Damage Due to a Slow Synchronizing Breaker,” proceedings of the 24th Annual Western protective Relay Conference, Spokane, WA,

Oct. 1997.

[10] IEEE Guide for AC Generator Protection, IEEE standard C37.102-2006.

[11] W. M. Strang, C. J. Mozina, B. Beckwith, T. R. Beckwith, S. Chhak, E.

C. Fennell, E. W. Kalkstein, K. C. Kozminski, A. C. Pierce, P. W.

Powell, D. W. Smaha, J. T. Uchiyama, S. M. Usman, and W. P.

Waudby, “Generator synchronizing, industry survey results,” IEEE

Trans. Power Del., vol. 11, no. 1, pp. 174-183, Jan. 1996. [12] J. V. Mitsche and P. A. Rusche, “Shaft torsional stress due to

asynchronous faulty synchronization,” IEEE Trans. Power App. Syst.,

vol. PAS-99, no. 5, pp. 1864-1870. Sep. 1980. [13] K. Malmedal, P. K. Sen, and J. P. Nelson, “Application of out-of-step

relaying for small generators in distributed generation,” IEEE Trans.

Ind. Appl., vol. 41, no. 6, pp. 1506-1514, Nov./Dec. 2005. [14] R. P. O’leary, R. H. Harner, “Evaluation of Methods for Controlling the

Overvoltages Produced by the Energization of a Shunt Capacitor Bank,”

International Conference on Large High Voltage Electric Systems, Paris, Aug. 1988.

[15] R. W. Alexander, “Synchronous Closing Control For Shunt Capacitors,”

IEEE Trans. Power App. Syst., vol. PAS-104, no. 9, Sep. 1985. [16] A. Nied, J. Oliverira, R. F. Campos, R. P. Dias, and L. C. S. Marques,

“Soft Starting of Induction Motor With Torque Control,” IEEE Trans.

Ind. Appl., vol. 46, no. 3, May/June. 2005. [17] Y. Cui, S. G. Abdulsalam, S. Chen, and W. Xu, “A Sequential Phase

Energization Technique for Transformer Inrush Current Reduction –

Part I: Simulation and Experimental Results,” IEEE Trans. Power Del., vol. 20, no. 2, Jan. 2005.

[18] A. J. Rodolakis, “A comparison of North American (ANSI) and

European (IEC) fault calculation guidelines.” IEEE Trans. Ind. Appl., vol. 29, no. 3, pp. 515-521, May/Jun. 1993.

[19] Technical Requirements for Connecting to the Alberta Interconnected Electric System (IES) Transmission System, ESBI, Alberta, 1999.

[20] IEEE standard for Salient-Pole 50 Hz and 60 Hz Synchronous

Generators and Generator/Motors for Hydraulic Turbine Applications Rated 5 MVA and Above, IEEE Standard C50.12-2005.

[21] P. Kundur, “Power System Stability and Control”, in McGraw-Hill,

1994, pp. 581-600.

Final Version

9

Abstract—This paper proposes a novel method to enable open-

loop synchronization of a microgrid to the main grid. The idea is

to pre-insert an impedance to reduce the synchronization

transients and then bypass it after the initial transients are over.

With this method, infrastructure cost and complexity of

synchronization can be reduced significantly since the

communication link between the breaker and the microgrid

generators is no longer required. In addition, the extra effort

required for generator adjustments, especially for multiple units,

can be avoided. Technical considerations and design method for

the selection of pre-insertion impedance size is presented. A

simulation study is conducted to evaluate the performance of the

method. The results prove that the transient levels can be

effectively reduced and open-loop synchronization is indeed

achievable.

Index Terms—Impedance insertion, islanding, microgrid,

open-loop synchronization.

VI. INTRODUCTION

ICROGRIDS are becoming an integral feature of the

power systems due to their environmental and

economical benefits. A microgrid includes a variety of

distributed energy resource (DER) units (including distributed

generation (DG) and energy storage devices and different

types of load. Improved reliability and sustainability are some

of the desired characteristics affecting the distribution level

provided by the implementations of microgrids [22].

Microgrids are able to operate in both grid-connected and

islanded modes. For effective operation in islanded mode,

several methods have been proposed for controlling the

microgrid [2-4].

Among the various types of distributed generators proposed

for microgrids, the synchronous generators (SG) are actually

the most common type. Examples are combined heat and

power (CHP), internal combustion engine, and small hydro

[26]. Each time the microgrid switches from islanded to grid-

connected operation mode, the grid and its SGs need to be

synchronized with the main grid. Precaution is necessary to

make sure that synchronization criteria are met to avoid

equipment damage and power quality concerns [27], [28].

In the traditional synchronizing practices, voltage across the

The authors are with Electrical and Computer Engineering Department of

University of Alberta, Edmonton, AB T6G 1H9, Canada (e-mail:

wxu@ualberta.ca).

breaker at the substation location is monitored and sent to the

operator at the generation site. The operator then uses these

measurements to adjust governor and exciter settings to meet

the synchronization criteria [29]. Automatic synchronization

methods based on feedback control strategies have been

presented in literature [30], [31]. Recently, an active

synchronization strategy that takes into account voltage

unbalances and harmonic distortions is presented in [32]. In

order to implement such feedback control systems in practice,

infrastructure support such as communication links are

necessary to carry the signals from the breaker location

(substation) to the local machines as shown in Fig. 1. The

problem is that the microgrid might be located in a rural or

forestry area far away from the point of connection to the main

grid. For example, in [30] it is stated that these rural

microgrids are becoming common in Brazil. In such cases, the

cost of communication infrastructure can be prohibitively

high.

Additional difficulties arise when there are multiple DG

units in the microgrid. In this case, multiple communication

links are necessary to allow the tuning of individual DG units.

Therefore, synchronizing by feedback control can drive up the

cost of infrastructure support and increase synchronization

effort.

~

Microgrid

Synchronizing

Controls

Communication Link

Main Grid

Machine

Controls

CB

Fig. 1. Schematic of the synchronization methods with feedback control

In response to the above challenges, a novel open-loop

synchronization method is proposed in this paper. As the name

implies, the method does not require a communication link

between the DG units and the breaker location. The proposed

method is inspired by the practices of controlling switching

transients, since synchronization is an event where switching

transients is the main concern. The idea here is to pre-insert an

impedance to reduce the synchronization transients. With this

method, infrastructure cost and complexity of synchronization

A Method to Enable Open-Loop

Synchronization of a Microgrid to Main Grid

Yaxiang Zhou, Student Member, IEEE, Wilsun Xu, Fellow, IEEE, Moosa Moghimi Haji, Student Member,

IEEE, Jing Yong, Member, IEEE

M

Final Version

10

can be reduced significantly since the communication link

between the breaker and the microgrid generators is no longer

required.

The rest of the paper is organized as follows. The proposed

scheme is presented in Section II. The impedance design

method is explained in Section III. A case study is conducted

in Section IV to evaluate the effectiveness of the proposed

method. Finally, the conclusion is presented in Section V.

VII. PROPOSED OPEN-LOOP SYNCHRONIZATION METHOD

The main concern on synchronizing two systems or a

generator with a system is the transients produced when the

synchronizing circuit breaker is closed. Excessive transients

can lead to large inrush current and transient torque, damaging

generator and other equipment [33], [34]. The current practice

of reducing synchronization transients is to limit the voltage,

angle, and frequency differences between the two parties.

There are other ways to reduce switching transients [35]. A

good example is the impedance insertion/bypass scheme used

to limit capacitor switching transients. In this scheme, the

inserted impedance increases the total circuit impedance and

thus reduces the inrush current. The impedance is bypassed

after the system has reached the steady-state. In view of the

excellent performance of this scheme on reducing switching

transients, one may wonder if it can be used to simplify the

synchronization of a microgrid to a system.

Before synchronization, the microgrid and the system are

expected to operate within their respective power quality

limits. This means that each party has a known operating

region of voltage and frequency at the synchronization point.

The proposed idea is to establish a value of the pre-insertion

impedance such that the resulting transient is always within

acceptable limit as long as the two parties are operating within

their power quality limits at the time of synchronization.

Consequently, feedback control is not needed to adjust the

microgrid operating point. Fig. 2 illustrates the proposed

scheme.

It is clear that a larger Zinsert will result in more reduction of

inrush currents. However, it may increase the transient when

the impedance is bypassed. In addition, larger Zinsert will

weaken the synchronizing power between the microgrid and

the system so instability may occur. Therefore, the following

issues must be addressed for the proposed scheme:

a) What is the acceptable synchronization transient level?

b) What is the minimum value of Zinsert which can lead to

acceptable synchronization transients?

c) What is the maximum value of Zinsert that can maintain

system stability and results in acceptable bypass

transients?

In the following section, the above issues are addressed one

by one. A method to design the proposed scheme is developed

accordingly.

VIII. IMPEDANCE SIZE DESIGN

Current surges and power oscillations should be taken into

account for designing the impedance size. The first step is to

find the maximum acceptable surge level and power

oscillation. Maximum surge level could be expressed by

maximum current and maximum torque. These values are

calculated based on the synchronization criteria from the

standard IEEE C50.12 for salient-pole synchronous generators

[38]. The limits are:

- Angle ±10° - Voltage 0 to 5% (of nominal)

- Slip ±0.067𝐻𝑧

These limits are established to ensure an acceptable

transient level experienced by the generator in terms of stator

current and shaft torque. In order to find the worst transient,

the limits have been transformed into an acceptable region

shown in Fig. 3. As the synchronizing condition moves away

from the origin, the transient level will increase due to an

increase in the voltage across the breaker. As a result, the four

corners of the region should be examined through simulations

to determine the maximum acceptable current and torque,

since they are farthest from the origin.

Fig. 3. Acceptable transient region defined by standard

The maximum power oscillation level is expressed by the

maximum rotor angle which usually happens at the first

swing. A limit is required to ensure that the synchronous

generator of the DG unit will successfully be synchronized

with the grid and will not become unstable. The industry

practice is to consider 45 degrees as the maximum rotor angle

for transient stability studies [39]. The same value is used in

this paper as the maximum acceptable rotor angle.

The current transients and transient stability of the

Fig. 2. The scheme of the proposed method and power quality limits of the

both sides

Final Version

11

microgrid during the open-loop synchronization are analyzed

in the next subsections. Then the impedance design method is

presented.

A. First switching surges

As mentioned before, the method is meant to enable open-

loop synchronization of the micro grid at any normal operating

point of the microgrid and the main grid shown in Fig. 4. To

make sure that the current and torque transients are both lower

than their acceptable limits, all of the operating points of the

two systems should be examined. However, it is evident that

the largest voltage and frequency mismatch will result in the

largest synchronizing transients. Therefore, there are four

combinational cases to consider: case (2, 3), case (4, 1), case

(3, 2) and case (1, 4). Where, the first number refers to the

operating point in the microgrid side, and the second number

refers to the system side shown in Fig. 4. It should be

mentioned that 10 degrees angle difference is used to find the

transient limits according to IEEE standard [38]. In practice,

according to an IEEE industrial survey [40], the maximum

closing angle can reach up to 20 degrees when

synchronization is done manually by an operator. Therefore, a

worst case angle difference of 20 degrees is used for

impedance design.

The worst (highest) transient current and torque could be

determined by simulating the four aforementioned cases.

However, to have a better insight, an analytical expression is

derived for the case of having only one synchronous

generator.

The microgrid is modelled as a synchronous generator with

a load connected to its terminal. Before the breaker is closed,

the current through the impedance is zero. However, the stator

current of the synchronous machine consists of a steady state

flow of load current as shown in Fig. 5(a). At the instant of the

first breaker switching, this switch can be represented by two

opposing voltage sources in series, as shown in Fig. 5(b). By

applying the superposition theorem, the circuit shown in Fig.

5(b) can be split into two equivalent circuits shown in Fig.

5(c).

The circuit on the left side of Fig. 5(c) represents the same

circuit before breaker is closed. The current in this circuit

represents the steady state component. The right side circuit

can be used to calculate the transient resulting from the

switching. Assuming ∆f is equal to zero, the analytical

expression for the transient component based on Kirchhoff’s

voltage law (KVL) for the circuit is

( )( ) 2 sin( )

di tL Ri t V wt

dt (1)

where, L and R are the total equivalent inductance and

resistance, respectively. is the angular frequency and is

the voltage phase difference across the breaker. The solution

to the differential equation (1) is

/

2 2

2( ) sin( ) sin( ) Rt L

eq

Vi t wt e

X R

(2)

where, eqX includes the series impedances X”d, Xinsert, and Xsys.

is the impedance angle given by tan-1

(Xeq/R).

The peak stator current is the sum of load current and the

synchronizing current, where the synchronizing current is

given by (2). It is worth mentioning that (2) is derived on the

basis ∆f=0 because the frequency difference effect on the

transient current is negligible. An explanation is realized by

considering two voltage sources connected together with

different frequencies. If one voltage source is taken as a

reference running at angular frequency 1 , the instantaneous

voltage difference when the breaker is closed at t=0 is given

by

1 1( )

1 2| | | |j w t jw tj wtV V e e V e

(3)

where, ∆ω represents the angular frequency difference

between the two voltage phasors having magnitudes V1 and V2,

respectively.

Based on the time window of subtransient, the peak inrush

current typically occurs within t < 2 cycles from the

synchronization instant. Since ∆wt is almost zero, the impact

of ∆f on ∆V across the breaker will be negligible.

B. Transient Stability Evaluation

Every time the microgrid has to be synchronized with the

main grid, transient stability is one of the main concerns. The

effect of inserting the impedance on the transient stability will

be evaluated in this section. Then, the worst case which causes

the maximum rotor angle will be identified. The ultimate goal

is to ensure that the maximum rotor angle in the worst case

does not exceed the transient stability limit of the generator.

Fig. 4. Operating regions of the microgrid and main grid

Fig. 5. Equivalent circuit representation of closing breaker 1. (a) Circuit before

breaker 1 is closed. (b) Circuit after breaker 1 is closed. (c) Equivalent circuits

after breaker 1 is closed.

Final Version

12

First, a simple case including a single generator shown in

Fig. 6 is considered. The rotor dynamics can be expressed by

the classical definition of the swing equation with damping

ignored [41] 2

2

2 ( )

( )( )

m e

syn

syn

H d tP P

w dt

d tw t w

dt

(4)

where, Pm and Pe are mechanical and electrical power,

respectively. H is the machine inertia constant. ω(t) and ωsys

are the generator and system angular frequency, respectively.

After breaker is closed, the electrical power transfer from

the internal voltage of generator to the system can be

expressed as '

'sin( )s

e

d t sys

E VP

X X X

(5)

where, E' represents the internal voltage behind direct axis

transient reactance, Vs is the voltage magnitude of the system,

and δ is the power angle between E' and Vs. The following

equation can be derived from the first equation of (4) max

max

00

2

( )m e

syn

H dP P d

w dt

(6)

For generator synchronization, since the mechanical power

is close to zero, (6) can be integrated and expressed in a closed

form to obtain the value of δmax in the first swing

1 2

max 0 1 2'cos cos( ) [2 ( )]T

s syn

HXf f

EV w

(7)

where, XT is the total series reactance.

In the case of microgrid to system synchronization, an

equivalent diagram for stability analysis is shown in Fig. 7(a).

Before the breaker is closed, the power is mainly consumed by

the load as shown in Fig. 7(b). After the breaker is closed, the

power is exchanged with the main grid, as shown in Fig. 7(c).

By modeling the loads as constant impedances and applying

Y-∆ transformations, the circuit shown in Fig. 7(c) is

transformed to the circuit shown in Fig. 8. From network

theory, the real power at node 1 of Fig. 8 is given by

*

1Re EI or

'2 '

_ 11 11 12 12cos( ) cos( )e after sP E Y E V Y (8)

where,11 10 12Y Y Y ,

10 101/Y Z , and 12 121/Y Z .

12Z is the

series impedance of the transmission network, including

transformers, lines and the inserted impedance. 10Z is the

equivalent shunt impedance connected to the machine

terminal, which includes any local loads.11 and

12 are the

impedance angle corresponding to 11Y and

12Y , respectively.

Similar to synchronization of single generator, the

theoretical δmax for the first swing when microgrid

synchronizes to system can be obtained by substituting (8) into

(6), which gives a non-linear equation

12 12 max 12

2

1 2 max

' [sin( ) sin( )]

/ [2 ( )] ( )( ) 0

s o

syn m c o

E V Y

H W f f P P

(9)

where,'2

11 11cos( )cP E Y represents the power dissipation in

the network from first term of (8).

To consider the transient stability in designing the

impedance, first the worst case should be identified. Then

stability of the system under the worst case should be ensured.

The effect of increasing the loading level on the transient

stability is shown in Fig. 9. As it can be seen, increasing the

loading level of microgrid reduces the stability margin, where

the margin is defined as (Pmax-Pm)/Pmax. The reason is that the

initial angle of the rotor will be higher as the load is increased.

The effect of inserted impedance on the transient stability is

shown in Fig. 10. As it was expected, the inserted impedance

will reduce Pmax and consequently reduce the transient stability

margin. As it will be seen in the case study, since the

frequency difference between the two grids is small, the

inserted impedance will only have a small impact on the

maximum rotor angle.

To sum up, the worst case that should be examined in terms

of maximum rotor angle is based on the peak loading level and

the designed impedance value. In addition, the voltage levels

(microgrid and system sides) should be at the lowest of the

Fig. 6. Generator to system synchronization example

Fig. 7. Circuits for transient stability analysis (a) microgrid to system

synchronization single line diagram. (b) Before closing breaker (c) After

breaker is closed.

Fig. 8. Simplified equivalent circuit of the microgrid after applying Y-∆

transformations

Final Version

13

power quality limits according to (6) and (8). The frequency

difference should be positive (i.e. higher microgrid frequency

than system) to result in higher rotor angle within the first

swing. As a result, the worst case open loop synchronization

scenario from a transient stability point of view is case (1, 3)

shown in Fig. 4.

C. Second Switching Surges

Similar to the first breaker switching, impedance bypass

also produces a transient effect on the system which must be

evaluated to ensure an acceptable disturbance level. According

to the principal of superposition, bypass switching can be

represented by two opposing voltage sources as shown in Fig.

11, which is equivalent to a single voltage source in the

transient circuit. This circuit is quite similar to the first

switching circuit, but with ∆f=0. The voltage difference across

the impedance before the bypass depends on both the

impedance value and the amount of current flowing through

the impedance.

The active and reactive power flowing through the

impedance after closing the first breaker depend on the

governor and AVR settings of the synchronous generator. The

reason is that due to the nature of open loop synchronization,

the voltage set-point of AVR is not adjusted during the

process of synchronization. Therefore, the AVR is regulating

the terminal voltage to the level prior to synchronization. As a

result a reactive power flow is expected which can flow in

either direction depending on the voltage levels of the

synchronous generator and the main grid.

Active power flow through the inserted impedance is also

expected. The speed governors of the synchronous generators

are operating in droop control mode to assure proper load

sharing between the DG units. When the system reaches the

steady-state before impedance bypass, the microgrid

frequency becomes equal to the system frequency. The

governor responds to the change in frequency by a percentage

change in the mechanical torque or output power [41] as in

(10). Therefore, a real power exchange between the microgrid

and the main grid exists which can be in either direction

depending on the pre-synchronization frequencies of the both

sides.

(%) 1( )

microgrid sys

m

nom

f ffP pu

R f R

(10)

where, R is the droop constant and fnom is the rated machine

frequency.

The larger the frequency and voltage differences prior to

synchronization, the larger the power flowing through the

impedance. The worst case caused by the highest voltage

difference should be identified to make sure that the worst

possible transient is lower than the acceptable limit. Possible

worst case bypass voltages can be identified as case (2, 3),

case (4, 1), case (3, 2) and case (1, 4) shown in Fig. 4. The

voltage across the impedance can be found through load flow

studies. Then the highest voltage difference is compared to the

acceptable voltage limit. If the maximum voltage is lower than

the limit, the bypass is realized in one step. Otherwise, more

bypass steps are required. The voltage difference limit is

0.1855 p.u. based on standard IEEE C50.12 [38].

D. Impedance Design Summary

The design method for the pre-insertion impedance size can

be summarized as follows:

- The acceptable transient current and torque limits are

determined.

- The four possible worst cases identified before are

established to find the case with the worst transient.

- The worst case is then used to determine the minimum

impedance size through analytical approach or

Fig. 9. Impact of load to microgrid synchronization

Fig. 10. Impact of impedance pre-insertion to microgrid synchronization

Fig. 11. Representing impedance bypass by equivalent voltage source

Final Version

14

simulation.

- Analytical approach or simulation is then used to find

the maximum acceptable impedance to maintain the

transient stability of the generators.

- An acceptable impedance range is now developed

according to transient current and transient stability

limits. This impedance size is selected in the lower

range in this paper to have the maximum transient

stability margin.

- Power flow studies are utilized to find the case causing

the highest voltage across the inserted impedance

among the four cases identified before. The highest ∆V

is then compared to the maximum acceptable value to

find how many steps are required to bypass the inserted

impedance.

IX. CASE STUDY

To evaluate the effectiveness of the method, it is applied to

a system shown in Fig. 12. The system consists of one 6.6

MVA synchronous generator connected through a 30 mile

(25kV) feeder to the main substation. A total load of 6MW is

distributed along the feeder as shown in the figure. The system

and synchronous generator data can be found in Appendix.

The impedance pre-inserted at the circuit breaker location is

considered to be purely inductive.

The system is simulated in MATLAB/Simulink. Based on

EMTP simulations, maximum acceptable transient current and

torque are found to be 1.5 p.u. and 1.7 p.u., respectively. The

respective simulation results are shown in Fig. 13.

A. Impedance Design

The 4 possible worst cases for the first switching discussed

in Section III are simulated. The Peak current and torque for

these cases are presented in Fig. 14. As it can be seen, the

worst transients occur in case (2, 3). Now that the worst case

is identified, the minimum size of impedance required to limit

the transients can be found either by EMTP simulations or by

the analytical expression given in (2). A comparison of the

two approaches is shown in Fig. 15. The analytical method

provides an upper bound for the peak stator current. For this

case study, the impedance value determined analytically is 0.6

p.u. or 56.82 Ω.

For transient stability analysis, the worst case occurs in case

(1, 3). Similar to the previous step, either EMTP simulations

or analytical approach could be used to determine the

predicted maximum rotor angle. The theoretical prediction of

the maximum rotor angle is obtained by simplifying the circuit

in Fig. 12 to Fig. 8 using Y-∆ transformations. The maximum

swing angle can then be calculated by (9). The maximum rotor

angles found by simulation and analytical approaches for

different impedance sizes are compared in Fig. 16. It can be

seen that the results of the two methods are close to each

other. In addition, increasing the inserted impedance size does

not have a significant effect on the maximum rotor angle.

According to the maximum acceptable transient level and

stability limit, the feasible impedance range is within 0.5p.u.-

1.1p.u. However, to ensure a high stability margin, 0.6pu is

chosen as the required impedance size.

The last design step is to evaluate the severity of bypass

transients. Load flow studies are used to find the worst bypass

voltage among the four possible cases mentioned in Section III

and the results are presented in Fig. 17. As it can be seen, case

Fig. 12. Single line diagram of the case study

Fig. 13. Time domain simulation to obtain acceptable transient limits

Fig. 14. Worst case transients based on power quality limits

Fig. 15. Peak current versus impedance size for the worst case

2.01 2.00 1.90 1.93

2.34 2.332.22 2.26

0.00

0.50

1.00

1.50

2.00

2.50

3.00

case (2, 3) case (4, 1) case (3, 2) case (1, 4)

Pea

k S

tato

r C

urr

ent

an

d

To

rqu

e (

pu

)

I_peak Te_max

Worst Case transient

Final Version

15

(4, 1) has the highest complex power which leads to the

highest voltage across the impedance. For the selected

impedance size (0.6p.u.), this complex power incurs a voltage

difference of 0.178 p.u., which is lower than the acceptable

level of 0.1855p.u. Therefore, the impedance can be bypassed

in one step while the transient level is acceptable.

B. Simulation Results

To better show the performance of the proposed method,

the whole synchronization process using the designed

impedance is shown in Fig. 18 to Fig. 20. For these

simulations it is assumed that the synchronization is done

under case (2, 3) which has the worst transient at the

synchronization instance. The voltage angle difference across

the breaker is assumed to be 20 degrees. For comparison, the

synchronization under the same condition but without the

inserted impedance is also shown in these figures.

As it can be seen in Fig. 18 and Fig. 19, without the inserted

impedance the current and torque could reach as high as 2 p.u.

and 2.34 p.u., respectively, which are higher than the

acceptable limits. After inserting the impedance, the peak

current for the first and second switching are 1.37 p.u. and

1.38 p.u. which are within the acceptable limit. Also the peak

torque for the first and second switching are 1.54 p.u. and 1.36

p.u., respectively, which are below the maximum acceptable

limit.

Comparing the rotor angles in Fig. 20 shows that although

inserting the impedance increases the oscillations, the

oscillations will not causing the rotor to pass the maximum

stability limit and the generator will be stable after the

synchronization.

X. CONCLUSION

An impedance insertion method has been proposed in this

paper to enable open-loop synchronization of a microgrid.

Since the primary concern in synchronization is the switching

transients, the idea is to pre-insert an impedance before the

breaker closing to reduce the transients to an acceptable level.

The impedance is then bypassed using a second breaker. The

necessary requirements are considered in the impedance

design to ensure acceptable transient level and transient

stability. Using the open-loop scheme, both synchronization

Fig. 16. Feasible impedance range base d on short circuit and stability analysis

Fig. 17. Worst possible power flows through the impedance from load flow

studies

Fig. 18. Stator current during synchronization with and without impedance

insertion

Fig. 19. Torque during synchronization with and without impedance

insertion

Fig. 20. Rotor angle during synchronization with and without impedance

Final Version

16

cost and effort is reduced because feedback control is not

necessary. The proposed scheme has been evaluated through a

case study. The results demonstrated that the transient levels

can be effectively reduced and the microgrid can be

successfully synchronized in an open-loop manner. This

scheme presents an attractive solution to synchronizing remote

microgrids.

APPENDIX

TABLE I

THEVENIN EQUIVALENT (SUB) DATA [30]

Short circuit power (MVA) 346

Nominal voltage (kV) 25

X/R Ratio 7

TABLE II

SYNCHRONOUS GENERATOR DATA [30]

Nominal Power (MVA) 6.6

Nominal Voltage (kV) 4.16

Pair of poles 2

Inertia constant (s) 2.52

Stator resistance (pu) 0.00

𝑋𝑑(pu) 1.26

𝑋𝑑′ (pu) 0.26

𝑋𝑑′′ (pu) 0.18

𝑋𝑞(pu) 1.24

𝑋𝑞′′ (pu) 0.21

𝑇𝑑𝑜′ (pu) 1.4446

𝑇𝑑𝑜′′ (pu) 0.0218

𝑇𝑞𝑜′′ (pu) 0.0690

TABLE III

EXCITATION AND GOVERNOR-TURBINE DATA [41]

AVR IEEE

Type 1

Governor- Turbine

System

Type: Hydro

𝑇𝑟 5ms 𝑅𝑝 0.05

𝐾𝑎 300 𝑇𝑔 0.2s

𝑇𝑎 50ms 𝑇𝑤 1.0s

𝐾𝐸 1.0 𝑅𝑡 0.38

𝑇𝐸 0.65s 𝑇𝑟 5.0s

𝐾𝑓 0.048

𝑇𝑓 0.95s

𝐸𝑓𝑚𝑖𝑛 -5

𝐸𝑓𝑚𝑎𝑥 8

ACKNOWLEDGMENT

The authors wish to thank the supports from the NSERC

Smart Microgrid Research Network (NSMG-Net) during the

course of this project.

REFERENCES

[22] A. Bidram and A. Davoudi, “Hierarchical Structure of Microgrids

Control System,” IEEE Trans. Smart Grid, vol. 3, no. 4, pp. 1963-1976, Dec. 2012.

[23] M. Savaghebi, A. Jalilian, J. C. Vasquez, and J. M. Guerrero,

“Secondary Control Scheme for Voltage Unbalance Compensation in an Islanded Droop-Controlled Microgrid,” IEEE Trans. Smart Grid, vol. 3,

no. 2, pp. 797-807, Jun. 2012

[24] M. S. Golsorkhi Esfahani and D. D. C. Lu, “A Decentralized Control

Method for Islanded Microgrids under Unbalanced Conditions,” IEEE

Trans. Power Del., doi: 10.1109/TPWRD.2015.2453251.

[25] P. Sreekumar and V. Khadkikar, “A New Virtual Harmonic Impedance

Scheme for Harmonic Power Sharing in an Islanded Microgrid,” IEEE Trans. Power Del., doi: 10.1109/TPWRD.2015.2402434.

[26] N. W. A. Lidula and A. D. Rajapakse, “Microgrids research: A review

of experimental microgrids and test systems,” Renewable Sustainable Energy Rev., vol. 15, no. 1, pp. 186-202, Jan. 2011.

[27] L. C. Gross, L. S. Anderson, and R. C. Young, “Avoid Generator and

System Damage Due to a Slow Synchronizing Breaker,” proceedings of the 24th Annual Western protective Relay Conference, Spokane, WA,

Oct. 1997.

[28] IEEE Guide for AC Generator Protection, IEEE standard C37.102-2006. [29] M. J. Thompson, “Fundamentals and Advancements in Generator

Synchronizing Systems,” 65th Annual Conference for Protective Relay

Engineers, pp. 203-214, Apr. 2012. [30] T. M. L. Assis, G. N. Taranto, “Automatic Reconnection From

Intentional Islanding Based on Remote Sensing of Voltage and

Frequency Signals,” IEEE Trans. Smart Grid, vol. 3, no. 4, pp. 1877-1884, Dec. 2012.

[31] C. Cho, J. Jeon, J. Kim, S. Kwon, K. Park, and S. Kim, “Active

Synchronizing Control of a Microgrid,” IEEE Trans. Power Electron, vol. 26, no. 12, pp. 3707-3719, Dec. 2011.

[32] F. Tang, J. M. Guerrero, J. C. Vasquez, D. Wu, and L. Meng,

“Distributed Active Synchronization Strategy for Micro-grid Seamless Reconnection to the Grid Under Unbalance and Harmonic Distortion,”

IEEE Trans. Smart Grid, vol. 6, no. 6, pp. 2757-2769, Nov. 2015 [33] H. Laaksonen, D. Ishchenko, and A. Oudalov, “Adaptive Protection and

Micro-grid Control Design for Hailuoto Island,” IEEE Trans. Smart

Grid, vol. 5, no. 3, pp. 1486- 1493, May. 2014 [34] K. Malmedal, P. K. Sen, and J. P. Nelson, “Application of out-of-step

relaying for small generators in distributed generation,” IEEE Trans.

Ind. Appl., vol. 41, no. 6, pp. 1506-1514, Nov./Dec. 2005 [35] R. C. Dugan, M. F. McGranaghan, H. W. Beaty, “Electrical Power

Systems Quality”, in McGraw-Hill, 1996, pp. 100-103.

[36] Technical Requirements for Connecting to the Alberta Interconnected Electric System (IES) Transmission System, ESBI, Alberta, 1999.

[37] A. J. Rodolakis, “A comparison of North American (ANSI) and

European (IEC) fault calculation guidelines.” IEEE Trans. Ind. Appl., vol. 29, no. 3, pp. 515-521, May/Jun. 1993.

[38] IEEE standard for Salient-Pole 50 Hz and 60 Hz Synchronous

Generators and Generator/Motors for Hydraulic Turbine Applications Rated 5 MVA and Above, IEEE Standard C50.12-2005.

[39] N. Mohan, "Electric Power Systems: A First Course", in John Wiley &

Sons Inc., 2012, pp. 159-160. [40] W. M. Strang, C. J. Mozina, B. Beckwith, T. R. Beckwith, S. Chhak, E.

C. Fennell, E. W. Kalkstein, K. C. Kozminski, A. C. Pierce, P. W.

Powell, D. W. Smaha, J. T. Uchiyama, S. M. Usman, and W. P. Waudby, “Generator synchronizing, industry survey results,” IEEE

Trans. Power Del., vol. 11, no. 1, pp. 174-183, Jan. 1996.

[41] P. Kundur, “Power System Stability and Control”, in McGraw-Hill, 1994, pp. 581-600.