Page 1 of 10

Code: _D.5_______

Distributed Synchronous Coordination Field Testing Of an Actual Automated Distribution

Feeder System

[Naibo Ji, OMICRON electronics Asia Limited, +852 3767 5521, noble.ji@omicronenergy.com]

[Stephan Geiger, OMICRON electronics GmbH, Austria]

[Overview] A Distribution Feeder Automation system deployed at Wake Electric using IEC61850 “GOOSE” over a

WiMAX wireless communication provided the real world test bed to prove a new testing method and proc

edure. The Wake system consists of 7 reclosers connected in a mesh network supplied from three differe

nt substations. The system typically performs automation sequences in less than 0.5 seconds.

This paper will discuss how this complex high-speed feeder automation and associated communication syst

ems were tested to prove correct and predictable protection isolation and restoration sequences. The testing

was first performed in a laboratory environment to verify test protocols, followed by the actual field test.

For comprehensive testing, the test equipment was deployed at all 7 recloser locations, and simulated fault

s were applied on different line sections. The test equipment operated synchronously to simulate a real-life

event at all locations as would be the case when real faults were to occur. All test sets used to perform t

he required transient current and voltage injections were GPS synchronized and controlled through the inst

alled WiMAX system from a single PC located at one of the recloser locations. This paper will in additio

n provide a comparison of a test performed to a real event that occurred on the system.

[Key Word] RECLOSER CONTROL, DISTRIBUTION RESTORATION, IEC61850, NETWORK SIMULATION

[1. Introduction]

Testing a standalone protection device for its protection functions and operating times has become

commonplace due to the availability of accurate test equipment. But testing wide area automation systems

consisting of several devices is a very difficult task since many factors need to be tested including

communication between the devices, dynamically changing power system parameters etc. One of the prominent

wide area automation systems is the high speed feeder automation system, whose main goal is to locate the fault,

isolate a faulted line section and restore power to the unaffected line sections as fast as possible. This system is

often called on to perform in stormy, chaotic conditions where it is difficult to predict the behavior of a power

system due to dynamic changes in power system parameters at different points on the feeder. In addition,

healthy communication between the devices is very crucial at all times for correct operation. These

communication networks have several limits in terms of bandwidth and latency which are affected by weather

conditions and also real time changes in power system parameters. All of these factors must be taken into

consideration when designing and testing a feeder automation system.

To achieve good performance, these systems must be thoroughly tested. Some of the challenges of testing

include:

Synchronous injection of all devices in the system

Collection of data from all the test units in the system and analysis of system performance

Simulation of adverse scenarios affecting the communication network

The latest testing tools have advanced simulation software which can be used to address the items listed above.

The remainder of this paper will provide details of how these tools were utilized to test an actual in service

feeder automation system.

Page 2 of 10

[2. Wake Electric High Speed Feeder Automation Project]

Wake Electric Cooperative, Youngsville, NC chose to automate a mesh connected system which includes three

substations and four midline reclosers. The system was deployed using a WiMAX communication network. All

WiMAX base stations were installed on a radio station tower (WCPE) at an appropriate height to allow good

coverage of all required locations. The topology of the system is represented in the figure below.

P1 P2 P3 P4 P5

P6 P7

WCPETOWERLSA

LSB

LSC LSD

LSE

Heritage Sub

Walkers Sub

98E Sub

Figure 1: Wake Feeder Automation System Topology

Two critical automation sequences performed by this system were selected for synchronous testing.

Sequence 1: Source Transfer (When both Heritage and Walkers substation are lost):

Figure 2: Automatic Sequence resulting from loss of Sources HE5 and WX3

As shown in the above figure, the source transfer sequence steps include:

1. Heritage (HE5) and Walkers (WX3) Substation Reclosers (P1 and P6) are opened

2. Normally Open point Recloser P3 is closed to feed line sections LSA and LSB from 98East (98E)

Substation

3. Normally Open point Recloser P7 is closed to feed line section LSE from 98East (98E) Substation

Sequence 2: Fault on line section between heritage substation and heritage line recloser

In order to minimize the outage time on the feeder serving the WCPE radio tower as well as other customers,

Wake EMC decided to implement the new protection philosophy utilizing a differential principal. In the event of

P1 P2 P3 P4 P5

P6 P7

WCPETOWERLSA

LSB

LSC LSD

LSE

HE5

WX3

98E

P1 P2 P3 P4 P5

P6 P7

WCPETOWERLSA

LSB

LSC LSD

LSE

HE5

WX3

98E

P1 P2 P3 P4 P5

P6 P7

WCPETOWERLSA

LSB

LSC LSD

LSE

HE5

WX3

98E

1

2

P1 P2 P3 P4 P5

P6 P7

WCPETOWERLSA

LSB

LSC LSD

LSE

HE5

WX3

98E3

Page 3 of 10

a fault, differential protection should pinpoint the fault and isolate the faulted line section from both ends.

Following this operation, the recloser at the normally open point should close to provide service to the unfaulted

line section from another source. This approach helps to restore service on unaffected line segments even before

the first autoreclose operation is performed on the faulted feeder section. This fault isolation and reconfiguration

sequence takes 350 - 450 ms, where conventional protection system with time coordinated curves might take

seconds or even minutes performing a number of autoreclose operations and further steps to reestablish the

service on the rest of the feeder.

Along with fast operating speed, the differential scheme is known for its selectivity. The differential zone can be

applied on each feeder section individually without a need for time coordination between individual reclosers.

To isolate a fault inside a zone, each recloser evaluates the information from upstream and downstream reclosers.

A true differential scheme typically requires the comparison of real time currents. In order to minimize the

amount of data, reclosers convert magnitude changes in the phase currents to logical signals called “Positive

jump” and “Negative jump”. (hence the name of the algorithm “Jump Differential Protection” or jDiffTM). These

signals can be transmitted through the network to the reclosers at the opposite side of the protected zone with

less bandwidth requirements. In the event of fault, system performs the steps shown in figure 3.

P1 P2 P3 P4 P5

P6 P7

WCPETOWERLSA

LSB

LSC LSD

LSE

HE5

WX3

98E

P1 P2 P3 P4 P5

P6 P7

WCPETOWERLSA

LSB

LSC LSD

LSE

HE5

WX3

98E

P1 P2 P3 P4 P5

P6 P7

WCPETOWERLSA

LSB

LSC LSD

LSE

HE5

WX3

98E

P1 P2 P3 P4 P5

P6 P7

WCPETOWERLSA

LSB

LSC LSD

LSE

HE5

WX3

98E

1

2

3

Figure 3: Automatic Sequence resulting from a Fault on line section between P1 and P2 (LSA)

As shown in the above figure, the automatic sequence steps for a LSA fault include:

1. Heritage (HE5) Substation recloser is opened to clear the fault

2. Recloser P2 is opened to isolate faulted line section A from rest of the system

3. Normally Open point P7 is closed to feed line section B (LSB) from Walkers (WX3) Substation

4. Recloser P1 performs its auto reclose attempts to check if the fault is cleared. In the event of a temporary

fault, P1 stays closed after the reclose attempt. In the event of permanent fault, P1 exercises all its reclose

attempts, opens, and goes to lockout

[3. Synchronous Lab Testing]

As noted previously, the test units were setup in a lab environment to verify test protocols and performance prior

to field testing. The steps performed are as follows:

Page 4 of 10

All communication devices are programmed and configured in the lab according to the network design

All Feeder Automation controllers are programmed and mounted in their respective recloser, switch or

breaker cabinets as necessary to simulate the original feeder topology

Required test sequences are configured in test software and loaded into the test set. These sequences are

run to make sure desired voltages and currents are injected into the respective devices to simulate the

faulted conditions.

The system undergoes a thorough testing for faults on all line sections, as well as loss of source and load

balance conditions. Each test has level assessment and time assessment steps to indicate pass or fail

results. These assessments determine correctness and accuracy of the system.

Software based HMI is tested to verify all commands and indications function properly.

All the devices are GPS synchronized.

3.1 Lab Test Setup

The Wake Electric Feeder Automation system is replicated in the lab over a WiMAX communication network.

Each protection point is injected by a test unit which injects currents and voltages into the Feeder Automation

Controllers. The primary contact operations are also simulated by the test units. All test units communicate with

each other over the existing WiMAX network in the lab to simulate field operation. A PC running the

RelaySimTest software is also given access to this network to be able to communicate to all test units. Individual

GPS antennas are used with all the test units to synchronize them.

To test the correct sequence of events for an automated distribution feeder e.g. from trip to reconfigure, it is

important to have a realistic and consistent simulation of currents and voltages according to the breaker states.

To achieve this, power system data provided by Wake EMC was entered into the test set software to provide

simulation of different fault scenarios.

In case of a short circuit incident the controller will issue a trip. If the fault current does not extinguish after the

circuit breaker trip time, the controller will assess it as a circuit breaker failure and will run a different sequence.

Historically, to achieve a simulation in which the current is extinguished would require state sequences to be

predefined, which was complex and error prone. Alternatively, a real time simulation system had to be used,

which is costly, inappropriate for field tests and won’t work for a distributed system. To provide a system which

could overcome the short comings of a real time system while including its strengths, the simulation software

used an iterative closed loop approach.

Figure 4 shows the first two iterations of testing a permanent fault on line section A between devices P1 and P2.

The test case is already defined by only placing a fault on line segment A. The first iteration simulates a

permanent fault. As the software detected a trip of P1 it issued a second iteration. Assuming that the P1

controller trips consistently when simulating the same fault quantities, the software creates a circuit breaker

event at the time of the last trip and includes additional time for circuit breaker operation. Following the

previous iteration, the software would record a trip from the P2 controller triggering the next iteration and so

forth until the last iteration does not record any further events of the automated feeder system. In other words it

is a step wise approximation of a real-time simulation.

Page 5 of 10

First Iteration

Second Iteration

Trip requiring a second iteration

Figure 4: First two software iterations for a fault on line section A

3.2 Lab Results

Two critical sequences for this system were considered for synchronous testing.

The first sequence is the source transfer when both Walkers and Heritage Feeds are lost. The automatic sequence

resulting from this source’s loss includes operation of 4 devices in a way that all the line sections are supplied

power from the third source (98 East Substation). Hence RelaySimTest software performs 4 iterations to test and

record the entire event:

Source breakers P1 and P6 are opened

Normally open point configuration devices P3 and P7 are closed to feed all the line sections from the

98East substation.

As shown in figure 5, the total transfer time is 660 ms. GPS synchronization over WiMAX in the lab was

inconsistent due to availability of only one GPS transmitter in the lab and closely placed high gain GPS

antennas. This resulted in additional delays in the closed loop iterative process.

Figure 5: Software iterations showing automatic source transfer sequence result

Page 6 of 10

Three reclose attempts till lockout

Reconfiguration

IsolateTrip

Figure 6: Software iterations showing automatic sequence for a permanent fault on line section A

The second sequence shown in Figure 6 is a permanent fault on line section A between devices P1 and P2. The

automatic sequence resulting from this fault includes 3 automation operations and 4 protection operations (3

reclose shots from breaker P1). Hence RelaySimTest software performs 7 iterations to record the entire event:

Source breaker P1 opens after jDiffTM determines the fault

Device P2 opens to isolate the fault

Device P7 closes to feed line section B from Walker’s substation

Device P1 performs 3 autoreclose attempts before it opens and locks out.

S1 Trip + CB = 294ms

S2 Isolate + CB = 456ms

S7 Reconfiguration + CB = 665ms

Figure 7: Software Iterations showing fault Isolation and reconfiguration times

From the above figure, the breakdown in times for each step with reference to fault inception is as follows

Device P1 took 294 ms to trip

Device P2 opened after 456 ms to isolate the faulted line section A

Device P7 closed after 665 ms to reconfigure the system feeding line section B from Walkers Substation

Page 7 of 10

[4. Synchronous Field Testing]

4.1 Field Test Setup

In the field, the protection points are located a few miles away from each other. The field test setup for

synchronous testing is shown on figure 8.

Figure 8: Field Test Setup

Prior to the test the following field actions were performed:

The testing team assembled field test kits for all locations with additional hardware to streamline

installation. Each kit included:

Two sets of wired connectors between the automation controller and the test unit to simulate

primary contact operation and provide feedback

Pre-wired current and voltage plugs for local injection from the test set

An unmanaged switch at each location to connect into the existing WiMAX network

By passing all primary switchgear at the 7 protection points

Connecting each test unit to a GPS antenna

The PC running the RelaySimTest software was connected to the WiMAX network at the Heritage

substation. The test software accessed all GPS synchronized test units at each location using WiMAX

communications.

The same two sequences previously completed in the lab are tested in the field. Several iterations are

performed and recorded by the software to analyze the system performance.

Simultaneously, fault records and event data were downloaded from the devices that were exercised in

each test sequence.

Page 8 of 10

4.2 Field Test Results recorded by the Test Software

S1 Trip + CB = 298ms

S2 Isolate + CB = 410ms

S7 Restore + CB = 563ms

Figure 9: Software Iterations showing field results of automatic sequence for fault on line section A

From the above figure, the breakdown in times for each step with reference to the fault inception is as follows

Device P1 took 298 ms to trip

Device P2 opened after 410 ms to isolate the faulted line section A

Device P7 closed after 563 ms to reconfigure the system feeding line section B from Walkers Substation

Figure 10: Software Iterations showing field results of automatic source transfer sequence

From the above figure, the total transfer time from the source loss to the close of normally open points P3 and

P7. The total transfer time is 488.92 ms.

Page 9 of 10

4.3 Comparison of the system performance during the Field Test with a Real time event

Several months after the installation a severe storm caused outage on the transmission line interrupting the

service to Heritage and Walkers substations. The system detected loss of both substations and transferred the

load over to the available substation 98E. The fault record from this transfer event is shown in figure 11, where:

1. Voltage at Heritage Substation dropped below the threshold

2. Voltage at Walkers Substation dropped below the threshold

3. Undervoltage element picked up at Heritage Substation

4. Undervoltage element picked up at Walkers Substation

5. Heritage Recloser (P1) is opened;

6. Walkers Recloser (P6) is opened;

7. Midline Recloser (P3) is closed, service to the major part of the system is restored (495 ms from source loss);

8. Walkers Line Recloser (P7) is closed, service to the line section E is restored (663 ms from source loss);

Analyzing the real transfer event and comparing it with the software field test results from figure 8, it was

concluded that:

1. The system executed source transfer sequences correctly

2. The operation was as fast as the primary gear allowed

3. 234 ms delay and status contact chatter was caused by old VWE reclosers at Walkers 3. Software test result

from figure 10 doesn’t include this delay because the primary contacts were simulated by the test unit.

4. Operational sequence execution is fast enough to eliminate cold load scenario at 98 East Substation.

5. Sequential operation time per switching step is 158 ms. This includes additional WiMAX communication

delays during the storm which is not the case when the synchronous field testing was performed.

IA IB IC

VA VB VC

t/s

0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8

Source HOTRecloser Closed

IA IB IC

VA VB VC

t/s

IA IB IC

VA VB VC

t/s

0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8

0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8

Recloser ClosedSource HOT

P1 - Heritage

Sub Recloser

Current

Voltage

Current

Voltage

P3 - Midline

Recloser (NOP)

Current

Voltage

P6 - Walkers

Sub Recloser

500A

7.5kV

750A

8kV

7.5kV

200A

3

5

2

4 6

7

8

1

S1 OFF + P1 OFF + P3 ON = 495ms

S1 OFF + P1 OFF + P3 ON +P7 ON = 663ms

Recloser Closed

Figure 11: Fault Record from real source transfer event

Page 10 of 10

[5. Conclusion]

5.1 Advantages

It was possible to perform a distributed synchronous field test on an automated feeder system by

injecting all the system devices simultaneously.

This approach will provide the ability to test a wide area system using various scenarios.

These tests will help identify the problems or concerns for a high speed automation system during

adverse weather conditions where the power system parameters, including currents and voltages, change

rapidly.

By injecting all the devices in the field simultaneously, all controllers generate a significant amount of

GOOSE messages which provides a good test for the communication network.

The test units, being GPS synchronized help analyze the system performance accurately.

The real time event sequence and speed of operation correlates with the test results

5.2 Shortcomings

Primary switchgear operating time was simulated instead of expressly monitored.

Any network issues which cause loss of connection to any test unit during the test resulted in the test

software aborting the iteration with a warning.

[6. About the Author]

Naibo JI

BEng (Hons), BBA, MSc, MIET, MIEEE, MCIGRE, MCSEE, MIEEJ

Mr. Naibo Ji was born in 1985 in China. He received his bachelor's degree in Electrical

Engineering at the Hong Kong Polytechnic University in 2011. He joined OMICRON

electronics in 2011 where he worked as Regional Application Specialist. More than 400

protection engineers and technicians from Asia Pacific have been directly trained by Noble

in the past 50 trainings. Noble also attended and presented technical papers at international

conferences like CIRED, DPSP, CEPSI.

Stephan Geiger

Mr. Stephan Geiger graduated from the Technical College of Bregenz (Austria) with a focus

on Electrical Engineering in 2007. He accepted a job with OMICRON electronics GmbH in

Austria in 2008 as an Application Engineer for protective relay applications, where he

worked out application notes and custom applications, provided trainings and later took over

project management responsibilities. Since 2012 he is Product Manager and responsible for

OMICRON's recloser control testing solutions.