Engineering, 2013, 5, 7-12
doi:10.4236/eng.2013.51b002 Published Online January 2013 (http://www.SciRP . org/journal/eng)
Development of Fault Management Dispatcher Training
Simulator for TDAS in Korea
In-Yong Seo, Sang-Ok Kim, Bok-Nam Ha
Smart Energy Lab, KEPCO R esearch In s titute, Munjiro, Yuseong , D a e j e on, Korea
Received 2013
ABSTRACT
A fault ma nage men t dispa tcher training s i mulator for lar ge-s cale Distribut ion Auto mation Syste m (TDAS) is deve lop ed
to train ope rators in distribution control ce nter. T his simulato r is composed of independent simulatio n server and o per-
ator consoles and can be used for networ k anal ysis, net wor k op eratio n, faul t ma nageme nt and eval uatio n. TD AS DB i s
duplicated online to the simulation server keeping the data security. The system can model distribution network pene-
trated with distributed generat ions (DG) using the real d ata from the TDAS DB. Network fault scenarios are automati-
cally generated by calculating fault current and generating fault indicators. Also, manual entry of cry wolf alarm is
available. Mo reover, op eration solution for scenario of fault isolation and ser vice restoration is generated automatically
so that trainee can check their operation result. Operator actions during training session are saved and can be played
back as well as displayed on one-line dia gram pictures.
Keywords: Distribution Automatio n System; Fault Ma nagement Dispatcher T raining Simulator; Distributed
Generation
1. Introduction
With high penetration of distribution generation (DG),
power system operations are becoming increasingly
complicated and critical. There is a need to provide more
directed power system operation training than that
provided by on-the-job experience. Dispatcher training
simulator (DTS) can realistically simulate electricity
system in the actual production, so that the trainees are
familiar with the system, to develop its capability of
handling vari ous incidents.
The handling of network failures (e.g. short circuits,
earth faults) offers a lot of possibilities for automation.
The main tasks are fault location isolation and service
restoratio n (FISR). The network operator needs excellent
knowledge about the controlled network area to perform
these tasks efficiently. Without efficient aids by a u t oma-
tion, the operator has to achieve and to keep these skills
by spending a lot of time with a dispatcher training
simulator.
Lots of research have been concentrated on DTS re-
lated to Energy Management System (EMS) and Super-
visor y Control and Data Acquisition (SCADA) [1-3] but
not many papers for fault management DTS (FM-DTS)
were reported [4,5]. In this paper a fault management
system for DG interconnected distribution networks is
described which automates the fault localization and the
determinatio n of isolation and restoration measures to
relief the operator from these demanding tasks during
network failures.
KEP CO has been operating 186 TDAS for urban areas.
As can be seen in Figure 1, it has a dual server, three
human–machine interface (HMI) terminals, and a front-
end processor (FEP). It adopted DNP 3.0 protocol over
fiber optic cable, dedicated metallic line, wireless data
packet, trunked rad io system, etc.
Figure 1. C e ntral control system of TDAS.

I.-Y. SEO ET AL.
8
2. Fault Management DTS
2.1. Configurations and Features
The developed FM-DTS shown in Figure 2 has three
main sub systems:
Simulat i on Ser ver : This server saves coded TDAS DB ,
scenario, solutio n of FIS R, trai nee’s trai n histor y, traine e’s
switch operation actions, user information and score
statistics.
Simulation Console: This console uses simulation
software and displays for generating fault scenarios with
which users can perform self-training. Several application
pro grams ar e runni ng in the cons ole whic h per for m short
circuit analysis and power flow analysis, and generating
fault event and solution of FISR. Also, it converts fault
scenarios to the format for competition console.
Competition Console: This console is used for practical
test and written test. All operator’s actions are logged so
that instructors can review their mistakes during the
practical tes t. W rit te n te st progr a m is an inter net b a sed te st
(IBT) that includes short-answer question and multiple
choice about operation of distribution networks. Test
results for both tests are a u tomatically scored and analyzed
for statistics by the system, and transferred to the simulation
server.
This simulator, from software implementation point of
view co mprises of the following modules
a) DB Manager
b) Simulator Manager
c) Scenario Manager
d) Application P rogra ms
e) Operation History Manager
f) HMI (NZed, agO LD, AgWorks, Host-server)
2.2. Online Duplication of DB for Simulation
TDAS is one of most important national security facili-
ties in Korea. TDAS DB should be secure in any case but
real DB is needed for training the service restoration of
distribution networks. T he DB is duplic ated in two ways,
i.e. real time and periodic req uest by the DB handler.
Figure 3 shows configuration of DB online duplica-
tion, and the procedures for duplication is described in
the followin g:
Step 1. Make a backup DB and checksum file in OA
link server
Step 2. Encode the backup DB
Step 3. Copy and save the encoded DB and checksum
file to simulation server
Step 4. Copy and decode the encoded DB and check-
sum file to simula tion console
Step 5. Generate a checksum file for the decoded DB
in simulatio n console.
Step 6. Compare the c hecks um fil es i n step 1 and 5 . If
those files are identical, we can assure integrity of
T DAS.
Step 7. Delete all DB data in simulation console at the
end of simulation program automatically.
2.3. Simulator Manager
Simulator manager is installed in the simulation console
and has fo llowi ng fu nct ions;
- Execution o f host-server and HMI
- Creation of distr ibution li ne fault sce nar ios
- Creation of s ub station fault sc e nar ios
- Management of scenarios
- Inquiry of train history
- Management of user access
- Close of simulation program
Figure 2. Configuration of fault management DTS.

I.-Y. SEO ET AL.
9
Figure 3. Configur a tion of DB online copy.
Figure 4 . DB handler prog ram.
Figure 5. Simulator manag er user interf ac e.
2.4. Scenario Manager
Scenario manager produces fault scenarios and cry wolf
alarm shown in Figure 6 and 7 respectively. The fault
types for scenarios are 1 phase ground, 2 phase ground, 3
phase short. Fault location can be designated using the
switc h and d istance from t he swit ch. Al so, this can gene-
rates scenarios for different protection coordination cases
such as complete, incomplete and impossible protection
coordination.
Figure 8 represents a scenario generated by the scena-
rio manager. Using this scenario, trainees can start fault
isolation and service restora tion train.
2.5. Wide Area Operation
In TDAS, wide area operatio n (WAO) is available which
means oper ators in distribution control center ( DCC) can
manage switches in distribution line which belong to
regional control center (RCC) by transferring control
authority. Each control center including DCC and RCC
has its own middleware (M/W) and naming service.
To realize WAO in the FM-DTS, M/W configuration
is changed as shown in Figure 9 in which the M/W
mana ges several naming services .
Figure 6 . Scenario generation.
Figure 7. Cry w olf alarm addi ti o n.
Figure 8 . One Scenario c ase.
Figure 9. Configuration of FM-DTS M/W.

I.-Y. SEO ET AL.
10
3. Application Programs of FM-DTS
3.1. Short Circuit Analysis Program
This application calculates short circuit current in DG
interconnected distributio n networ k system ba sed on IEC
60909 standards. Several kinds of fault currents at dif-
ferent time instants after fault occurrence can be gener-
ated. This module is composed of Networ k Connectivity
Processor (NCP) and Short Circuit Analysis (SCA) [6].
The flowchart for fault current calculation is shown in
Figure 10.
NCP: This application performs the function changing
the state of system topology, presented to node state on
the kind of b reaker o r switch, into the form of ‘bus’ using
the open or close state of those kinds. Also, the topological
island and t he equipm ent st ate( L i v e or Dead ) are classif ied .
SCA: Based on positive, negative and zero sequence
impedance, the fault current for the each section of the
power distribution system is calculated according to IEC
60909 in case of fault occurrence.
Figure 10. Flowchart of short-circuit current calculation.
3.2. Automatic Service Restoration Program
This program which utilized the algorithm in [7] gene-
rates operation procedures for fault isolation and system
restoration so that trainee can perform self-tr aini ng refe r-
ring to those procedures. The proposed restoration strat-
egy consists of two steps: Candidate Set Generation and
Fuzzy Decision Making. The fault is classified in 6
classes as shown in Figure 11, i.e., Self Restoration,
Single Group Restoration (SGR), Double Group Restora-
tion (DGR), Triple Group Restoration (TGR), Single
Group & Level-2 Load Transfer ( S GRLT), Double
Group & Le vel -2 Load Transfer (DGRLT). The Fuzzy
rule makes decisions to minimize the number of switch-
ing, live load transfer, and to balance the load of the
network [8].
Figure 12 shows FISR solution produced by this pro-
gram (a) and operation actions manipulated by trainer (b).
The solution is separated into several fault areas to be
restored because the ope ration list is long. And it shows 3
solution candidates from the highest priority for each
area.
3.3. Topology Considering DG
Network topology for FM-DTS is developed to prevent
islanding operation of DGs in distribution network. If
any switch installed in source side changes to open, all
the vacuum circuit breakers (VCB) of DG after the
switch are tripped, while in case of DAS only energized
status of line after the switch will be dead.
Figure 11. Six basic restoration schemes (a)Self-restoration.
(b)SGR (c)DGR (d)TGR. (e)SGR_LT (f)DGR_LT.

I.-Y. SEO ET AL.
11
Figure 1 2. FLIS pr ocedure (a) Solut ion (b) Play bac k.
Figure 13. An example of network topology.
Figure 13 shows an example of DG topology. When
switch G2 is open, only VCB2 changes to trip and lines
after G2 will be d e-energized, while the status o f DG2 is
unchanged.
4. Operation
The simulator is implemented on a HP workstation with
Windows 2003 Server operating system and MS-SQL
Server 2008. The MMI is XP window based one with
MS-SQL Server 2008.
Figure 14 shows a simulatio n flow chart of this simulator.
A brief description of running of the simulator is as
follows:
To star t the simulatio n proces s, the si mulator re ads the
static data of the network over which simulation i s to be
performed.
Users input all fault information that is need to define
the fault such as t ype of fault, fault location, r esistance to
ground, etc. W ith these infor mation, applicatio n progra m
calculates fault currents, and generates fault indicator (FI)
at each switches automatically.
After scenario is generated, instructor can modify the
scenario by adding cry wolf alarms like communication
error, battery low of FRT U, switch open etc.
Figure 14. Flowchart of simulation.

I.-Y. SEO ET AL.
12
Now, trainer can start fa ult isola tion and service resto-
ration by manipulating the switche s on one-line diagra m.
Trainer can refer to solution of FISR during and after
the train session which is created b y application program.
After self-training, trainer can play back the operation
actions what he did during the train session in step by
step or batch mode.
Once session is finished, trainer can save the scenario
and st art another session by gener ating new scenarios.
The above process gets r epeated till the end of simul a-
tion.
5. Conclusion
A fault management dispatcher training simulator for
large-scale Distribution Automation System in Korea is
developed to train operators who work in distribution
control center. This simulator can be used for training of
network analysis, network operation, fault management
and evaluation. The simulator can model distribution
network penetrated with DG using the real data from the
TDAS DB. Network fault scenarios are automatically
generated by calculating fault current and generating
fault events. Also, manual entry of cry wolf alarm is
available. Moreover, operation solution for FISR scena-
rio is generated automatically so that trainee can check
their operation result. Operator actions during training
session are saved and can be played back as well as dis-
played on one-line diagram pic tures.
This simulator will be installed nation-wide next year.
It will be very helpful to train operators resulting in
shortening the outage time as well as number of human
errors.
REFERENCES
[1] P. I. Hwang, S. J. Ahn, S. I. Hur, Y. T Yoon, S. I. Moon,
“Adaptive S tep Size M eth od for the Po wer Syste m Mod el
of Dispatcher Training Simulator,” Proceedings of the
2010 Power and Energy Society General Meeting, July
2010, pp. 1-7.
[2] S. V. Vadari, M. J. Montstream, H. B. Ross, “An Online
Dispatcher Training Simulator Function for Real-Time
Analysis and Training,” IEEE Transactions on Power
Systems, V ol. 10, No . 4, 1995, pp. 1798-1804
[3] D. Bolborici, “ExpDTS - Hybrid Dispatcher Training
Simulator Expert System for operator assistance and
evaluation,” Proceedings of the 2009 Power Tech, July
2009, pp. 1-5.
[4] R. Apel, C Jaborowicz and R. Kussel, “Fault Manage-
ment in Electrical Distribution Networks,” Proceedings of
CIRED2001, June 2001, Conference Publication No. 482.
[5] Krish na. K. Agarwal, J. S. Candlish, T. L. Carney, “Au-
tomatic Fault Location and Isolation System for the Elec-
tric Traction Overhead Lines,” Proceedings of the 2002
ASME/IEEE Joint Rail Conference, April 2002, pp.
117-122.
[6] C. H. Chu, Y. S. Yun, S. C. Kwon and I. K. Song, “De-
sign of Application Model for Network Analysis in Smart
Distribution Management,” Proceedings of the 21st In-
ternational Conference and Exhibition on Electricity Dis-
tribution, Frankfurt, 6-9 June 2011, Pape r 02 24 .
[7] S. H. Lee, S. I. Lim and B. S. Ahn, “Ser vi ce r es toration of
primary distribution systems based on fuzzy evaluation of
multi-criteria,” IEEE Transactions on Power Systems,
Vol. 13, No. 3, 1998, pp. 1156-1163.
[8] S. I. Lim, S. J. Lee, M. S. Choi, D. J. Lim and B. N. Ha,
“Service Restoration Methodology for Multiple Fault
Case in Distribution Systems,” IEEE Transactions on
Power Systems, Vol. 21, No. 4, 2006, pp. 1638-1644.