You are on page 1of 43

105

CHAPTER 4
DISTRIBUTION SYSTEM RELIABILITY
EVALUATION SERVICES

4.1

INTRODUCTION
Reliability assessment is the most important factor in designing and

planning of distribution systems that should operate in an economic manner


with minimal interruption of customer loads.

Over the past decades

distribution systems have received considerably less attention regarding


reliability modelling and evaluation than devoted to generating systems. The
main reasons are that the generating systems are very capital intensive and
generation inadequacy will have greater impact on both the society and its
environment. A distribution system is relatively cheap and outages have a
much localized effect. On the other hand, the analysis of customer failure
statistics shows that distribution systems make the greatest individual
contribution to the unavailability of customer supply.
The goal of a power system is to supply electricity to its customers
in an economical and reliable manner. It is important to plan and maintain
reliable power systems because cost of interruptions and power outages can
have severe economic impact on the utility and its customers.

In the

distribution systems, most of the outages or failures would result in direct


impact on the customers. A customer connected to an unreliable distribution
system could receive poor energy supply even though the generation and
transmission systems are highly reliable. This fact clearly illustrates the

106

importance and necessity of conducting reliability evaluation in the area of


distribution systems. Attention should be focused to enhance the distribution
system reliability on par with the generation and transmission systems to
achieve greater customer satisfaction without additional operational cost. The
power utilities require a variety of networked interconnections and
telecommunication technologies to monitor and control power system
operations especially distribution system operations. Reliability assessment
of a distribution system is concerned with the performance at the customer
load points.
4.2

ADEQUACY INDICES IN DISTRIBUTION SYSTEM


RELIABILITY EVALUATION
In the classical concept of reliability, the basic parameters used to

evaluate the reliability of a distribution system can be categorized as load


point indices and system reliability indices. The basic load point indices are
the load point failure rate (), the average outage time (r) and average annual
unavailability or outage (U). The set of system reliability indices includes
interruption indices and energy oriented indices. As most power system
interruptions are due to breakdowns in distribution systems, it is essential for
a distribution company to have proper planning tools to assess and improve its
reliability and performance.
The load point and system reliability indices are normally
determined on annual basis. Because of the stochastic nature of a power
system, the indices for any particular year are random values and are
functions of the component failure rates, repair times and restoration times
within the year. A complete representation of these indices involves
knowledge of the underlying probability distributions. It is relatively easy to
compute the average values as the associated analytical techniques are highly
developed for both radial and meshed distribution systems.

107

4.2.1

System Reliability Indices


The three primary load point indices introduced above are very

important from a customer standpoint. The system performance can also be


assessed on an overall distribution system including system reliability indices.
These indices reflect the adequacy of overall system supply and indicate the
system behaviour and response. The system basic reliability indices are
defined as follows:
System Average Interruption Frequency Index
The index represents the average number of sustained interruptions
experienced by a customer in a unit time (generally 1 year). The definition of
service area is flexible in the sense that the number of customers and the
interruptions experienced by them changes with the definition of the enclosed
area. For instance, a feeder SAIFI indicates the average number of
interruptions a customer serviced by the particular feeder would experience in
a year. Similarly SAIFI reported for a substation or a distribution system
encloses the total customers in the service area.

The system average

interruption frequency index is given in Equation (4.1)

SAIFI

Total Number of Customer Interruptions


Total Number of Customers Served

(4.1)

In order to calculate the index, data of individual sustained


interruptions in a year are required. For each of these interruptions the number
of customers affected comprises the customer interruptions for the particular
outage. The denominator is the total number of customers in the service area
under consideration. Thus, the SAIFI is represented by the Equation (4.2):

108

SAIFI

Ni

(4.2)

NT

where Ni is the number of interrupted customers for each interruption event


during the reporting period and NT is the total number of customers served in
the area. The method to improve on the SAIFI levels of a system is by
reducing the number of sustained interruptions that occur. This can be
achieved through a proper maintenance cycle for each of the component in the
system and through the use of automation and improved protective equipment
that sense faults and attempt to clear the same before it turns into a permanent
outage.
System Average Interruption Duration Index
The index indicates the average time a customer has an interruption
during a time cycle (1 year). It is usually specified in customer minutes or
customer hours of interruption / year. SAIDI (system average interruption
duration index) is the average interruption duration per customer served as
given in Equation (4.3). It is determined by dividing the sum of all customer
interruption durations during a year by the number of customers served.

SAIDI

Customer Interruption Durations


Total Number of Customers Served

(4.3)

SAIDI can be improved by reducing the number of interruptions or


the duration of the interruptions.

For rural areas and long distance feeders,

the time taken to reach the outage spot is comparatively larger than the actual
time of repair. In such cases, SAIDI can be reduced by using optimal crew
dispatch techniques like having a decentralized scheme to crew movement
wherein crew are available at multiple locations throughout the system and
the ones nearest the fault fix the problem. Use of automation is another

109

method to achieve improvements in SAIDI. For a given service area, system


average interruption duration index SAIDI is represented as given in
Equation (4.4).

SAIDI

ri N i
NT

(4.4)

where Ni is the number of interrupted customers for each interruption event


during the reporting period, NT is the total number of customers served in the
area and ri is restoration time for each interruption event. The number of
customers affected and the time it took for the restoration for each
interruption event are the parameters required to estimate the system average
interruption duration index. The restoration time includes, the time taken to
notice an outage, the time taken to locate and reach the location and the time
to repair the fault.
Customer Average Interruption Duration Index
Customer Average Interruption Duration Index (CAIDI) is the
average interruption duration for those customers interrupted during a year. It
is determined by dividing the sum of all customer interruption durations by
the number of customers experiencing one or more interruptions over a one
year period.
The index is the ratio of SAIDI to SAIFI as given in Equations (4.5)
and (4.6) respectively. It represents the average time taken to restore service
to the customers when a sustained interruption occurs.

CAIDI

Customer Interruption Durations


Total Number of Customer Interruptions

(4.5)

110

CAIDI can be improved by reducing the length of interruptions by


faster crew response time and repair times. The value of CAIDI for a given
service area is given as:

CAIDI

ri N i
Ni

SAIDI
SAIFI

(4.6)

where Ni is the number of interrupted customers for each interruption event


during the reporting period and ri is restoration time for each interruption
event.
The performance indices such as SAIFI, SAIDI and CAIDI express
interruption statistics in terms of system customers. A customer here can be
an individual, firm, or organization who purchases electric services at one
location under one rate classification, contract or schedule.

If service is

supplied to a customer at more than one location, each location shall be


counted as a separate customer.
Customer Total Average Interruption Duration Index
The customer total average interruption duration Index (CTAIDI)
represents the average time the customers facing interruptions spent without
power. Unlike CAIDI, only the customers that actually had interruptions are
included in the computation of this index as defined in Equation (4.7).

CTAIDI

Customer Interruption Durations


Total Number of Customers Interrupted

(4.7)

The difference arises in the way an interrupted customer is


accounted.

For the computation of CTAIDI, every customer facing an

interruption is counted only once, irrespective of the number of interruptions

111

seen for the reporting period. The CAIDI uses total number of customer
interruptions in its denominator while the CTAIDI has total number of
customers interrupted.

Thus, CTAIDI can be represented as given in

Equation (4.8).

CTAIDI

ri N i

(4.8)

CN

where, CN is the total number of customers facing an interruption during the


reporting period.
Customer Average Interruption Frequency Index
Customer average interruption frequency index (CAIFI) gives the
average frequency of sustained interruptions for those customers experiencing
interruptions as given in Equation (4.9).

CAIFI

Total Number of Customer Interruptions


Total Number of Customers Interrupted

Ni
CN

(4.9)

SAIFI is the average frequency of interruptions experienced by a


customer and includes even the customers that havent experienced an outage,
while CAIFIs computation involves only the inclusion of customers that have
experienced atleast one interruption. Also, the customer is counted once no
matter how many times they have been interrupted.
Average Service Availability Index
The average service availability index (ASAI) gives the fraction of
time the customer has power during the reporting time. Higher ASAI values

112

reflect higher levels of reliability. Equation (4.10) is used to calculate the


value of ASAI for a given service area:

ASAI

Customer hours service availabili ty


Customers hours service demand

N T 8760

ri N i

N T 8760

(4.10)
Average System Interruption Frequency Index
Unlike SAIFI, ASIFI is an index that uses the load interrupted
rather than the number of customers interrupted. Thus, it is a measure of the
expected number of times load is interrupted during the specified interval of
time. Thus ASIFI for a system is be computed as given in Equation (4.11).

ASIFI

Total connected KVA of Load Interrupted


Total Connected KVA Served

Li
LT

(4.11)
where, Li is the load interrupted due to each outage while LT is the total load
connected to the system under consideration. ASIFI becomes equal to SAIFI
when the load distributed to each customer is equal.
Average System Interruption Duration Index
Similar to ASIFI, ASIDI is load based and computes the average
duration for which load is interrupted when a sustained outage occurs. Thus
ASIDI is computed as given in Equation (4.12).

ASIDI

Connected KVA Duration of Load Interrupted


Total Connected KVA Served

ri Li
LT

(4.12)

113

ASIDI becomes equal to SAIDI when the load is homogeneously


distributed amongst the customers in a system. The units of ASIDI are hours
or minutes for which load is interrupted.
4.2.2

Energy Oriented Indices


The most important parameters required in the evaluation of load

and energy oriented indices is the average load at each load point busbar. The
average load La is given in Equations (4.13) and (4.14).
La = Lp f
where

(4.13)

Lp = peak load demand


f

= load factor

The average load La is also defined as follows:

La

total energy demanded in period of interest


period of interest

Ed
t

(4.14)

Energy not Supplied Index


Total energy not supplied by the system is estimated using
Equation (4.15)
ENS

where,

L a(i) U i

(4.15)

La(i) and Ui respectively are the average connected load and the

average annual outage time at load point i.

114

Average Energy not Supplied Index


The average energy not supplied by the system is estimated using
Equation (4.16).

AENS

total energy not supplied


total number of customers served

L a(i) U i
Ni

(4.16)

where, La(i) and Ui respectively are the average connected load and the
average annual outage time at load point i and Ni is the number of customers
at load point i.
The system wide indices are used to describe the average
performance of the system. For customers requiring high levels of reliability,
using the system wide indices to generalize the quality of service may not be
appropriate. In such cases the use of customer level indices are used to
compute the reliability of the service. Generally, three indices used for the
same are:
Customer interruption frequency which is the frequency of
interruptions seen by the customer during the year
Customer outage duration which is the average time the
customer spends in the interrupted state
Customer service availability is the fraction of the year the
customer has power supply.
Apart from the indices defined so far certain utilities also define
metrics for prioritizing customers based on importance. These indices can be
used not only to assess the past performance of a distribution system but also
to predict the future system performance.

115

4.3

RELIABILITY EVALUATION OF RADIAL DISTRIBUTION


FEEDER TEST SYSTEM
A service oriented model is proposed for predictive reliability

assessment of radial distribution feeder system using Failure Modes and


Effects Analysis (FMEA).

The radial feeder can be represented as an

interconnection of components whose failure characteristics can be used to


predict the system behaviour. The predictive methods are commonly used to
predict the distribution system reliability. The test system, which is shown
in Figure 4.1, consists of a switching station and a typical radial distribution
feeder arrangement (Yeddanapudi et al 2005). The test feeder consists of five
laterals L1, L2, L3, L4, and L5 with various number of customers at load
points C1, C2, C3 and C4 with capacities as shown in Figure 4.1. Fuses and
substation breakers are the protective devices and are shown as F1, F2, F3 and
M1, M2, M3 respectively. It is assumed that the switches are 100% reliable
and the average switching time of each of the switches (S1, S2, and S3) is 0.5
hours. The failure rate information by assuming a single mode of failure of
each segment in the radial feeder test system is given in Table 4.1. It is
assumed that the time taken to repair the fault in the circuit breaker M1 is 4
hours and hence customers at the load points C1, C2, C3 and C4 experience
an outage of 4 hours.
In reliability analysis, huge amount of data are being exchanged
among interconnected systems. The exchange of power system data using
XML offers trouble-free integration with the Web and Intranet / Internet
applications.

In order to compute the power system reliability indices,

utilities often use large databases where outage histories are maintained often
termed as Outage Management Systems (OMS). These databases include
details of the location, date / time of the failure event, the component involved
and the number of customers interrupted due to each outage. Also recorded
are the entities like the time taken to restore service to the affected customers,

116

the time to repair / replace the failed component and more importantly the
cause of the interruption. Events where service was restored in stages due to
switching actions or reconfiguration of the system are also recorded. Put
together, the outage database provides information on each and every event
that happens on a distribution system. The history of outages in power system
is needed for calculating the distribution system reliability indices.

The

outage data consist of number of customers, components failure data, repair


data and forced outage data. SOAP communication with attachment model is
proposed to predict the analytical reliability indices of the radial feeder
distribution test system.

Substation

BRK
L1

M1

C1
F1
900 Customers
1800 KVA

S1 (NC)

M2
L2

L3

C2
F2

550 Customers
1100 KVA

S2 (NC)
NC - Normally Closed
BRK - Breaker

M3

L4
C3

L5
C4
F3

450 Customers
825 KVA

125 Customers
300 KVA

Figure 4.1 Radial Distribution Feeder System

117

Table 4.1 Failure Rate Information of Radial Feeder Test System


Failure
Component
M1
M2
M3
L1
L2
L3
L4
L5

Failure Rate ( )
Frequency/yr
0.10
0.25
0.30
0.20
0.40
0.10
0.10
0.00

MTTR
(hours)
4
4
4
3
3
1
2
0

FOR
hrs/yr
0.40
0.12
0.15
0.60
0.00
0.05
0.05
0.00

The XML data representation for distribution system reliability


analysis is as follows:
<distribution_system_reliablity_data>
<general>
<no_of_customers>Total No. of Customers at Loading Point
</no_of_customers >
<clr type="KVA"> Customer Load Rating </clr>
<id>Customer System Data </id>
</general>
<loadpoint_data>
<loadpoint name="C1">
<failure_device_name device="M1">
<failure_rate>Failure Rate of the Device</failure_rate>
<mttr type="hour">Mean Time To Repair</mttr>
<for>Forced Outage Rate</for>
</failure_device_name>
</loadpoint >
</loadpoint_data>
</distribution_system_reliability_data>

118

The reliability data pertaining to the Radial Feeder Test System is


represented using XML annotations and the corresponding XML document
and its XML schema are generated.

The data required for FMEA analysis

are represented in a class LoadPointData, which encapsulates the load point,


protective device identification, failure rate of each component, mean time to
repair and forced outage rate. The class LoadPointData is described as
follows:
class LoadPointData {
String loadPoint;
String failureModeDevice;
double failure_rate;
int mttr;
double forcedOutageRate;
public LoadPointData(String lp, String fmd, double fr,
int mr, double outage_rate)
{
loadPoint=lp;
failureModeDevice=fmd;
failure_rate=fr;
mttr=mr;
forcedOutageRate=outage_rate;
}
}

In order to represent each member of the above class as an XML


element in an XML document, the class LoadPointData has to be annotated
with @XmlRootElement annotation.

As it is aimed to estimate the

distribution system reliability indices, an instance to the LoadPointData


class is encapsulated with other members such as number of customers and

119

customer load rating in a class named as DistributionSystemReliability, thus


making the members of the LoadPointData class as child elements in the
XML document.

The @XmlRootElement annotation associated with the

DistributionSystemReliability class has two elements, name and namespace


and described as follows:
@XmlRootElement(name ="reliability",
namespace ="http://www.reliability.com/radialfeeder")
public class DistributionSystemReliability {
int noOfCustomers;
int customerLoadRating;
LoadPointData loadPointData;
DistributionSystemReliability(int noOfCustomers,
int customerLoadRating,
LoadPointData lpd)
{
this.noOfCustomers=noOfCustomers;
this.customerLoadRating=customerLoadRating;
loadPointData = new LoadPointData();
loadPointData.loadPoint= lpd.loadPoint;
loadPointData.failureModeDevice= lpd.failureModeDevice;
loadPointData.failure_rate= lpd.failure_rate;
loadPointData.mttr= lpd.mttr;
loadPointData.forcedOutageRate= lpd.forcedOutageRate;
}

The code segment to generate the formatted XML document using


the above annotation is as follows:

120

JAXBContext context =
JAXBContext.newInstance(DistributionSystemReliability.class);
Marshaller marshaller = context.createMarshaller();
marshaller.setProperty(Marshaller.JAXB_FORMATTED_OUTPUT, true);
OutputStream os=new FileOutputStream("dsr.xml");
marshaller.marshal(new DistributionSystemReliability(900, 1800, new
LoadPointData("C1", "M1", 0.10, 4, 0.40)), os);

The formatted XML document stored in the dsr.xml file, which


represents the distribution system reliability estimation data for the radial
feeder test system, is given below.
<?xml version="1.0" encoding="UTF-8"?>
<ns2:reliability xmlns:ns2="http://www.reliability.com/radialfeeder">
<noOfCustomers>900</noOfCustomers>
<customerLoadRating>1800</customerLoadRating>
<loadPointData>
<loadPoint>C1</loadPoint>
<failureModeDevice>M1</failureModeDevice>
<failure_rate>0.1</failure_rate>
<forcedOutageRate>0.4</forcedOutageRate>
<mttr>4</mttr>
</loadPointData>
</ns2:reliability>

The necessary XML schema documents are generated using


schemagen utility for the DistributionSystemReliability class. The XML
schema definition to map the DistributionSystemReliability class is as
follows:

121

<?xml version="1.0" encoding="UTF-8"?>


<xs:schema version="1.0"
xmlns:xs="http://www.w3.org/2001/XMLSchema">
<xs:complexType name="reliability">
<xs:sequence>
<xs:element name="loadPointData" type="LoadPointData"/>
<xs:element name="noOfCustomers" type="xs:int"/>
<xs:element name="customerLoadRating" type="xs:int"/>
</xs:sequence>
</xs:complexType>
<xs:complexType name="loadPointData">
<xs:sequence>
<xs:element name="customerId" type="xs:string"/>
<xs:element name="failureModeDevice" type="xs:string"/>
<xs:element name="failure_rate" type="xs:double"/>
<xs:element name="forcedOutageRate" type="xs:double"/>
<xs:element name="mttr" type="xs:int"/>
</xs:sequence>
</xs:complexType>
</xs:schema>

Subsequently, the above schema is used to generate the XML


document to represent the reliability data for the radial feeder distribution
system and can be added directly to the SOAP envelope as request payload,
while a request is initiated to invoke the distribution system reliability
evaluation services.

122

4.4

SOAP COMMUNICATION MODEL FOR RELIABILITY


ESTIMATION OF ELECTRIC POWER DISTRIBUTION
SYSTEMS
The major components of the proposed SOAP communication

model are Distribution System Reliability (DSR) Service Provider and the
Service Requester. The main aim is to represent the estimation of distribution
system reliability indices as a service and to exchange the required data using
XML over SOAP. The SOAP communication model for distribution system
reliability analysis is shown in Figure 4.2.

DSR Service Provider


Service Requesters
Power System
Clients

System Reliability Indices


Interfaces
Reliability service invocation

SOAP with Attachment


API
Run Time Environment

Binding

(SAIFI service)
(SAIDI service)
(CAIDI service)
(CAIFI service)

(CTAIDI service)
(ASAI service)
(ASIFI service)
(ASIDI service)

Distribution System Reliability


service description using WSDL

Exchange of reliability data


Using XML over SOAP

Binding

Transport protocol (HTTP)

Figure 4.2

SOAP Communication Model for Distribution System


Reliability Analysis

The DSR services are categorized into estimation of Interruption


Indices such as System Average Interruption Frequency Index (SAIFI), the
System Average Interruption Duration Index (SAIDI), the Customer Average
Interruption Frequency Index (CAIFI), the Customer Average Interruption
Duration Index (CAIDI), Customer Total Average Interruption Duration

123

Index (CTAIDI), the Average Service Availability Index (ASAI),

the

Average System Interruption Frequency Index (ASIFI), Average System


Interruption Duration Index (ASIDI) and estimation of energy indices such as
the Energy Not Supplied (ENS) and the Average Energy Not Supplied
(AENS).
The DSR Service provider offers the above services and describes
the services to its clients using Web Services Description Language (WSDL).
Since the described services are Distribution System Reliability Services, the
WSDL file is named as DSRS-WSDL, which is in XML format. A service is
a reusable function that can be invoked by another component through a well
defined interface.

Services are loosely coupled, that is, they hide their

implementation details and only expose their interfaces. In this manner, the
power system client need not be aware of any underlying technology or the
programming paradigm which the service is using.

The loose coupling

between services allows for a quicker response to changes than the existing
conventional applications for power system operations. This results in a
much faster adoption to the need of power system industries.
4.4.1

Implementation of the Proposed Model


A radial feeder test system is considered for reliability evaluation

using the proposed model. Synchronous SOAP request-response model is


proposed in which the request payload encapsulates service invocation details
and reliability data in XML format and the SOAP response messages
represent system reliability indices, which are also in document style format.
The various stages involved in the implementation of the proposed SOAP
communication model for distribution system reliability analysis are XML
data representation, reliability service interface, service implementation,
service configuration, description, service mapping, service binding and
service invocation.

124

In the proposed interoperable SOAP communication model for


power distribution systems reliability evaluation, the reliability data is
included as attachment within the SOAP message. The power system
reliability services are designed to enhance the interoperability based on
standards such as HTTP, XML and SOAP.

The calling sequence of

reliability evaluation services along with required data is configured as a


SOAP message. The proposed SOAP communication model in the form of
document style will provide a common computational environment for
interaction between various power system clients and reliability service
providers. Analytical techniques for distribution system reliability assessment
are effectively used to evaluate the wide range of system reliability indices.
4.4.2

Reliability Service Interfaces


The service interface provides the contract between the power

system client and reliability service provider.

The service interface is

responsible for all of the implementation details needed to perform the


communication between the clients and service provider. In this proposed
model, eight interfaces are created for the representation of various reliability
indices estimation services such as System Average Interruption Frequency
Index estimation, System Average Interruption Duration Index estimation,
Customer Average Interruption Duration Index estimation etc. Decoupling
the service interface from the service implementation enables the system to
deploy two codebases on separate tiers, potentially increasing the deployment
flexibility. The service interface for computing System Average Interruption
Frequency Index is as follows:
package reliability;
public interface saifi extends Remote {
public String estimateSAIFI() throws RemoteException;
}

125

The service interface encapsulates all aspects of the network


protocol used for communication between clients and service provider. All
other interfaces are described similarly. The implementation of the service is
modified without affecting the user who consumes that service. The services
related to computing of power distribution system reliability indices are
configured as follows:
<service name="ReliabilityService"
targetNamespace="urn:Reliability"
typeNamespace="urn:Reliability"
packageName="reliability">
<interface name="reliability.saifi"/>
<interface name="reliability.saidi"/>
<interface name="reliability.caifi"/>
<interface name="reliability.caidi"/>
..........................
<interface name="reliability.asidi"/>
</service>

The above XML document contains the information and details


about the deployed distribution system reliability services (SAIFI, SAIDI
etc.,) and metadata such as their service name (ReliabilityService) and
namespace (Reliability).
4.4.3

Reliability Service Descriptors


A service description is document-based that defines or references

the information needed to use, deploy, manage and control the reliability
services.

The service descriptor includes the information and behaviour

models associated with a service to define the service interface. The purpose
of service descriptor is to facilitate interaction and visibility, particularly

126

when the providers and participants are in different ownership domains. The
RSDs describe how the service provider and client communicate with each
other. It also provides the information about the data type, binding and
address information for invoking the services from the service provider. The
SAIFI service is described as follows:
<definitions name=ReliabilityService
targetNamespace="urn:Reliability">
<types>
<schema targetNamespace="urn:Reliability">
<doubleType name="estimateSAIFI">
</doubleType>
<element name="estimateSAIFIResponse" type ="string"/>
</types>
<message name="SAIFI_estimateSAIFI">
<portType name="saifi">
<operation name="estimateSAIFI">
<input message="tns: SAIFI_estimateSAIFI"/>
<output message="tns:
saifi_estimateSAIFIResponse"/></operation>
</portType>
<binding name="saifiBinding" type="tns: saifi">
<soap:binding transport="http://schemas.xmlsoap.org/soap/http"
style="document"/><operation name="estimateSAIFI">
</binding>
<service name="ReliabilityService">
<port name="saifiPort" binding="tns: saifiBinding">
<soap:address location=" http://loacalhost:8080/reliability"/>
</service>
</definitions>

127

The name of the service is defined as ReliabilityService and


declares the namespace as Reliability. The Java based data types used in the
interfaces are converted into JAX-RPC based data types that would be used to
describe the distribution system reliability data within the Web service. The
abstract definition of the operation estimateSAIFI that uses a concrete
protocol SOAP to represent messages is described in the Web service
descriptor file. The SOAP message represents a logical definition of the data
being transmitted between the client and the service provider. The service is
bound to an endpoint address named as saifiPort through which the clients
can invoke the reliability estimation services.

In this proposed SOAP

communication model, all the reliability service objects are translated and
mapped into XML for interoperability in a heterogeneous environment. The
mapping file describes how the Java objects like package, type, port, method,
and endpoint are mapped into XML and vice versa. While invoking the
reliability service, the method call and its parameters are mapped into XML
and sent through SOAP communication protocol. When received at the client
or server end, the request / response parameters must be mapped from XML
to their proper types or objects to make interoperability inherently.
4.4.4

SOAP Handlers for Reliability Estimation Service Messages


SOAP Handlers are effectively utilized in order to maintain power

system operations in a pre-defined sequence.

In the proposed SOAP

communication model, the SOAP request sent by the client is received by the
service provider as a SOAP MessageContext. A handler is introduced to
extract the request payload from the MessageContext and to parse it to extract
the actual data, which is required for reliability estimation services. A handler
chain is used to control the sequence of operations to obtain the desired
results. Separate handlers are introduced to carry out the intended task in
sequence to obtain the desired reliability index. In order to estimate the

128

customer average interruption duration index (CAIDI), the proposed model


has to invoke the XMLised power system reliability data generation Web
service, the service associated with the estimation of system average
interruption frequency index (SAIFI) and followed by invoking the service
associated with system average interruption duration index (SAIDI). The
sequence of operations is effectively controlled using <handler> elements
within a handler chain XML document.

The handler chain used in the

proposed model to estimate CAIDI is given as follows:

<?xml version="1.0" encoding="UTF-8"?>


<handler-chains xmlns="http://java.sun.com/xml/ns/javaee">
<handler-chain>
<handler>
<handler-name>XMLised Data Generation</handler-name>
<handler-class>power.XMLisePSDataImpl</hander-class>
</handler>
<handler>
<handler-name>Estimation of SAIFI</handler-name>
<handler-class>reliability.SaifiImpl</handler-class>
</handler>

<handler>
<handler-name>Estimation of CAIDI</handler-name>
<handler-class>reliability.CaidiImpl</handler-class>
</handler>
</handler-chain>
</handler-chains>

129

The handler chain has been associated with the CAIDI estimation
service using @HandlerChain annotation with the file element initialized
with

handler-chain.xml.

The

CAIDI

estimation

service

calls

handleMessage() method, which will receive the request payload as SOAP


MessageContext. All service handlers listed in sequence within the handlerchain.xml file are effectively managed by the handleMessage() method and
the SOAP response for the CAIDI estimation service is sent to the requested
client. Each handler is associated with handleFault() method to report the
errors and exceptions to the client while executing the services in sequence.
This model provides an automation of sequencing the services and it is of
great importance while handling complex power system processes where
many operations with variable response times run simultaneously in which the
response of one operation will become the request for subsequent operations.
4.4.5

SOAP Request and Response for Reliability Services


The reliability service descriptor represents information about the

interface and semantics of how to invoke a reliability service. The power


system

client

has

to

establish

the

SOAP

connection

SOAPConnectionFactory for invoking the desired services.

using

The SOAP

message is created using MessageFactory reference to an endpoint interface.


The method call and its parameters are mapped into XML and sent through
SOAP communication protocol. When received at the client or server end,
the request / response parameters must be mapped from XML to their proper
types or objects. The following code segment delineates how the SAIFI
service is being invoked by the power system clients.

130

String destination =
"http:/192.168.1.1:1078/powersystem/ReliabilityService";
SOAPConnectionFactory soapConnFactory =
SOAPConnectionFactory.newInstance();
SOAPConnection connection = soapConnFactory.createConnection();
SOAPMessage reply = connection.call(message, destination);

The SOAP request with the attachment of power distribution


system reliability data for invoking the reliability estimation service is shown
below:
<SOAP-ENV:Envelope
.. <!-- method invocation request -->
</SOAP-ENV:Body></SOAP-ENV:Envelope>
<SOAP:Attachment>
<distribution_system_reliability_data>
<general>
<cust>900</cust>
<clr type="kva">1800</clr>
</general>
<loadpoint_data>
<loadpoint_customer one="c1">
<failuremode_device_name brk="m1">
<frd>0.10</frd>
<mttr type="hour">4</mttr>
<for>0.40</for>
</failuremode_device_name>
<failuremode_device_name brk="m2">
<frd>0.25</frd>
<mttr type="hour">0.50</mttr>
<for>0.125</for>
</failuremode_device_name>
</loadpoint_customer>
</loadpoint_data>
</distribution_system_reliability_data>
</SOAP:Attachment>

131

The SOAP response for SAIFI and SAIDI services obtained using the
proposed model is shown below:
<env:envelope xmlns:env="http://schemas.xmlsoap.org/soap/envelope/"
xmlns:enc="http://schemas.xmlsoap.org/soap/encoding/"
xmlns:ns0="urn:reliability"
xmlns:xsd="http://www.w3.org/2001/xmlschema"
xmlns:xsi="http://www.w3.org/2001/xmlschema-instance">
<env:body>
<ns:computeresponse>
<result>
<distribution_system_evaluation>
<dsrindices>
<saifi>1.04999</saifi>
<saidi>2.02257</saidi>
..
</dsrindices>
</distribution _system _evaluation>
</result>
</ns:computeresponse>
</env:body>
</env:envelope>
In the proposed SOAP communication model, the various
reliability services can be invoked by the clients without any limitations. It
has been proven that the proposed model is capable of representing the power
system problems in heterogeneous environments.

The data required for

distribution system reliability evaluation are attached in the SOAP body. The
provider sends an entire document rather than sending a set of parameters to
the clients. SOAP message handlers are used to intercept both SOAP request
and response messages and for sequencing the operations. The proposed
SOAP communication model for reliability analysis makes the distribution

132

system reliability service provider and its clients to exist in a loosely coupled
environment.
4.5

SECURE AND RELIABLE SYNCHRONIZED EVENT


NOTIFICATION SYSTEM
Over decades, computer network and distributed information

processing technologies have seen profound growth. In the meantime,


deregulation is starting and will be spreading in the electric power industries.
The deregulation of power sectors is bringing about major changes in the
utilities, with new players involved in the network and increased quality and
profitability requirements. Consequently, customers request low cost power
supply and high quality services from the power suppliers. Corresponding to
these requirements caused by deregulation, Seki et al (2002) have developed a
prototype power systems information delivering system using Internet
technologies supplying useful power systems information such as power
failures. The purpose of using Internet technologies is to reduce the cost and
time to develop applications and integrate them with existing systems. The
prototype system consists of an application layer based on a Web browser, an
information model layer which reflects the power systems behaviour and a
distributed object management layer using CORBA and OS / Network layer.
Due to deregulation policies, an electric power company needs to compete
with other power companies and hence more efficient operations are required
and various services are to be provided to the customers. There will be stiff
competition on cost and services associated with supplying power.

It is

necessary to provide timely power systems operational information to the


customer.
Originally the software was designed and built in the EMS /
SCADA systems, which was dependent on the hardware since the data had
been received from the sensors. Later the EMS / SCADA for power systems

133

have been developed using Internet Protocol (IP) for the Wide Area Network
(WAN) connected between the control centres and substations. This system
provides information on power failures to customers through the Internet, in
addition to the operators of the electric power sector. Moreover, the system is
able to supervise the status and failure of both the power transmission systems
and the power distribution systems simultaneously.

This system detects

power failures using the operation status and activated status of the switch and the
relay. The power systems information delivering system has the following
features:
Power failure monitoring using Web browser
Reproducing at any time the power systems status in the past
The integration of the information on power systems
configuration and status for the power transmission and the
power distribution network
The integration of power systems information and geographical
map information
The provision of statistical or historical information for offline
use, such as calculated blackout time or number of customers
affected
Kivikko et al (2003) presented a Web application for viewing
customer-specific interruption data.

The application is integrated with

commercial Distribution Management System (DMS) and it uses the same


database as DMS system.

The authors have clearly described about the

content of interruption database and the construction of interruption sector.


They have proposed a Web application for customer interruption monitoring.
The application presents the existing outage data measured and saved by the
SCADA and DMS systems on the browser. They have also explored the

134

possibilities of new technologies, which make the development of Web and


mobile applications. Microsoft Visual Studio .Net programming tool and C#
programming language are used in the implementation of the Web
application. They have described a logic how the interruption data is gathered
to the DMS database and how the number and durations of interruptions can
be searched for a certain customer and illustrated a methodology for
construction of interruption sector.

The major aim is to develop a Web

application, which uses the existing interruption data gathered by other


systems and can be integrated with existing commercial data systems.
Yves Chollot and Martine Pauletto (2003) of Schneider Electric
Industries have proposed Web based solutions which provide an efficient and
reliable sharing of the network information between all the clients and which
increase the quality of their services.

They have developed a general

architecture that describes the system that is Web enabled and designed for
medium voltage / low voltage substation monitoring and maintenance. The
developed system utilizes Web, GSM and GPRS technologies, which are
becoming lesser and lesser expensive and makes it possible to:
inform the right persons when there is an alarm in a substation
via e-mail
inform through SMS
display on the Web through a PC with a Web browser
provide real time information from the substation and also
provide historical information
Even though Web enabled power system applications provide faster
and timely information and collaborate efficiently with other network

135

technologies, they are still facing challenges due to heterogeneous nature and
the way they are communicating with each other.
Kannammal et al (2006) have proposed a secure model based on
the concept of shared objects and mobile agents to secure the business
database present in the e-business environment.

The models have been

implemented that would control the direct access to the business database
while maintaining synchronized data transfer.

This model faces security

challenges because of the introduction of mobile agents. The author uses


public key cryptographic algorithm to secure communication using mobile
agents, which causes additional overhead to the overall performance of the
system.
Wang Xin Fang et al (2003) had applied XML and SOAP in
transformer substation management information system for communication
between client and server and for providing data exchange with other electric
utilities. They have made possible that transformer substation can exchange
data with other power utilities through firewalls. Cao Hou-ji et al (2007) have
described how transmission substations exchange large quantity of outage
data among the heterogeneous platforms by using SOAP message in a
predefined format with strong firewall passing ability and without the need
for additional installation of special software to receive them.
Pasteur et al (2007) have proposed to utilize SOAP based Industrial
Messaging Specification (SIMS) to exchange power plant process data with
increased real time performance. The proposed SIMS messaging is based on
two components: a client and a server. The server component is a data
provider, which provides read, write and subscribe Web services. The client
component is the data consumer and it aims to invoke services and provides
call-back functions. These call-back functions are also Web services used by
the server to notify the client that new data are available. They have designed

136

the messaging server to have an access to any type of data sources and
capable of providing these data if its Web services are invoked.
The Messaging server provides two Web services read() and write()
that can be called by the client to write or read current values. Also client can
subscribe to some data in order to be notified if any of the attributes of this
data changed. If data changes are made, a notification is created and stored in
a buffer and the client will be intimated using notify message that new
values are available for the data to which it has subscribed. The authors have
demonstrated this SIMS messaging based on Web services to exchange data
between a hydraulic power plant and a telecontrol center.
The SIMS messaging was first used in SCADA application to
telecontrol 15,000 MW hydro power plants from four dispatching centres
(Laurent Bacon and Cedric Bellec 2006).

Three levels tele-control

architecture was defined to ensure the robustness of the control systems. First
level concerns about the power unit control, the second level called as plant
local control deals with power distribution among all power units and the
third level called as dispatching center deals with control, modification and
supervision of generation schedules. SIMS is used as the communication
protocol between a dispatching center and hundreds of local control plants.
Open connectivity unified architectures are evolving based on service
oriented architecture to have effective message based communication.
Standard specifications are emerging to define a base set of generic services
to browse and query namespaces, read / write data and publish / subscribe
events and data changes.
Jeongje Park et al (2010) have developed Web based on-line real
time reliability integrated information system for monitoring reliability of
electrical energy supply including wind turbine generator. In order to supply

137

information about the quality, reliability and security of the electric service,
several techniques and methodologies have been adapted that include
traditional operation planning, real time control functions and a redesign of
control system hardware and software architectures. As the utilization of
renewable resources has been receiving considerable attention in recent years,
the information system requirement is increased. The reliability information
system is more important for implementing the smart grid. The proposed
integrated information system not only supplies the information about the
reliability indices of the power system but also the estimation of CO2
emission.
The system developed by them is used to evaluate probabilistic
production energy with reduction in production cost and to obtain CO2
emission by inserting renewable energy to the power systems. Probabilistic
reliability indices have been used extensively for generation expansion
planning. The basic reliability indices, namely the loss of load expectation
(LOLE), the expected energy not supplied (EENS) and the energy index of
reliability (EIR) can be calculated using effective load duration curve. From
these reliability indices, probabilistic production energy, production cost,
capacity factor and CO2 emission can be obtained. The proposed on-line
reliability information system is successfully established and applied to Jeju
Island Power System in South Korea. The functioning of this system can be
viewed in the Website, http://worris.gsnu.ac.kr/PraWin. The users of this
system are the system operators, decision makes and information seekers and
they will access the system with browser via the Web. Jaeseok Choi et al
(2010) have extended the above work for grid constrained probabilistic
reliability evaluation of power systems including wind turbine generators.
They have developed a multi-state model for composite power system
reliability evaluation based on the composite power system effective load

138

model in order to consider wind turbine generators. The proposed work is


integrated with the information system developed by Jeongje Park (2010).
In order to enhance the security of operational data of different
power sectors in the deregulated environment and to protect the history of
outage information, a new SOAP based communication model is proposed
using remotely accessible shared objects between the messaging server and
client. The proposed model ensures synchronized communication between
the operations monitoring service providers like SCADA applications and the
power system operators to enhance reliable notification of failure events and
other operational data required for decision making to keep the capacity
reserve at the required levels to meet the demand and to avoid catastrophic
events.
The proposed model uses remote sharable bean components and
Web services to update the power system operators automatically with new
information due to data changes with respect to the operational status of the
power systems. The service that resides in the database server is informed
about the new information by triggering a function. Then the Web service
updates the bean with the recent outage information by invoking a remote
method. This sharable component is accessed by another Web service on the
client side, which sends the information to the decision makers and
information seekers. This approach improves security, as clients are not
aware of the location of the outage history database and makes power system
events notification application more scalable by deploying Web services.
4.5.1

Securing Outage History Database for Power System


Applications
While integrating power system applications with the Internet

offers potentially unlimited opportunities for increasing efficiency and

139

reducing cost, it also offers potentially unlimited risk. In the automation of


power system operations, the system parameters are maintained consistently
in a database. The outage history thus logged in the database throughout the
operation of power systems reflects the quality of power system maintenance.
In the deregulated environment, the power utilities should keep their
operational data secure especially the outage history table. Critical decisions
have been made based on the operational data and based on the outage history
table. A knowledgeable and malicious client or even the system operator can
execute unauthorized procedures or SQL queries inside the outage history
database.
An unauthenticated user with browser access to a Web server
hosting the power system application can exploit the database. This activity
will lead to unwanted decisions, which will pose a false hope to the planners
as well as consumers and create a negative impact on the quality of service of
the system. If an unauthorized user voluntarily records a false outage data to
the outage history table even when the power system operation is in line with
the normal and stable operating conditions, it will cast an illusion of poor
maintenance of power system.

The risk of exposure to power system

operational data as well as outage history database is high, as any client with
browser access and specialized knowledge can exploit these vulnerabilities.
At the same time, the occurrence of the critical events should be notified to
the system operators in a synchronized manner without any information loss
or omission in order to take timely decisions.
A Web service model has been proposed for event notification
system to improve scalability and a concept of remotely accessible sharable
bean component is introduced to improve the security of Web applications.
This sharable bean component named as shared object is used not only to
improve the security, but also to enable synchronized and reliable

140

communication between server side applications. Using the recently stored


contents of the sharable bean component, the Web service updates the
operators and decision makers automatically with new information. The Web
service associated with the database server is informed about the new
information by triggering a function. This information is copied onto the
static sharable bean component, which is accessed by another Web service,
which will forward the same to the system operators and to the clients. The
block diagram representation of the architectural design of synchronized
event notification system is shown in Figure 4.3.

Remote
Outage
History
Table

Web
Service

Sharable Bean
Component

Web
Service

Power
System
Clients

Trigger

Figure 4.3

Block Diagram Representation of the Event Notification


System

Power system utilities maintain data about the various outages


occurring in the system, especially in distribution systems. The outage data
includes failure histories which comprise details of fault occurrence times,
fault restoration times, feeder information and interruption type. A radial
system with 2 feeders (F1 and F2) having an outage history as shown in
Table 4.2 has been considered (IEEE Guide, 2003) to illustrate the event
notification system. Assuming Feeder 1 has a total of 900 customers along
with a load of 1800 kVA and Feeder 2 has 1850 customers with a load of
3700 kVA, the total number of customers in the system is: 900+1800=2700
while the total load connected to the system is: 1900+1125= 4025 kVA. An
interruption is loss of power supply to the customer and its effect is variable,
which is classified as sustained and momentary based on the duration of
failures.

The interruption types S and M stand for sustained and

momentary respectively.

141

Table 4.2 Historical Outage Data for a Radial Feeder system

Date

Time of
fault

23 March

12:02:20

12:20:30

F1

900

1800

15 April

16:13:56

16:14:26

F1

550

1100

5 May

00:23:10

01:34:29

F1

450

825

12 June

23:17:00

23:47:14

F2

400

800

6 July

09:30:10

09:31:10

F2

1850

3700

20 August

15:45:39

20:12:50

F1

450

825

31 August

08:20:00

10:20:00

F2

900

1800

3 September

17:10:00

17:20:00

F2

950

1900

2 October

10:15:00

10:55:00

F2

1850

3700

31 October

01:47:25

03:35:15

F2

900

2600

23 November 15:00:05

15:20:00

F1

550

1100

13 December

09:06:15

F2

1850

3700

09:05:10

Time of
No. of
Load Interruption
Feeder
restoration
Customers kVA
type

The notification system includes one or more system operators and


service consumers to whom the failure events are to be informed in time to
take appropriate actions. Figure 4.4 depicts the architecture of the proposed
secure, reliable and synchronized event notification system.
The power system operators and decision makers have to
continuously enquire the database server by sending requests to the client side
Web service. The database server has to send responses to all the requestors
only during occurrence of new events in the system.

142

B1

DS 1

SCADA
System
(custom)

Outage
History
Table
Trigger

Create
Store

Browser
JAX-RPC
based
Web service

Invoke
Web service

E-mail
FAX/
Printer
SMS

Java Bean
Sharable
Component

Power system
Client side
Web service

Figure 4.4 Secure and Reliable Synchronized Event Notification System


Web services have been deployed to act as intermediaries between
the database server and the remotely accessible sharable bean and between the
sharable bean and the power system clients respectively. The sharable bean
component contains fields to hold the information present in the outage
history database due to insertion of new failure event information or due to
update of existing outage data. And also it has a flag to indicate whether the
recent contents are consumed by the power system clients or not. The Web
services access the sharable bean using a middleware component, which is
common to both forwarding and notification services. The middleware
application in turn uses the RMI framework to access the sharable bean
component. The sharable bean component has been declared as static within
the remote object to enable sharing of data between services.

143

The outage history information as shown in Table 4.2 is stored in a


relational database using a table name power. The table has been associated
with the following trigger, which has to be initiated while inserting new
outage record or updating the contents of the existing record.
/*An insert or update trigger, which invokes in turn a Web Service
to forward the outage information*/ create or replace trigger message_send
after insert or update on power for each row
declare
fault_occurrence_time date;
restoration_time date;
feederno varchcar2(10);
interruption_type varchar2(2);
newdata varchar2(36);
mobile_no varchar2(15);
email_id varchar2(30);
begin
fault_occurrence_time := :new.intime;
restoration_time := :new.restorationtime;
feederno := :new.feedernumber;
interruption_type := :new.interruptiontype;
mobile_no := :new.mobileno;
email_id := :new.emailid;
newdata:= 'Outage Data ' || to_char(fault_occurrence_time,'hh:mi:ssam') ||
to_char(restoration_time,'hh:mi:ssam') || feederno ||
interruption_type;
communicationService(newdata, mobile_no, email_id);
end;

144

The trigger has to invoke the SQL procedure, communicationService


to forward the outage data or restoration data to the Web service associated
with the database server. The Web service in turn invokes a remote method,
storeData() to copy the outage information in the static sharable bean
component and sets the flag in the bean as true.

The procedure for

communicationService is given in Appendix 4. The communicationService


will access the JAX-RPC based Web service using the endpoint reference,
http://localhost:8080/power/ForwardingService?WSDL
When there is an information update, the database server sends the
recent outage information to the Web service logically named as
ForwardingService

through

the

message_send

trigger.

The

ForwardingService checks the flag attribute of sharable component to find


out whether the information provided by previous update or insert has
been consumed or not by the Web service on the client side, logically named
as NotificationService.

If the flag is set to false by the notification

service, it means that the previous information is consumed, and then the
forwarding service updates the outage information attributes of the sharable
bean by invoking the remote method storeData(). Also, the flag attribute of
the bean is set to true to indicate that new information is available for the
notification service to consume.

The NotificationService has been

implemented using SOAP communication in AXIS2 platform.


The notification service on the client side continuously monitors the
remotely accessible sharable bean for the availability of new outage
information. When the flag attribute is true then it means that the new
information is available, and hence notification service retrieves the
information and sets the flag to false, to indicate that the information is
consumed.

The flag field is updated accordingly in order to ensure

synchronized and reliable data transfer between the database server and the

145

clients. The power system clients have to contact the notification service
continuously by sending request messages for any information update. The
clients need not contact the database server for the information. Also, the
location of sharable bean and the database server are hidden from the clients.
By this approach, the outage history database is made secure by avoiding
direct access to the database server. Since the forwarding and notification
services are deployed as Web services, they are inherently scalable and will
respond to a large number of clients who are sending request messages for
updated information.
As the system is designed to notify the fault events during power
systems operations, it is not sufficient to have only scalable and synchronized
services but also reliable services are required to notify the system events to
the power system operators at appropriate time to enable them to take
effective decisions. The reporting facility is extended to various devices to
enhance the reliable communication.

The notification service has been

developed to report the consumed information from the sharable bean


component to the power system operators using various modes of
communication such as PC, Web browser, printer or fax, SMS using a mobile
device or sending the outage information via e-mail.
This model enhances the security of operational data of different
power sectors in the deregulated environment and protects the history of
outage information as the clients are completely detached from the location of
the operational database. A common middleware component connects the
forwarding service and the notification service, which in turn uses an RMI
framework to access the static sharable bean component. The reliability of
notification system is enhanced by providing the information to various
devices simultaneously.

146

4.6

CONCLUSION
The predictive reliability assessment of radial distribution feeder

system using Failure Modes and Effects Analysis has been presented and
analyzed in the distributed environment.

The basic distribution system

reliability indices used in practice are illustrated.

The reliability data

pertaining to the radial feeder test system is represented using XML


annotations.

An effective SOAP communication with attachment model has

been developed to evaluate the reliability of power distribution systems.


SOAP handlers are used to intercept the SOAP messages and for
manipulating the request sent by the client. A handler chain is introduced in
the distribution system reliability estimation model in order to control the
sequence of invocation of services starting from data generation to sending
the SOAP response to the requested clients. Each handler is associated with
the error handling procedure to report the exceptions to the clients while
executing the services in sequence. Analytical techniques for distribution
system reliability assessment are effectively used to evaluate the wide range
of system reliability indices.

Other power system services can also be

plugged into this model and the services are made available anytime and
anywhere for the power system planning and operations.
A secure and scalable Web application framework has been
developed using SOAP communication for continuous monitoring of power
distribution systems and maintenance of outage history information which is
required for planning while further expansion. The model is designed to
report the status of the power distribution system corresponding to outage and
restoration state in a synchronized way. This model enhances the security of
operational data of different power sectors in the deregulated environment and
protects the history of outage information as the clients are not aware of the
location of the operational database. The SOAP communication model uses

147

sharable bean components and Web services to update the power system
operators automatically with outage or restoration information due to data
changes with respect to the operational status of the power systems. The
service that resides in the database server is informed about the new
information by triggering a function. The reporting facility is extended to
various devices such as PC, Fax and other electronic gadgets to enhance the
reliable communication to avoid abnormalities. The developed notification
model is scalable as the forwarding and notification services are designed as
Web services and the critical events are notified to the operators as well as to
the decision makers instantly in a synchronized way, which enables them to
take appropriate actions in time to improve the quality of service and
maintenance.