Sei sulla pagina 1di 12

INFORMATION MODEL FOR THE COMPUTATION OF THE

POWER SUPPLY CONTINUITY INDICATORS FOR A POWER


DISTRIBUTION BRANCH
RCANU VALENTIN GRIGORE 1, ALBERT HERMINA 2, GOLOVANOV NICOLAE 3, PUN
DAN STELIAN 1, PUN CLAUDIU MIRCEA 1,

Keywords: Information model, Power continuity indicators, Spanning tree,


Connectivity status function, Data production and aggregation areas
This paper presents an information model conceived to support the computation of the
power continuity indicators. The model consists of three layers: the physical layer, the
data layer and the application layer. A tree-like structure that spans the graph structure
of the power grid and the hierarchical structure of the power distribution branch is used
to describe the relationships between this branch and its consumers. The model allows
the computation of the quality indicators at various granularity levels by defining a
number of monitored areas that span sub-trees of the spanning tree of different sizes.
For the description of the connectivity state of the consumer to the power grid during
the monitored period of time a status function is also defined.

1. INTRODUCTION
The quality of the power supply service, especially for the industrial
consumers, has an important influence on the economic performance of a company.
Although all power quality indicators have their own importance for the
evaluation of the performance level of a power network operator, among them the
information about the short- and long-term power interruptions is of outmost
importance due to the fact that these events have the greatest impact on the
consumer.
In this respect, the Performance standard for the electrical energy distribution
service issued by ANRE (The National Energy Regulating Agency) defines a series
1

) Felix IT SA, str. Ing. George Constantinescu nr. 2, sector 2, cod 020339, Bucureti, tel: 021242.01.92, fax: 021-242.70.41, e-mail: vrascanu@yahoo.com
2
) ISPE SA Bucureti, Bd. Lacul Tei nr. 1-3, cod 020371, sector 2, tel: 021-206.13.07, fax: 021206.13.07, e-mail: halbert@ispe.ro
3
) Facultatea de Energetic, U.P. Bucureti, Splaiul Independenei nr. 131, cod 060042, sector 6, tel.:
021-316.96.43, fax: 021-316.96.45, e-mail: nicolae_golovanov@yahoo.com
Rev. Roum. Sci. Techn. lectrotechn. et nerg., 54, 1, p.

, Bucarest, 2009

12

Rcanu Valentin Grigore et al.

of quality indicators regarding the frequency and the duration of the power supply
interruptions. These indicators describe:
- the power supply continuity of a consumer;
- the power supply continuity of a residential area (e.g. : transformer
points);
- the power supply continuity of a busbar (node) of the electrical grid;
- the power supply continuity of the consumers of an electrical energy
supplier (at the LV, MV and HV levels).
The computation of the power quality indicators requires the continuous
acquisition and processing of the data from the monitored nodes of the power grid.
At present, only particular solutions for data acquisition and computation of
the power supply continuity indicators are available on the Romanian market.
These solutions depend on the type of the equipment being used for the data
acquisition and on the particular structure of the distribution branch and are
difficult to modify, reuse and even reconfigure.
This paper presents an information model that solves the problem presented
above by using a general purpose tree-like structure that describes the relationship
between the customer and the distribution branch for the computation of the power
supply continuity indicators.
2. THE POWER GRID STRUCTURE
The interconnected power system propagates the perturbations caused by
different events to all the grid supply points. The impact of the perturbation
depends on the short-circuit current in the analyzed node and on the impedance
between this node and the location where the event took place.
Basically, the structure of an electrical distribution network, depicted in
figure 1, includes [1, 2]:
- the 110kV network (usually containing loops) which supplies electrical
power to the 110/20kV substations (level I);
- the 110 kV/20kV substation which supply electrical power to the
consumers that are directly connected (usually in a radial network) to the
110kV busbar or to the consumers connected to the 20kV busbars by
means of the 110kV/20kV transformers (level II);
- the 20kV network (usually having a radial or a looped topology) which
supplies electrical power either directly to the consumers connected to it
or to the consumers connected by means of the 20/0.4 kV transformers.
Starting form the 20kV busbars, the energy may be fed to the feed points
too (level III);

Information model for the computation of the power supply continuity indicators

11

the LV network (mainly a radial network), which supplies electrical


power to the small consumers from the residential or rural areas (level
IV).
110 kV

110 kV
110 kV

20 kV
PT

PT
PA

PA
PT

PA

S
S

Fig. 1 The structure of the power distribution grid

The data required for the computation of the power supply continuity
indicators, at each of these voltage levels, are the number of power interruptions,
the duration of these interruptions, the number of the affected consumers and the
interrupted power. It is also important to know if the interruption was planned or it
happened by accident, knowing that in case of the planned interruptions the costs
can be avoided.
In a radial network, these data can be directly acquired either from the
consumers (level IV), from the transformer points (level III), from the 20kV
busbars of the distribution substation (level II) or from the 110kV bus bars of the
substation (level I).
We have to take into account, when we analyze the acquired data, the use on
the 20kV busbars, for power reservation purposes, of the automatic throw over
switches (ATS) and of the isolating switches used to configure a loop in a radial
network.

12

Rcanu Valentin Grigore et al.

3. THE POWER SUPPLY CONTINUITY INDICATORS


According to the present legislation, every distribution branch must compute
the yearly values of the continuity indicators SAIDI, SAIFI and AIT and of the
ENS quantity.
The SAIFI indicator (The System Average Interruption Frequency Index) is
commonly used as a reliability indicator by electric power utilities. SAIFI is the
average number of interruptions with duration longer than 3 minutes that a
customer would experience. It is usually measured over the course of a year,
usually taking values that are smaller than 2, and it is calculated as:
SAIFI = (Nk=1 The number of interrupted customers during the k
interruption) / Total number of customers served

(1)

where N is the number of power supply interruptions during a year.


SAIFI is measured in units of interruptions per customer.. The calculation of
this continuity indicator requires information about the number of customers
connected to every feeder leaving a substation or a feed point and who experience
interruption as the respective feeders are disconnected from the power system.
The SAIDI indicator (The System Average Interruption Duration Index) is
another indicator commonly used as a reliability indicator by electric power
utilities. SAIDI expresses the average outage duration for each customer served,
and is calculated as:
SAIDI = (Nk=1 The duration of the k interruption x The number of
interrupted customers during the k interruption) / Total number of
customers served

(2)

where N is the number of power supply interruptions during a year.


SAIDI is measured in units of time, often minutes or hours. It is usually
measured over the course of a year and it requires information about the time the
circuit breakers stayed tripped-off during every interruption. The SAIDI indicator
could take values that are less than 200 minutes pro year.
The ENS (Energy Not Supplied) quantity may be determined using the
following formula:
ENS = (Nk=1 The power interrupted during the k interruption x The
duration of the k interruption)

(3)

where N is the number of power supply interruptions during a year.


The power that is interrupted may be determined using the information
delivered by the power meters, if they exist or else a constant value, like the power
to be delivered by contract, may be used.

Information model for the computation of the power supply continuity indicators

11

The AIT indicator (Average Interruption Time) may be determined using the
formula:
AIT = 8760 * 60 * ENS / AD

(4)

where AD (Annual Demand) its the total energy demanded by the consumers
over an entire year (MWh/year, no losses), while the ENS (Energy Not Supplied)
unit of measurement is the MWh.
The calculation of the AIT indicator its based on the calculated ENS value,
while the AD value it is computed using the value of the power level that was
contracted by the consumers.
For the computation of the above-mentioned power supply continuity
indicators we need an information model that will allow us to systematically tackle
the complex structure of the electrical grid.
The proposed information model covers the HV and MV substations
belonging to a distribution branch. Although it has been conceived to support the
computation of the continuity indicators at the branch or even at company level, the
model is highly scalable and allows to computation these indices at the desired
granularity level (distribution branch, substation, busbar or feeder).
4. THE PRESENTATION OF THE INFORMATION MODEL
The information model (figure 2) consists of three layers: the physical layer,
the data layer and the application layer. Besides these three layers there is a fourth
one, the user layer, which doesnt actually belong to the model. It is nonetheless
useful for the description of the way in which the continuity indicators are used by
the users in the problem domain.
A layer provides for data services to the layer above it and uses the data
services of the layer below it.
The physical layer is implemented at the electrical grid level by means of
which the electrical power is supplied to the consumers. This grid, which is often
referred as a complex system, can be described from a topological point of view by
a graph. The data types used by the model at this layer are the primary data (e.g.:
r.m.s. values of the voltage magnitude).
The data layer is dedicated to the production of the events that describe the
continuity of the power supply to the consumer. These events are produced the
moment when the r.m.s. value of the supply voltage magnitude goes under the
separation threshold Vsep for more than 3 minutes (where Vsep= 0.05 * UN).
User layer

12

Rcanu Valentin Grigore et al.

Application layer
Data layer
Physical layer
Fig. 2 The information model used for the computation of the continuity

The application layer is the place where the acquired events are processed
and aggregated and the continuity indices are computed.
The user layer is the place where the data coming from the application layer
are being used in decision systems in order to facilitate the solving of the problems
caused by the poor electrical energy quality or to generate reports.
The primary data from the physical level are obtained by continuously (every
half-period) measuring the r.m.s. value of the supply voltage magnitude. In order to
determine the events related to the power supply interruptions for the set of
consumers of a distribution branch connected at a certain voltage level C={c 1, c2,
,cn}, we attach each of these consumers a connectivity status function S=f(ui).
The function S: R -> {0,1} is defined as following:
S(u) = 0, u < 0.05 * Un
S(u) = 1, u >= 0.05 * Un
where:
u=f(t), u:R->R is a function of time that gives the r.m.s. value
of the supply voltage magnitude in the feed node;
Un= the r.m.s. value of the contracted voltage magnitude.
The S function, defined on the R set and taking values in the {0,1} set,
indicates if at any particular moment of time the consumer is connected to the
power grid (taking the 1 or true value) or not (taking the 0 or false value).
We can say that the set of the values these functions take at the time moment
t=t1 {S (u1 (t=t1)), S (u2 (t=t1)) S (un (t=t1))} completely defines the connectivity
status of the members of the C set at that particular moment. We can conclude that,
if we are going to record the all the values that the S i functions took for a particular
period of time, we can determine the power supply interruptions by selecting those
situations when the function S took the 0 value for more than 3 minutes.
Function value
0

Moment of change
[dd mmm yyyy hh:mm:ss.ms]
10 May 2008 22:43:59.250

12 May 2008 12:23:45.120

Duration [ms]
200.000

123

Due to the fact that the S functions can take only to values (0 or 1), we need
to record only the moments when the function change its output value together
with the new value of the function and the change duration.

Information model for the computation of the power supply continuity indicators

11

In the table above, the rows that record the moments when the function S
takes the value 0 for more than 3 minutes are of particular interest, because they
represent the events of the interruption of the power supply to the customer.
In order to compute the continuity indicators at the application layer, we need
to define a topological structure that will match both the physical structure of the
electrical grid (it has to include the nodes that produce the events which are used in
the computation of the continuity indicators) and the organizational structure of the
power distribution system (it has to describe the relationships between the entities
of this system and the consumer). The most adequate solution for describing the
relationships between the distribution branch and the consumer is the use a tree
structure named the spanning tree (figure 3). This name stems from the fact that
this tree spans the entire above mentioned structures.

Fig. 3 The structure of a spanning tree of a distribution branch

A spanning tree is a hierarchical structure that places elements in nodes along


branches that originate from a unique node called the root. The nodes in a tree are
subdivided into levels, where the topmost level holds the root node. Any node in a
tree can have multiple successors at the next level, excepting the nodes belonging
to the last level (the leaf nodes). Each node in a tree is the root of a subtree, which
consists of the respective node and all of its descendants.
Using analogies from a family tree, we associate the terms parent and
children to describe the relationship between a node and its successor nodes. The

12

Rcanu Valentin Grigore et al.

fact that each nonroot node has a single parent ensures that there is a unique path
from any node to itself or one of its descendants. Practically, in the case of the tree
structure, starting from the root node we can visit all the nodes exactly one time.
The spanning tree is made of nodes that represent either physical objects (the
busbars and the circuit breakers), which are data producers, or abstract objects (the
distribution company, the distribution branch, the substation) which are data
aggregators.
The root node is allocated to a data aggregator (the distribution branch,
depending where we start the hierarchy), the intermediate nodes hold data
aggregators (the distribution branch, the substation, and the busbars) and the leaf
nodes hold data producers (the circuit breakers which connect the consumer to the
network grid). The spanning tree is described from the point of view of the
relationships between nodes (has a or subordination relationships) by a number
of configuration files.
This storage of the events takes place in the leaf nodes of the spanning tree
which partially overrides the graph structure of the electrical grid, including only
those vertices that directly participate in the event production (the busbars and the
circuit breakers). The line that connects the leaf nodes of the spanning tree
represents practically the border between the consumers and the distribution
network operator (figure 4).

Fig. 4 The structure of a distribution branch viewed from the perspective of its relationship with the
customers

The physical nodes are monitored by means of specialized equipments that


continuously acquire real-time data regarding the continuity of the supply service,

Information model for the computation of the power supply continuity indicators

11

data which is transformed in events that are eventually stored in historical


databases. The structure of the historical database records that store these events
represents an interface that separates a variable domain (the data format) from the
rest of the model. The model allows the import of data that are acquired in other
formats.
Two assumptions have been made regarding the decision of wether a vertex
of the electrical grid graph is included in the spanning tree or not.
First, taking into account that the absence of the supply voltage at any point
on the path between the busbar and the consumer automatically leads to the
absence of supply voltage at the consumer connection point, no intermediate vertex
(usually implemented by a circuit breaker) belonging to this path shall be included
in the tree.
Second, the path between the busbar and the circuit breaker that connects the
consumer to the electrical grid represents an abstract relationship (although it is
based on a physical reality). So, from the model standpoint, it is immaterial
whether the power is feeded from the respective busbar or from another power
source (as in the case when we reconfigure the network by building a loop).
One important attribute of the leaf nodes is their multiplicity. The node
multiplicity represents the number of the consumers directly connected to that node
at the same voltage level. Keeping this thing in mind, a MV substation represents
only one consumer for a HV substation, no matter how many MV consumers are
connected to the first one. This fact holds for a MV substation with respect to the
LV consumers connected to a feed/transformer point too (they count as a single
consumer at the MV level to the substation).
The abstract nodes contained by the spanning tree belong to the
organizational structure of the power distribution system and they include the
electrical substation, the distribution branch or the distribution company. These
nodes do not directly produce events and they are included in the spanning tree
because, at the application layer, we need them to compute the continuity
indicators.
From the point of view of their position to the information production and
consumption processes, we can divide the nodes of the spanning tree between
servers (data producers) and clients (data consumers). In this respect, the leaf
nodes are servers, the intermediate nodes are both servers (the busbars) and clients
(the substations and the distribution branches) and the root node is a client (the
distribution branch or the distribution company).
The data collection which a server presents to its clients defines the address
space of that server. These data can be either attributes, which are primitive types
of data (e.g.: r.m.s. value of the voltage magnitude or mains frequency values), or
complex data types consisting of primitive data types and which have a certain
structure (e.g.: the events which record the power interruptions).

12

Rcanu Valentin Grigore et al.

10

Taking into account this point of view, the circuit breakers from the
substations or from the feed points form together a so called data production zone
(DPZ), while the busbars, the substations and the distribution branches which use
these data in order to compute the continuity indicators form around them a data
aggregation zone (DAZ). Both the DPZ and the DAZ contain subtrees of the
spanning tree. The root node in case of the DPZ is a busbar and in case of the DAZ
is a substation, a distribution branch or a distribution company, depending on the
area covered by this zone.

Fig. 5 The relationship between the structure of the spanning tree and the way data are produced
and consumed.

The correspondence between the structure of the spanning tree (figure 3) and
the way the data are produced and consumed during the power supply continuity
monitoring and computation processes are presented in figure 5.
The relationships between the spanning tree nodes and the way they belong
to the data production and aggregation zones are depicted by a number of
configuration files.
The configuration files store static information about the subtrees of the
spanning tree. The model uses two type of configuration files: the data production
zone (DPZ) configuration file which stores information about the nodes belonging

11

Information model for the computation of the power supply continuity indicators

11

to such a zone (e.g.: the circuit breakers) and the data aggregation zone (DAZ)
configuration file which stores information about the data production zones that are
included in a data aggregation zone (e.g.: the substations).
The DPZ configuration file describes a certain subtree of the spanning tree
which contains physical nodes and which has two layers: the root node (busbar)
layer and the leaf nodes layer (the circuit breakers which the customers are directly
connected to).
The nodes of the subtree depicted by a DPZ configuration file are described
by a number of entries containing an unique identifier (e.g.: the serial number of
the monitoring equipment and the identifier of its input channel connected to the
node), the number of the consumers connected to the node, the power that has to be
supplied to these users by contract, the area type to which the customer belongs
(urban or rural area), the voltage level supplied to the customer (HV or MV), the
path to the historical database where the dynamic information belonging to the
node is stored, the name of the data production node and the name of the respective
node.
The DAZ configuration file includes information about the DPZs that belong
to the respective data aggregation zone, this information being described by the
entries included the file. The file contains one entry for each DPZ, each entry
including the path to the database that stores the dynamic information acquired in a
DPZ, the path to the DPZ configuration file, the name of the DPZ and the name of
the DAZ.
The configuration files represent another interface that separates another zone
of variability (the specific structure of a distribution branch) from the rest of the
model.
In order to compute the power supply continuity indicators at the application
layer we need to use, besides the static information contained in the configuration
files, the dynamic information from the historical databases that store the data
acquired form the monitored grid nodes by the specialized equipments.
The power supply continuity indicators can be computed at different
granularity levels by reading the static information contained in the configuration
files and by corroborating this information with the one contained in the historical
databases.
5. CONCLUSIONS
The information model presented in this paper uses a general purpose treelike structure that describes the relationship between the customer and the
distribution branch, for the computation of the power supply continuity indicators.

12

Rcanu Valentin Grigore et al.

12

Compared with the solutions that are developed or used at present in the
distribution sector in Romania, this solution has the following advantages:
- proposes a general purpose information model for the computation of the
power supply continuity indicators;
- separates the variable part of the model (the input data format and the
distribution branch structure) being a number of clearly defined
interfaces;
- allows the description of the relationship between the distribution branch
and the consumer by means of an easily to be transversed tree structure;
- offers a flexible mechanism for the description, by means of this tree
structure, of the particular topologies of the distribution branches;
- the proposed model is scalable, allowing the computation of the power
supply continuity indicators at the granularity level chosen by the user
(distribution branch, substation, busbar, feeder).
The information model was experimentally implemented in an integrated
power quality monitoring system deployed in an Electrica distribution branch in
Romania
Received on (month day, year)

REFERENCES
1. Hermina Albert, I. Florea, Alimentarea cu energie electric a ntreprinderilor industriale, Editura
Tehnic, Bucureti, 1987.
2. N. Golovanov, .a., Instalaii electroenergetice i elemente de audit industrial, Editura N'ERGO,
Bucureti, 2008.

Potrebbero piacerti anche