Chapter 3
Optimal Protective Device Scheme for Radial Distribution Networks with Distributed
Generation using Binary Linear Programming
3.1 Introduction
In this research, the programs used were LINGO Software, and Microsoft Excel. The
researchers were able to develop test cases by locating the protective devices in every section of
the bus system. The location of the protective devices will be optimized with two fixed DGs
using the Binary Integer Programming. The types of DG used in the bus system are Wind
Turbine Generator (WTG) and Solar Cell Generator (SCG) in which their location supplies the
loads completely in an island [1].
The algorithm will be done using Microsoft Excel for the optimization process of the
reliability indices. For the verification of results, LINGO software will be used. Moreover,
Microsoft Excel will also be used in conducting T-test for the comparison of the analytical results
obtained using the two different methods. Using again the Microsoft Excel, the cost of the
project based on the reliability index EENS will be computed. Having the least amount of cost
and most optimal location of the protective devices will be chosen as the best case of the
research.
3.2 Research Flow Chart
The flows of the study are described as follows. First, the researchers develop test cases
for the IEEE RBTS bus-6 system having two fixed DGs and protective devices. The protective
devices included in this research are circuit breakers, reclosers, and fuses. The fuses are
classified into two: fuse-save and fuse blown. Using all possible combination of two fixed DGs
as it turns on and off and all possible locations of the protective devices for every section, test
cases will depend on this combination.
For the binary formulation of the indices, SAIFI and EENS will be developed as the
objective function. On the other hand, the formulation of the constraints should be considered in
terms of inequalities. These formulations will be used to obtain the optimum location of the
protective devices having fixed DGs with the use of Binary Linear Programming. Also, the
researchers used the LINGO software for the verification of results. T-test will be conducted to
compare the two results with different method using Microsoft Excel. The cost of the project will
be computed and the best case will be chosen. The flow of the research was shown in figure 3.2.1
Figure 3.2.1 Research Flow Chart
Start
Develop test cases for the
system
Optimize reliability indices
using the binary
programming technique
Verify results using
LINGO software
Statistical Analysis of
Results
Compute for Cost
Select the Best Case
End
3.2.2 Test Case of the System
The researchers will analyze the IEEE RBTS Bus6 system as their test case with the aim
of minimizing the reliability indices through determining the optimal device placement.
Figure 3.2.2 Test System
As seen in the distribution system, there are 23 load points, 23 distribution transformers
and two DGs. Section 1 and those nearest to the DGs are installed with circuit breakers while
sections 10, 16, 19, and 25 are automatically installed with reclosers. The analysis of the test
system will be simplified by dismantling portions from the system and analyzing them
separately. The aim is to optimize each portion. The final optimized results per portion will be
combined to achieve the overall reliability indices values and the optimal positioning of devices.
The test system has the reliability data per given component. The transformers in the
system, feeder lines, and load points have their reliability data given in the tables below.
Table 3.2.1 Reliability Data of the Components
Table 3.2.2 Reliability Data of the Protection Devices
The given length data per line is multiplied to the failure rate data of the lines since the
required unit for the computation is given by failures per year.
Table 3.2.2 Feeder Line Data
Each load point has different profiles with respect to the peak load and the number of
customers present. The load point data gives the factors used in computing the reliability indices.
The load is a factor for EENS while the number of customers is a factor of SAIFI.
Table 3.2.3 Load Point Data
The system is comprised of five sections as divided by the predetermined reclosers.
Failure rate, repair time, and unavailability hours are given on the table below. These data
provide the said parameters of each section present in the system. These also provide the factors
needed in computation for the reliability indices.
Table 3.2.4 Reliability Indices of Some Load Points
As seen on the table above, there are only five load point data provided. However,
looking at the test system, these load points are present per section. Therefore, the failure rate,
repair time, unavailability hours, and the ENS will be constant for load point 1 up to load point 7.
The same manner is applied to other load points.
Section one will cover the portion of the system from section 1 up to section 9. It has
seven load points, each of which have the distribution transformers installed.
Figure 3.2.3 Section 1 of the Test System
Section 2 will cover the sections 11 to 15. This section only has three load points with
distribution transformers.
Figure 3.2.4 Section 2 of the Test System
Section 3 is connected from section 1 as section 2 is. This section is connected through a
recloser at section 19. This section has five load points wherein each has a separate distribution
transformer connected to it.
Figure 3.2.5 Section 3 of the Test System
Section 4 will cover sections 16 to section 18. At the end of this section is a DG which
can switched on or off. Case studies will include the switching of the DGs present in the system.
This section has three load points and three distribution transformers.
Figure 3.2.6 Section 4 of the Test System
Section 5 will cover sections 25 up to section 30. Similar to the third section, this section
also has a DG installed to it. It can also be switched on or off like the DG of Section 3. This
section has five load points with transformers.
Figure 3.2.7 Section 5 of the Test System
The results table consists of 10 columns. The cases are described by the configuration of
the DGs, organization of protective devices, results under the SAIFI and EENS indices as well as
the computed cost of operation. Protective devices in each case are provided with their respective
locations. There are numerous combinations applicable to the test system represented as the
cases. In these tables, only 30 cases of those possible combinations are included.
The numbers under the Recloser, fuse-blown, and fuse-save columns are the location to
where the stated devices are to be installed. The None column represents the sections where
there is no device to be installed. If the case has reclosers set at the laterals, this is represented by
the load point number with a capital L.
For Sections 1,2 and 3, the DG will be assumed as off since it will not be affected unless
the sections with DGs are connected upon analysis.
Table 3.2.5: Section 1 Reliability and Cost Data
Case DG-1 DG-2 Recloser Fuse- Fuse- None SAIFI EENS Cost
Blown Save
1 OFF OFF 2,3 1,2,3 4,5,6,7 4,5,6,8
2 OFF OFF 2,4 1,2,3 4,5,6,7 3,5,6,8
3 OFF OFF 2,5 1,2,3 4,5,6,7 3,4,6,8
4 OFF OFF 2,6 1,2,3 4,5,6,7 3,4,5,8
5 OFF OFF 2,8 1,2,3 4,5,6,7 3,4,5,6
6 OFF OFF 2,3 4,5,6,7 1,2,3 4,5,6,8
7 OFF OFF 2,4 4,5,6,7 1,2,3 3,5,6,8
8 OFF OFF 2,5 4,5,6,7 1,2,3 3,4,6,8
9 OFF OFF 2,6 4,5,6,7 1,2,3 3,4,5,8
10 OFF OFF 2,8 4,5,6,7 1,2,3 3,4,5,6
11 OFF OFF 4,5 1,2,3 4,5,6,7 2,3,6,8
12 OFF OFF 4,6 1,2,3 4,5,6,7 2,3,5,8
13 OFF OFF 4,8 1,2,3 4,5,6,7 2,3,5,6
14 OFF OFF 4,5 4,5,6,7 1,2,3 2,3,6,8
15 OFF OFF 4,6 4,5,6,7 1,2,3 2,3,5,8
16 OFF OFF 4,8 4,5,6,7 1,2,3 2,3,5,6
17 OFF OFF 6,8 1,2,3 4,5,6,7 2,3,4,5
18 OFF OFF 6,8 4,5,6,7 1,2,3 2,3,4,5
19 OFF OFF 2,1L 2,3,4 5,6,7,8 3,4,5,6,8
20 OFF OFF 2,2L 1,3,4 5,6,7,8 3,4,5,6,8
21 OFF OFF 2,3L 1,2,4 5,6,7,8 3,4,5,6,8
22 OFF OFF 2,4L 1,2,3 5,6,7,8 3,4,5,6,8
23 OFF OFF 2,5L 1,2,3 4,6,7,8 3,4,5,6,8
24 OFF OFF 2,6L 5,6,7,8 2,3,4 3,4,5,6,8
25 OFF OFF 2,7L 5,6,7,8 2,3,4 3,4,5,6,8
26 OFF OFF 2,1L 2,3,4 5,6,7,8 3,4,5,6,8
27 OFF OFF 2,2L 1,3,4 5,6,7,8 3,4,5,6,8
28 OFF OFF 2,3L 2,3,4 5,6,7,8 3,4,5,6,8
29 OFF OFF 2,4L 2,3,4 5,6,7,8 3,4,5,6,8
30 OFF OFF 2,5L 2,3,4 5,6,7,8 3,4,5,6,8
Table 3.2.6: Section 2 Reliability and Cost Data
Case DG-1 DG-2 Recloser Fuse- Fuse- None SAIFI EENS Cost
Blown Save
1 OFF OFF 12 - 8,9,10 13,14
2 OFF OFF 12 8 9,10 13,14
3 OFF OFF 12 9 8,10 13,14
4 OFF OFF 12 10 8,9 13,14
5 OFF OFF 12 8,9 10 13,14
6 OFF OFF 12 8,10 9 13,14
7 OFF OFF 12 9,10 8 13,14
8 OFF OFF 12 8,9,10 - 13,14
9 OFF OFF 13 - 8,9,10 12,14
10 OFF OFF 13 8 9,10 12,14
11 OFF OFF 13 9 8,10 12,14
12 OFF OFF 13 10 8,9 12,14
13 OFF OFF 13 8,9 10 12,14
14 OFF OFF 13 8,10 9 12,14
15 OFF OFF 13 9,10 8 12,14
16 OFF OFF 13 8,9,10 - 12,14
17 OFF OFF 14 - 8,9,10 13,14
18 OFF OFF 14 8 9,10 13,14
19 OFF OFF 14 9 8,10 13,14
20 OFF OFF 14 10 8,9 13,14
21 OFF OFF 14 8,9 10 13,14
22 OFF OFF 14 8,10 9 13,14
23 OFF OFF 14 9,10 8 13,14
24 OFF OFF 14 8,9,10 - 13,14
25 OFF OFF 8L - 9,10 12,13,14
26 OFF OFF 8L 9,10 - 12,13,14
27 OFF OFF 9L - 8,10 12,13,14
28 OFF OFF 9L 8,10 - 12,13,14
29 OFF OFF 10L - 8,9 12,13,14
30 OFF OFF 10L 8,9 - 12,13,14
Table 3.2.7 Section 3 Reliability and Cost Data
Case DG- DG- Recloser Fuse- Fuse-Save None SAIFI EENS Cost
1 2 Blown
1 OFF OFF 20,22 - 14,15,16,17,18
2 OFF OFF 20,22 14 15,16,17,18
3 OFF OFF 20,22 15 14,16,17,18
4 OFF OFF 20,22 16 14,15,17,18
5 OFF OFF 20,22 17 14,15,16,18
6 OFF OFF 20,22 18 14,15,16,17
7 OFF OFF 20,22 14,15 16,17,18
8 OFF OFF 20,22 14,16 15,17,18
9 OFF OFF 20,22 14,17 15,16,18
10 OFF OFF 20,22 14,18 15,16,17
11 OFF OFF 21,23 15,16 14,17,18
12 OFF OFF 21,23 15,17 14,15,16
13 OFF OFF 21,23 15,18 14,16,17
14 OFF OFF 21,23 16,17 14,15,18
15 OFF OFF 21,23 16,18 14,15,17
16 OFF OFF 21,23 17,18 14,15,16
17 OFF OFF 21,23 16,17,18 14,15
18 OFF OFF 21,23 15,17,18 14,16
19 OFF OFF 21,23 15,16,18 14,17
20 OFF OFF 21,23 15,16,17 14,18
21 OFF OFF 22,24 14,17,18 15,16
22 OFF OFF 22,24 14,15,16 15,17
23 OFF OFF 22,24 14,16,17 15,18
24 OFF OFF 22,24 14,15,18 16,17
25 OFF OFF 22,24 14,15,17 16,18
26 OFF OFF 22,24 14,15,16 17,18
27 OFF OFF 22,24 14,16,17,18 15
28 OFF OFF 22,24 14,15,17,18 16
29 OFF OFF 22,24 14,15,16,18 17
30 OFF OFF 22,24 14,15,16,17 18
Table 3.2.8 Section 4 Reliability and Cost Data
Case DG-1 DG-2 Recloser Fuse- Fuse- None SAIFI EENS Cost
Blown Save
1 OFF OFF 16 - 11,12,13 17,18
2 OFF ON 16 11 12,13 17,18
3 OFF OFF 16 12 11,13 17,18
4 OFF ON 16 13 11,12 17,18
5 OFF OFF 16 11,12 13 17,18
6 OFF ON 16 11,13 12 17,18
7 OFF OFF 16 12,13 11 17,18
8 OFF ON 16 11,12,13 - 17,18
9 OFF OFF 17 - 11,12,13 16,18
10 OFF ON 17 11 12,13 16,18
11 OFF OFF 17 12 11,13 16,18
12 OFF ON 17 13 11,12 16,18
13 OFF OFF 17 11,12 13 16,18
14 OFF ON 17 11,13 12 16,18
15 OFF OFF 17 12,13 11 16,18
16 ON ON 17 11,12,13 - 16,18
17 ON OFF 18 - 11,12,13 16,17
18 ON ON 18 11 12,13 16,17
19 ON OFF 18 12 11,13 16,17
20 ON ON 18 13 11,12 16,17
21 ON OFF 18 11,12 13 16,17
22 ON ON 18 11,13 12 16,17
23 ON OFF 18 12,13 11 16,17
24 ON ON 18 11,12,13 - 16,17
25 ON OFF 11L - 12,13 16,17,18
26 ON ON 11L 12,13 - 16,17,18
27 ON OFF 12L - 11,13 16,17,18
28 ON ON 12L 11,13 - 16,17,18
29 ON OFF 13L - 11,12 16,17,18
30 ON ON 13L 11,12 - 16,17,18
Section 4 of the test system has a DG connected at its end. DGs behave differently with
respect to its configuration, whether it is on or off. For further analysis on the reliability indices
as affected by the DGs, the researchers will include its on and off configuration.
Table 3.2.9 Section 5 Reliability and Cost Data
Case DG- DG- Recloser Fuse- Fuse-Save None SAIFI EENS Cost
1 2 Blown
1 OFF OFF 26,28 - 19,20,21,22,23
2 OFF ON 26,28 19 20,21,22,23
3 OFF OFF 26,28 20 20,21,22,23
4 OFF ON 26,28 21 19,20,22,23
5 OFF OFF 26,28 22 19,20,21,23
6 OFF ON 26,28 23 19,20,21,22
7 OFF OFF 26,28 19,20 21,22,23
8 OFF ON 26,28 19,21 20,22,23
9 OFF OFF 26,28 19,22 20,21,23
10 OFF ON 26,28 19,23 20,21,22
11 OFF OFF 27,29 20,21 19,22,23
12 OFF ON 27,29 20,22 19,21,23
13 OFF OFF 27,29 20,23 19,21,22
14 OFF ON 27,29 21,22 19,20,23
15 OFF OFF 27,29 21,23 19,20,22
16 ON ON 27,29 22,23 19,20,21
17 ON OFF 27,29 21,22,23 19,20
18 ON ON 27,29 20,22,23 19,21
19 ON OFF 27,29 20,21,23 19,22
20 ON ON 27,29 20,21,22 19,23
21 ON OFF 29,30 19,22,23 20,21
22 ON ON 29,30 19,21,23 20,22
23 ON OFF 29,30 19,21,22 20,23
24 ON ON 29,30 19,20,23 21,22
25 ON OFF 29,30 19,20,22 21,23
26 ON ON 29,30 19,20,21 22,23
27 ON OFF 29,30 20,21,22,23 20
28 ON ON 29,30 19,20,22,23 21
29 ON OFF 29,30 19,20,21,23 22
30 ON ON 29,30 19,20,21,22 23
Achieving the optimal placing of devices per section of the test system, the researchers
are to combine the optimal results to determine the overall placing of the system as well as its
reliability indices and the cost of operation. This overall result is determined using the final table.
Table 3.2.10 Overall Reliability and Cost Data
Case DG-1 DG-2 Recloser Fuse- Fuse- None SAIFI EENS Cost
Blown Save
1 OFF OFF
2 OFF ON
3 ON OFF
4 ON ON
3.2.3 Binary Linear Programming Flowchart
Binary Integer Programming is a type of method under Linear Programming that
determines the maximum or minimum value of the objective function that is subjected to pre-
determined constraints, which are represented by inequality equations. The Binary Integer
Programming Flowchart is shown in Figure 3.2.8.
Figure 3.2.8 Binary Integer Programming Flowchart
The researchers are to focus only on minimizing the indices SAIFI and EENS. The binary
formulation of the SAIFI and EENS is a function of the location of the protective devices with
two fixed DGs. This formulation is the objective function of the optimization problem. Also, the
researchers should model the distribution system by formulating the constraints using such
variables. The constraints define the requirements of the test system through the coordination,
quantity, and organization of protective devices to be used in the researchers implementation.
The researchers aim to generate these constraints with the goal of minimizing the reliability
indices with respect to determining the optimal location for the protective devices installation.
Coordination constraints come into play when there are two or more devices normally
reclosers, that cannot achieve proper coordination throughout their operation. In other words,
these constraints help avoid such circumstances that may generate time operation delays due to
incoordination of the said devices. In the study, the researchers will constraint specific sections
on the test system where reclosers if installed, are assumed to be uncoordinated. This constraint
is to be used once per section.
Quantity constraints are present when a specific quantity is to be met by the optimization
process. For example, only two line reclosers are to be used all throughout the section, the fuses
are unlimited, etc. in the implementation study of the researchers, this type of constraints are to
be used only on reclosers for the allowable quantity to be used per section of the test system.
Fuses on the other hand is assumed to be unlimited.
Organization constraints are observed when a specific device is already installed at a
specific location. An example of this constraint is having a circuit breaker installed at the nearest
distribution section from the power grid or from the substation bus. These constraints are also
present when a device is not allowed at a certain location. In the study, the researchers will not
allow fuses to be installed at the main feeder and that every lateral must be installed with either a
recloser or a fuse.
The constraints to be utilized are summarized below:
Sections 1, 18 and 30 are automatically installed with a circuit breaker.
Fuses are not allowed at the main feeder.
All laterals must be installed with a protective device.
Not all sections on the main feeder must be installed with a protective device.
Reclosers at sections 4 and 5, 12 and 14, 17 and 18, 22 and 23 and, 28 and 29 are not
coordinated.
Only one recloser is allowed when the load points are three below.
Two line reclosers are allowed when the load point exceeds three.
The line reclosers that operate as sectionalizers are not included in the quantity
considered on installation.
There are no limitations for the quantity of fuses whether fuse-save or fuse-blown upon
installation.
All laterals are of category 3.
Using the objective function and constraints of the protective devices with fixed DGs, the
Binary Linear Programming would solve and determine the location of the protective devices
that would minimize SAIFI and EENS.
3.2.4 Lingo Software
After obtaining the results in Binary Programming, it will be verified using Lingo
Software. LINGO Software is a software tool designed to efficiently build and solve linear, non-
linear and integer optimization models. In LINGO Software, the researcher should create a
LINGO Model. A LINGO Model consists of Objective Function, Variables and Constraints.
Once the LINGO model has been input into the LINGO Model window, the model can be
solved. Figure 3.2.9 shows the LINGO Solver Status
Figure 3.2.9: Lingo Solver Status
If theres no error found, the LINGO would continue to the window that would produce the
optimal value of the objective function. As shown in Figure 3.2.10.
Figure 3.2.10: LINGO Solution Report
3.2.5 Statistical Analysis of Results
In determining the precision of the results, Statistical Analysis is used. The researchers
will use the T-test to compare the two results acquired from different methods. T-test would
determine if the outcome test is accepted or rejected. Using the Microsoft Excel, the researchers
could determine whether the null hypothesis is accepted or not. The significance tests provide
results within a predefined confidence level. If the P value is less than the significance level, the
hypothesis test is statistically significant.
Figure 3.2.5 MsExcel T-test interface
3.2.6 Computation of Cost and Selecting the Best Case
After the researchers optimized the locations of the protective scheme with fixed DGs,
the researchers will compute for the cost of the project such as the total cost of the materials and
each installation of the protective devices. The cost of the project depends on the reliability index
EENS. Using the Microsoft Excel, the cost can be computed. Moreover, the least amount of cost
with the optimum location of the protective devices will be selected as the best case in the
system.
Total Expenses
Total Expenses = Sum of Circuit Protection * Price
Sample Problem
Determine the SAIFI index for the given network.
Figure 3.3.2 Sample Test System
Table 3.2.5 Reliability Data for Sample Test System
The constraints are subjected as follows:
Only two line reclosers are available
There is no limitation on the number of fuses to be used
A proper coordination of reclosers at 13 and 14 is not possible
Fuses are not allowed on the main feeder
Assume fuses to be installed at locations 21, 31, and 41.
A breaker is placed at location 11.
All laterals are category three.
For uniformity of the analysis, variables use a three-digit subscript, the first two numbers of
it defines the location while the third defines the device assigned. Subscript 1 is used for
reclosers while subscript 2 is for the fuses. The decision variable, x will define the optimal
location of the protective device. If the value of x is calculated to be 0, then a device is to be
installed at that particular location.
SAIFI index is calculated through the formula, .
The formulas numerator can be expanded with respect to its position. For the main feeder, the
numerator expansion is given by:
Figure 3.3.3 SAIFI Numerator Formulation for Main Feeder
For the laterals, it is given by:
Figure 3.3.4 SAIFI Numerator Formulation for Lateral Feeders
Using the Formulas given, we can calculate the SAIFI numerator for the problem. Substituting
the values from the table we can get,
For the Laterals, for lateral 21:
For lateral 31:
For lateral 41:
The overall numerator of the index is given as the sum of the expansions. Overall, the system
index numerator is given by:
The generated numerators will be our objective functions to be subjected for minimization
through the defined constraints.
For the constraints, it is said that only two line reclosers are available. This can be represented
by:
There is no limitation on the number of fuses that can be installed. However, fuses are not to be
installed at the main feeder. following the constraint that fuses are installed at each lateral
therefore
Coordination of line reclosers that are to be installed at locations 13 and 14 is not possible. The
constraint is represented by:
Some terms on the objective function contains two or three decision variable, it generates
product term constraints. We assign a single variable to represent them.
The product term constraints are given as the following:
Minimizing the objective function using Binary Integer Programming through the method of Implicit
Enumeration
Subject To:
(1)
(2)
(3)
(4)
(5)
(6)
(7)
(8)
(9)
(10)
(11)
(12)
Iteration no. 1
S1 = {} set all variables to zero
V1 = {1,2} list all violated constraints
T1 = {x121, x131, x141} list all positive variables from the violated constraints
Iteration no. 2
S1 = {x131} select the variable with the greatest vertical sum
V1 = {} list all violated constraints
The iteration can now be terminated since there are no violated constraints. However, it is not
sure whether the computed variable is the minimum of the possible outcomes.
Iteration no. 3 - Backtrack
/S1 = {x131}
V1 = {1,2} list all violated constraints
T1 = {x121, x141} list all positive variables from the violated constraints
Iteration no. 4
/S1 = {x131, x141}
V1 = {}
Iteration no.5 Backtrack
//S1 = {x131, x141}
V1 = {1,2} list all violated constraints
T1 = {x121} list all positive variables from the violated constraints
Iteration no.6
//S1 = {x131, x141, x121}
V1 = {} list all violated constraints
Since there are no violated constraints, the iteration process can be terminated. This will leave a
value of 1 to x121 and 0 for x131 and x141 thus, a recloser is to be installed at locations 13 and 14.
This will give a SAIFI numerator of 2862.5 dividing it to 955 as the total number of customers,
the SAIFI index is computed to be 2.9974.
Figure 3.3.5 Iteration Flow
However, a constraint is violated in the process. It is that reclosers positioned at 13 and
14 are not possible to be coordinated therefore it is necessary to consider the reclosers to be
positioned at locations 12 and 13 to achieve coordination between the said devices. This would
be the optimal positioning for the devices because it has a lower SAIFI index value than having
the reclosers at 12 and 14.
Table 3.2.5 SAIFI Results
Case Recloser Fuse None SAIFI
1 12, 14 21, 31, 41 13 3.3507
2 12, 13 21, 31, 41 14 3.0236
3 13, 14 21, 31, 41 12 2.9974
Since SAIFI index is maintained to be at the lowest possible value, it is to be concluded that
reclosers at positions 13 and 14 and fuses at locations 21, 31, and 41 is the most reliable
protective scheme.
