Experiment 5: Load Flow Study

September 10, 2024

Tuesday Batch: Group 1

Kushagr Goyal − 2021EE10634

Pramsu − 2021EE10140
Aniket Abhiraj − 2021EE10676
Shubham Kumar − 2020EE30627
Sudhanshu Raj − 2021EE10132

1
1 Objective
1. Load Ability Margin Determination
2. Line Over Load Alleviation using Generator Rescheduling
3. Reactive Rescheduling to Improve Voltage Profile
4. Line Over Load Alleviation using Series Compensation
5. Ferranti Effect

2 Theory
This is the single line diagram that has been used in this experiment.

Figure 1: Single Line Diagram for the Network

The loads in Power Systems are modelled as ZIP model to incorporate load changes
with the change in system voltage.

The load modelling can be done as:

PLi = PLoi ∗ (a i + biV + c iV 2) (1)

2
Q Li = QLoi ∗ (di + ei V + fi V ) (2)

where PLoi and Q Loi are the nominal load, a, b, c and d, e, f are coefficients for
constant power, constant current and constant impedance components respectively.
In our experiment, we have taken a=c=d=f=0. Hence, we have taken loads directly
proportional to voltage only.

2
2.1 Bus Injection Multiplier
To incorporate changes in loads/generations, load multiplier λ is used. The new load
at bus i is given by:

PLi = PLoi ∗ (1 + λ) (3)

Q Li = Q Loi ∗ (1 + λ) (4)
Pgi = Pgoi ∗ (1 + λ) (5)

where PLo, QLo andPgo are the base case variables in a given loading interval.

2.2 Load Ability Margin Determination

Newton Raphson finds the unknown parameters of the buses after running for many
iterations till the result is close enough to the desired value. It updates the values in each
step. But it may not be able to give feasible solution for all the starting values of input
variables.
Hence, depending on input values, it may converge or divergence may be encountered.
In this experiment, we vary the real load by varying value of λ and then find it’s critical
value.

2.3 Line Over Load Alleviation using Generator Rescheduling

Concept of Generation Shift Factor (GSF) was used in this part. By changing the
generation of bus 7, we can change the power flowing in the lines.
This was done to ensure that the power in the lines gets redistributed. This will reduce
the power flowing in the line with largest power, hence making the system relatively safer
to handle.

2.4 Line Over Load Alleviation using Series Compensation

If we add series capacitive compensation in transmission line, then the capacitance
balances the inductance part, reduces the reactance of the transmission line. This in turn
helps to increase the SIL (Surge Impedance Loading) that is the loadability limit.
Due to this, redistribution of power takes place amongst lines. The difference between
line carrying maximum power for λ = 0 and λ = 2 increases on increasing the series
compensation. This is because loadability on increasing compensation, leading to increase
difference in the two loading condition.

2.5 Reactive Rescheduling to Improve Voltage Profile

After solving Newton Raphson, we would want all the voltages to have value near
1pu. But the value of voltages might not be near 1 for all the buses. So to bring them
close to 1, we can do the following three things:
3
 1. Increase voltage at bus 7: The value of most buses is away from 1 (less than 1), so
to bring it closer to 1, we will increase the voltage at bus 7, which will increase the
voltage and bring it closer to 1.
2. Changing Tap Ratio: Increasing the tap ratio will have same effect as increasing
the voltage of bus-7.
3. Adding Shunt Capacitance: Adding shunt will provide reactive power in the system.
We know from theory that the reactive power depends on the voltage of the buses,
hence the voltage will increase on adding shunt and get closer to 1.

2.6 Ferranti Effect

The effect in which the voltage at the receiving end of the transmission line is more
than the sending voltage is known as the Ferranti effect. Such type of effect mainly occurs
because of light load or open circuit at the receiving end.
In our experiment, this can be drawn equivalent to PQ buses having more voltage
than PV buses as PV bus act as source and PQ buses act as load in the power system.

3 Procedure
3.1 Load Ability Margin Determination
We varied the value of λ to find it’s critical value, that is the smallest value of λ upto
2 decimal places for which the Newton Raphson diverges.

3.2 Line Over Load Alleviation using Generator Rescheduling

We observed the change in power flowing through the line with largest power when
value of λ was changed from 0 to 1.5. The same process was repeated by changing the
power generation of bus-7 from 0.4pu to 0.6pu. Difference between λ = 0 and λ =
1 . 5 was calculated to cross check that this difference should decrease with increase in
power generated.

3.3 Line Over Load Alleviation using Series Compensation

Find the difference of line with maximum power for loading conditions λ = 0 and λ = 1.
This is to be done for series compensation value X = 0 and X = 0.05. Then compare the
difference between these loading conditions for both the values of series compensation.

3.4 Reactive Rescheduling to Improve Voltage Profile

1. Increase voltage of PV bus 7 to 1.08V and then observe the improvement in voltages
2. Change the tap ratio to 1.08 at transformer between bus 6 and bus 7 and observe
the improvement in voltages
3. Add shunt j0.10 at bus 6 and run Newton Raphon to observe the improvement in
voltages

4
3.5 Ferranti Effect
Run Newton Raphson for λ = 0 (specified load), and then run it for λ = -0.5 (light
load). Compare the results for voltages obtained in the two cases.

4 Observations
4.1 Load Ability Margin Determination

Value of λ Diverge / Converge

0 Converge
5 Diverge
2.5 Converge
3.75 Diverge
3.12 Diverge
2.81 Diverge
2.66 Diverge
2.58 Diverge
2.54 Diverge
2.52 Diverge
2.51 Converge

We observe that the critical value of λ for Newton Raphson is 2.51.

4.2 Line Over Load Alleviation using Generator Rescheduling

• λ = 0, P = 0.4 at bus 7

Busi Busj Pij Qij Pji Qji

1 2 1.2683 -0.1840 -1.2683 0.3361
6 2 -0.3606 0.1381 0.3634 -0.2029
6 5 0.1606 -0.0259 -0.1585 -0.0237
2 5 0.3194 -0.0355 -0.3133 0.0053
2 4 0.3035 -0.0009 -2.9997 -0.0291
2 3 0.2822 -0.0968 -0.2793 0.0735
5 4 0.0217 -0.0816 -0.0217 0.0598
4 3 -0.0768 -0.0807 0.0793 0.0265
6 7 -0.4000 -0.2122 0.4000 0.2094

Table 1: Power Flow Values for λ = 0, P=0.4

Here, we can observe that the line with the largest power flow is line 1-2 with
P12 = 1.2683pu. Hence, we will monitor this line.

5
• λ = 1.5 P = 0.4 at bus 7

Busi Busj Pij Qij Pji Qji

1 2 3.2631 1.4862 -3.2631 0.1823
6 2 -1.1034 0.5678 1.1373 -0.5902
6 5 0.5034 0.1202 -0.4762 -0.087
2 5 1.0152 0.175 -0.9386 -0.0305
2 4 0.9345 0.2767 -0.8883 -0.1309
2 3 0.8892 -0.0438 -0.8568 0.1329
5 4 0.0648 -0.1825 -0.0644 0.1681
4 3 -0.2474 -0.1872 0.2568 0.1671
6 7 -1.2 -0.988 1.2 1.2474

Table 2: Power Flow Values for λ = 1.5, P=0.4

∆P12 = 3.2631 - 1.2683 = 1.9948pu for change from λ = 0 to λ = 1.5 at P=0.4

at bus 7.
• λ = 0, P = 0.6 at bus 7

Busi Busj Pij Qij Pji Qji

1 2 1.0664 -0.1987 -1.0664 0.3076
6 2 -0.1755 0.1031 0.1764 -0.17
6 5 0.1755 -0.0216 -0.173 -0.0269
2 5 0.3123 -0.0371 -0.3064 0.0063
2 4 0.2984 -0.003 -0.2948 -0.0275
2 3 0.2795 -0.0976 -0.2766 0.0741
5 4 0.0294 -0.0794 -0.0293 0.0576
4 3 -0.0759 -0.0801 0.0766 0.0259
6 7 -0.6 -0.1814 0.6 0.1966

Table 3: Power Flow Values for λ = 0, P=0.6

• λ = 1.5, P = 0.6 at bus 7

Busi Busj Pij Qij Pji Qji

1 2 2.7467 1.1467 -2.7467 0.0118
6 2 -0.5517 0.3418 0.5609 -0.3904
6 5 0.5517 0.1278 -0.5202 -0.084
2 5 0.9899 0.165 -0.9185 -0.0333
2 4 0.9167 0.2645 -0.8731 -0.1301
2 3 0.8785 -0.0509 -0.8474 0.1348
5 4 0.0887 -0.1827 -0.0883 0.168
4 3 -0.2386 -0.1879 0.2474 0.1652
6 7 -1.8 -0.7696 1.8 1.1799

Table 4: Power Flow Values for λ = 1.5, P=0.6

6
 ∆P12 = 2.7467 - 1.0664 = 1.6803pu for change from λ = 0 to λ = 1.5 at P=0.6 at
bus 7.
We can observe that the change in power flow in the largest power flow line decreases
on increasing generator voltage.

4.3 Line Over Load Alleviation using Series Compensation

• λ = 0 and X = 0.1

Busi Busj Pij Qij Pji Qji

1 2 1.2683 -0.184 -1.2683 0.3361
6 2 -0.3606 0.1381 0.3634 -0.2029
6 5 0.1606 -0.0259 -0.1585 -0.0237
2 5 0.3194 -0.0355 -0.3133 0.0053
2 4 0.3035 -0.0009 -0.2997 -0.0291
2 3 0.2822 -0.0968 -0.2793 0.0735
5 4 0.0217 -0.0816 -0.0217 0.0598
4 3 -0.0786 -0.0807 0.0793 0.0265
6 7 -0.4 -0.2122 0.4 0.2094

Table 5: Power flow for λ = 0 and series compensation X = 0.1

• λ = 1.5 and X = 0.1

Busi Busj Pij Qij Pji Qji

1 2 3.2631 1.4862 -3.2631 0.1823
6 2 -1.1034 0.5678 1.1373 -0.5902
6 5 0.5034 0.1202 -0.4762 -0.087
2 5 1.0152 0.175 -0.9386 -0.0305
2 4 0.9345 0.2767 -0.8883 -0.1309
2 3 0.8892 -0.0438 -0.8568 0.1329
5 4 0.0648 -0.1825 -0.0644 0.1681
4 3 -0.2474 -0.1872 0.2568 0.1671
6 7 -1.2 -0.988 1.2 1.2474

Table 6: Power flow for λ = 1.5 and series compensation X = 0.1

∆P12 = 3.2631 − 1.2683 = 1.9948 for change from λ = 0 to λ = 1.5 at X=0.1.

7
 • λ = 0, X = 0.05
Busi Busj Pij Qij Pji Qji
1 2 1.2690 -0.3149 -1.2690 0.3941
6 2 -0.3589 0.1969 0.3622 -0.2604
6 5 0.1589 -0.0209 -0.1568 -0.0279
2 5 0.3202 -0.0362 -0.314 0.007
2 4 0.3042 -0.0011 -0.3004 -0.028
2 3 0.2826 -0.0964 -0.2796 0.0737
5 4 0.0208 -0.0791 -0.0207 0.0576
4 3 -0.0789 -0.0796 0.0796 0.0263
6 7 -0.4 -0.276 0.4 0.2765

Table 7: Power flow for λ = 0 and series compensation X = 0.05

• λ = 1.5, X = 0.05

Busi Busj Pij Qij Pji Qji

1 2 3.2417 0.7951 -3.2417 -0.0438
6 2 -1.1061 0.2821 1.1324 -0.3171
6 5 0.5061 0.0882 -0.4814 -0.0663
2 5 1.0059 0.1605 -0.9373 -0.0384
2 4 0.9282 0.2596 -0.8868 -0.1364
2 3 0.8826 -0.0593 -0.8533 0.1343
5 4 0.0688 -0.1953 -0.0684 0.1791
4 3 -0.2449 -0.1927 0.2533 0.1657
6 7 -1.2 -0.6702 1.2 0.8512

Table 8: Power flow for λ = 1.5 and series compensation X = 0.05

∆P12 = 3.2417 − 1.2690 = 1.9727 for change from λ = 0 to λ = 1.5 at X=0.05.

We can observe that the change in power flow in the largest power flow line decreases
on decreasing series compensation.

4.4 Reactive Rescheduling to Improve Voltage Profile

Initial values:
• Generator bus voltage at bus 7 = 1.04
• Tap ratio at bus 7 = 1.04
• λ = 0 without any changes in the circuit

8
 Bus Vmag Angle (rad)
1 1.06 0
2 1.0615 -0.1152
3 1.0628 -0.1538
4 1.0468 -0.1650
5 1.0451 -0.1644
6 1.0624 -0.1356
7 1.04 -0.098

Table 9: Converged voltages for λ = 0 without any changes

• λ = 1 without any changes in the circuit

Bus Vmag Angle (rad)

1 1.06 0
2 1.0302 -0.2434
3 1.0230 -0.3246
4 0.9875 -0.3493
5 0.9844 -0.3478
6 1.0302 -0.2868
7 1.04 -0.2090

Table 10: Converged voltages for λ = 1 without any changes

• λ = 1 with increasing the generator bus voltage to 1.06 at bus 7

Bus Vmag Angle (rad)

1 1.06 0
2 1.0393 -0.2412
3 1.0326 -0.3209
4 0.9980 -0.3453
5 0.9951 -0.3438
6 1.0433 -0.2850
7 1.06 -0.2097

Table 11: Converged voltages for λ = 1 with increasing the generator bus voltage to 1.06
at bus 7

We can see that the voltages of PQ buses are closer to 1 after increasing generator
bus voltage.

9
 • λ = 1 with changing the transformer tap ratio to 1.06 at bus 7

Bus Vmag Angle (rad)

1 1.06 0
2 10382 -0.2415
3 1.0315 -0.3213
4 0.9968 -0.3457
5 0.9939 -0.3443
6 1.0418 -0.2852
7 1.04 -0.2069

Table 12: Converged voltages for λ = 1 with changing the transformer tap ratio to
1.06 at bus 7

We can see that the voltages of PQ buses are closer to 1 after increasing the tap
ratio at bus 7.
• λ = 1 with adding a shunt capacitor of 0.05 at bus 5

Bus Vmag Angle (rad)

1 1.06 0
2 1.0334 -0.2426
3 1.0272 -0.3233
4 0.9932 -0.3483
5 0.9907 -0.3470
6 1.0328 -0.2854
7 1.04 -0.2079

Table 13: Converged voltages for λ = 1 with adding a shunt capacitor of 0.05 at bus 5

We can see that the voltages are closer to 1 after adding a shunt capacitor at bus 5.
Hence, we can observe that all the three methods here are improving the voltage
profile of the network.

4.5 Ferranti Effect

Bus Vmag Angle (rad)

1 1.06 0
2 1.0615 -0.1152
3 1.0628 -0.1538
4 1.0468 -0.165
5 1.0451 -0.1644
6 1.0624 -0.1356
7. 1.04 -0.098

Table 14: Bus voltages for λ = 0

10
 Bus Vmag Angle (rad)
1 1.06 0
2 1.0668 -0.0798
3 1.0701 -0.1071
4 1.0594 -0.1149
5 1.0582 -0.1145
6 1.0689 -0.0943
7 1.04 -0.0681

Table 15: Bus voltages for λ = - 0.3

Bus Vmag Angle (rad)

1 1.06 0
2 1.0695 -0.0567
3 1.074 -0.0766
4 1.0669 -0.0823
5 1.0660 -0.082
6 1.0725 -0.0675
7 1.04 -0.0488

Table 16: Bus voltages for λ = −0.5

Bus Vmag Angle (rad)

1 1.06 0
2 1.0723 -0.0225
3 1.0785 -0.0316
4 1.0766 -0.0343
5 1.0761 -0.0342
6 1.0771 -0.0279
7 1.04 -0.0205

Table 17: Bus voltages for λ = −0.8

Bus Vmag Angle (rad)

1 1.06 0
2 1.0730 -0.0113
3 1.0796 -0.0167
4 1.0794 -0.0184
5 1.0792 -0.0184
6 1.0784 -0.0149
7 1.04 -0.0112

Table 18: Bus voltages for λ = −0.9

We observe that the voltage of PQ buses are greater than that of the PV bus for light
load case, indicating Ferranti effect.
11
5 Conclusion
1. There exists some critical value of loading above which, Newton Raphson will not
converge.
2. We can change the loading of lines by changing the power generated by PV bus
which can be used to avoid the overloading of lines
3. Using series compensation, the power distribution over lines can be changed. On
increasing the series compensation, the SIL (Surge Impedance Loading) increases.
This increases difference between λ = 0 and λ = 1.5 on increasing value of compen-
sation.
4. Voltage profile can be improved (brought near to 1) by these 3 methods:
(a) Increase voltage of PV bus
(b) Changing tap ratio
(c) Adding shunt capacitance to increase reactive power
5. For low load, the load voltages (voltages of PQ bus) are more than the source
voltages (Voltages of PV bus).

6 Precautions
• One must always take a note of the rated machine values and never exceed them
while performing the experiment.
• One must ensure that all the connections are tight and have been verified properly
before powering on the circuit.
• One must wear proper clothes, shoes and handle the electrical outlets with care.

12