EEE 3201 Power System I

Lectures for [Link]. Students

Semester: L/T - 3/2

Md. Minarul Islam, Ph.D.

Assistant Professor
Department of Electrical and Electronic Engineering
University of Dhaka
Dhaka-1000, Bangladesh
Power Flow Analysis
Power Flow Analysis

❑ Information of the power system

❖ One-line diagram of the power system

❖ Line Data

❖ Bus Data

❖ Modeling of transformer, transmission

lines, generators, and loads.

❑ Solver:
Contents

❑Network Matrices

❑Power Flow Problems

❑Solvers:

✓ Gauss-Seidel Method

✓ Newton-Raphson Method
Network Matrices from Line Data
Table 1: Line Data

Line Data Table

Ybus Formation
3 − 𝑗9 −2 + 𝑗6 −1 + 𝑗3 0
−2 + 𝑗6 3.666 − 𝑗11 −0.666 + 𝑗2 −1 + 𝑗3
𝑌𝑏𝑢𝑠 =
−1 + 𝑗3 −0.666 + 𝑗2 3.666 − 𝑗11 −2 + 𝑗3
0 −1 + 𝑗3 −2 + 𝑗3 3 − 𝑗9
 Network Matrices
Node equations and formation of Ybus matrix 2
Yb Ye

Yc Yd Yf
KCL at node 1:
3 4
𝑉1 − 𝑉3 𝑌𝑐 + 𝑉1 − 𝑉2 𝑌𝑑 + 𝑉1 − 𝑉4 𝑌𝑓 = 0
1
𝑉1 (𝑌𝑐 + 𝑌𝑑 + 𝑌𝑓 ) − 𝑉2 𝑌𝑑 − 𝑉3 𝑌𝑐 − 𝑉4 𝑌𝑓 = 0 (1) Yg
Ya
I3 I4
0
KCL at node 3:
Reference
𝑉3 𝑌𝑎 + 𝑉3 − 𝑉2 𝑌𝑏 + 𝑉3 − 𝑉1 𝑌𝑐 = 𝐼3
−𝑉1 𝑌𝑐 − 𝑉2 𝑌𝑏 + 𝑉3 (𝑌𝑎 + 𝑌𝑏 + 𝑌𝑐 ) = 𝐼3 (2)

Similarly;

Node 2 −𝑉1 𝑌𝑑 + 𝑉2 (𝑌𝑏 + 𝑌𝑑 + 𝑌𝑒 ) − 𝑉3 𝑌𝑏 − 𝑉4 𝑌𝑒 = 0 (3)

Node 4 −𝑉1 𝑌𝑓 − 𝑉2 𝑌𝑒 + 𝑉4 (𝑌𝑒 + 𝑌𝑓 + 𝑌𝑔 ) = 𝐼4 (4)

 Network Matrices
Node equations and formation of Ybus matrix 2
𝑉1 (𝑌𝑐 + 𝑌𝑑 + 𝑌𝑓 ) − 𝑉2 𝑌𝑑 − 𝑉3 𝑌𝑐 − 𝑉4 𝑌𝑓 = 0 (1) Yb Ye
−𝑉1 𝑌𝑑 + 𝑉2 (𝑌𝑏 + 𝑌𝑑 + 𝑌𝑒 ) − 𝑉3 𝑌𝑏 − 𝑉4 𝑌𝑒 = 0 (3) Yd
Yc Yf
−𝑉1 𝑌𝑐 − 𝑉2 𝑌𝑏 + 𝑉3 (𝑌𝑎 + 𝑌𝑏 + 𝑌𝑐 ) = 𝐼3 (2) 3 4
−𝑉1 𝑌𝑓 − 𝑉2 𝑌𝑒 + 𝑉4 (𝑌𝑒 + 𝑌𝑓 + 𝑌𝑔 ) = 𝐼4 (4) 1
Ya Yg
I3 I4
If we arrange equations (1) to (4) in matrix 0
form, we have
Reference
𝑌𝑐 + 𝑌𝑑 + 𝑌𝑓 −𝑌𝑑 −𝑌𝑐 −𝑌𝑓 𝑉1 0
−𝑌𝑑 𝑌𝑏 + 𝑌𝑑 + 𝑌𝑒 −𝑌𝑏 −𝑌𝑒 𝑉2 0
= (5)
−𝑌𝑐 −𝑌𝑏 𝑌𝑎 + 𝑌𝑏 + 𝑌𝑐 0 𝑉3 𝐼3
−𝑌𝑓 −𝑌𝑒 0 𝑌𝑒 + 𝑌𝑓 + 𝑌𝑔 𝑉4 𝐼4

In general, for any network, we 𝑌11 𝑌12 𝑌13 𝑌14 𝑉1 𝐼1

𝑌21 𝑌22 𝑌23 𝑌24 𝑉2 𝐼
can form the equation (5) as = 2 (6)
𝑌31 𝑌32 𝑌33 𝑌34 𝑉3 𝐼3
follows:
𝑌41 𝑌42 𝑌43 𝑌44 𝑉4 𝐼4
 Network Matrices
Node equations and formation of Ybus matrix 2
Yb Ye
Equation (6) can be represented as
Ybus V = I Yc Yd Yf
3 4
𝑌11 𝑌12 𝑌13 𝑌14
1
𝑌 𝑌22 𝑌23 𝑌24
𝑌𝑏𝑢𝑠 = 21 (7) Ya Yg
𝑌31 𝑌32 𝑌33 𝑌34
I3 I4
𝑌41 𝑌42 𝑌43 𝑌44 0
Reference
𝑌𝑏𝑢𝑠 is called bus admittance matrix.

The usual rules to form 𝑌𝑏𝑢𝑠 are:

1. Diagonal elements 𝑌𝑖𝑖 = 𝑠𝑢𝑚 𝑜𝑓 𝑡ℎ𝑒 𝑎𝑑𝑚𝑖𝑡𝑎𝑛𝑐𝑒𝑠 𝑐𝑜𝑛𝑛𝑒𝑐𝑡𝑒𝑑 𝑡𝑜 𝑛𝑜𝑑𝑒 𝑖.
2. The off-diagonal element 𝑌𝑖𝑗 = −𝑛𝑒𝑡 𝑎𝑑𝑚𝑖𝑡𝑎𝑛𝑐𝑒 𝑐𝑜𝑛𝑛𝑒𝑐𝑡𝑒𝑑 𝑏𝑒𝑡𝑤𝑒𝑒𝑛 𝑛𝑜𝑑𝑒 𝑖 𝑎𝑛𝑑 𝑗.
𝑌𝑖𝑖 is called self-admittance, 𝑌𝑖𝑗 is called mutual admittance.
Contents

❑ Network Matrices

❑Power Flow Problems

❑ Solvers:
✓ Gauss-Seidel Method

✓ Newton-Raphson Method
 Power Flow Problems
Two scenario
Addition of new generations Addition of new loads

How do you
know grid’s
capability for
both scenario?

Load Flow Problem?

 Load Flow Analysis

❑ Important in planning and designing the future

expansion of the power system.

❑ Determining the best operation of the existing system.

❖Principal Information from Load Flow Analysis:

❖Magnitude and phase angle of voltage at each bus.

❖P and Q flowing in each line.

Load Flow Analysis
❖ What we have at the starting of LFA?

✓ The single-line diagram of the system and possibly the line and bus data.

❖ Form 𝑌𝑏𝑢𝑠 from the single-line diagram, where the elements of the 𝑌𝑏𝑢𝑠 can be

𝑌𝑖𝑗 = 𝑌𝑖𝑗 ∠𝜃𝑖𝑗 = 𝑌𝑖𝑗 cos 𝜃𝑖𝑗 + 𝑗 𝑌𝑖𝑗 sin 𝜃𝑖𝑗 = 𝐺𝑖𝑗 + 𝑗𝐵𝑖𝑗 (1)

❖ Other required information are:

✓ Transformer ratings and Z,

✓ Shunt capacitor ratings
✓ Transformer tap settings, etc.

❖ In advance of each LFA, certain bus voltages and power injections must
be given as known values.
 Load Flow Analysis
−𝑉1 𝑌𝑐 − 𝑉2 𝑌𝑏 + 𝑉3 (𝑌𝑎 + 𝑌𝑏 + 𝑌𝑐 ) = 𝐼3 (2)
−𝑉1 𝑌𝑓 − 𝑉2 𝑌𝑒 + 𝑉4 (𝑌𝑒 + 𝑌𝑓 + 𝑌𝑔 ) = 𝐼4 (4)
Let voltage at a bus i
2
Yb Ye
𝑉𝑖 = 𝑉𝑖 ∠𝛿𝑖 = 𝑉𝑖 (cos 𝛿𝑖 + 𝑗 sin 𝛿𝑖 ) (2)

Yc Yd Yf
So the net current injected at bus i is
3 4
𝐼𝑖 = 𝑌𝑖1 𝑉1 + 𝑌𝑖2 𝑉2 + ⋯ + 𝑌𝑖𝑁 𝑉𝑁 = σ𝑁
𝑛=1 𝑌𝑖𝑛 𝑉𝑛 (3) 1
Ya Yg
I3 I4
Now let 𝑃𝑖 and 𝑄𝑖 entering the network at bus i. Then 0

𝑃𝑖 − 𝑗𝑄𝑖 = 𝑉𝑖∗ 𝐼𝑖 = 𝑉𝑖∗ σ𝑁

𝑛=1 𝑌𝑖𝑛 𝑉𝑛 (4) Reference
𝑃𝑖 − 𝑗𝑄𝑖 = 𝑉𝑖∗ 𝐼𝑖 = ( 𝑉𝑖 ∠𝛿𝑖 )∗ σ𝑁
𝑛=1( 𝑌𝑖𝑛 ∠𝜃𝑖𝑛 )( 𝑉𝑛 ∠𝛿𝑛 )

𝑃𝑖 − 𝑗𝑄𝑖 = 𝑉𝑖∗ 𝐼𝑖 = σ𝑁
𝑛=1( 𝑌𝑖𝑛 𝑉𝑖 𝑉𝑛 ∠(𝜃𝑖𝑛 + 𝛿𝑛 − 𝛿𝑖 ) (5)
So, we have
𝑃𝑖 = σ𝑁
𝑛=1( 𝑌𝑖𝑛 𝑉𝑖 𝑉𝑛 𝑐𝑜𝑠(𝜃𝑖𝑛 + 𝛿𝑛 − 𝛿𝑖 ) (6)
𝑄𝑖 = − σ𝑁
𝑛=1( 𝑌𝑖𝑛 𝑉𝑖 𝑉𝑛 𝑠𝑖𝑛(𝜃𝑖𝑛 + 𝛿𝑛 − 𝛿𝑖 ) (7)
 Load Flow Analysis
Now let

❖ 𝑃𝑔𝑖 = scheduled power being

generated at bus i.

i
❖ 𝑃𝑑𝑖 = scheduled power Pgi Pi,sch
demand of load at bus i. Pi

Pdi
❖ Then 𝑃𝑖,𝑠𝑐ℎ = 𝑃𝑔𝑖 − 𝑃𝑑𝑖 =net
scheduled power injected
i
into the network at bus i. Qgi Qi,sch
Qi
❖ Now let, 𝑃𝑖,𝑐𝑎𝑙 =calculated Qdi
value of power at bus i.
Fig. Active and reactive power flow
in a bus i in a power network.
Load Flow Analysis
So, the mismatch i
Pgi Pi,sch
∆𝑃𝑖 = 𝑃𝑖,𝑠𝑐ℎ − 𝑃𝑖,𝑐𝑎𝑙 Pi
∆𝑃𝑖 = (𝑃𝑔𝑖 − 𝑃𝑑𝑖 ) − 𝑃𝑖,𝑐𝑎𝑙 (8)
Pdi
Similarly,
∆𝑄𝑖 = 𝑄𝑖,𝑠𝑐ℎ − 𝑄𝑖,𝑐𝑎𝑙 i
Qgi Qi,sch
∆𝑄𝑖 = (𝑄𝑔𝑖 − 𝑄𝑑𝑖 ) − 𝑄𝑖,𝑐𝑎𝑙 (9)
Qi

Qdi

Fig. Active and reactive power flow

If ∆𝑃𝑖 = 0 = ∆𝑄𝑖 at bus i, we may
in a bus i in a power network.
write the power balance equations

𝑔𝑖′ = 𝑃𝑖,𝑠𝑐ℎ − 𝑃𝑖,𝑐𝑎𝑙 = (𝑃𝑔𝑖 − 𝑃𝑑𝑖 ) − 𝑃𝑖,𝑐𝑎𝑙 = 0 (10)

𝑔𝑖′′ = 𝑄𝑖,𝑠𝑐ℎ − 𝑄𝑖,𝑐𝑎𝑙 = (𝑄𝑔𝑖 − 𝑄𝑑𝑖 ) − 𝑄𝑖,𝑐𝑎𝑙 = 0 (11)
 Load Flow Analysis
𝑔𝑖′ = 𝑃𝑖,𝑠𝑐ℎ − 𝑃𝑖,𝑐𝑎𝑙 = (𝑃𝑔𝑖 − 𝑃𝑑𝑖 ) − 𝑃𝑖,𝑐𝑎𝑙 = 0 (10)
𝑔𝑖′′ = 𝑄𝑖,𝑠𝑐ℎ − 𝑄𝑖,𝑐𝑎𝑙 = (𝑄𝑔𝑖 − 𝑄𝑑𝑖 ) − 𝑄𝑖,𝑐𝑎𝑙 = 0 (11)

Note:
❖ If bus i has no generation or load, the corresponding terms are set equal to zero
in equation (10) and (11).
❖ Each bus of the network has two such equations (10 and 11), and the load flow
problem is to solve equation (6) and (7) for values of the unknown bus voltages
which cause equation (10) and (11) to be numerically satisfied at each bus.
❖ If there is no scheduled value 𝑃𝑖,𝑠𝑐ℎ for bus i, then the mismatch ∆𝑃𝑖 = 𝑃𝑖,𝑠𝑐ℎ
− 𝑃𝑖,𝑐𝑎𝑙 cannot be defined and not required to satisfy equation (10), in the
course of solving the power flow problem.
❖ Similarly, for ∆𝑄𝑖 (11) not to be satisfied.

𝑃𝑖 = σ𝑁
𝑛=1( 𝑌𝑖𝑛 𝑉𝑖 𝑉𝑛 𝑐𝑜𝑠(𝜃𝑖𝑛 + 𝛿𝑛 − 𝛿𝑖 ) (6)
𝑄𝑖 = − σ𝑁
𝑛=1( 𝑌𝑖𝑛 𝑉𝑖 𝑉𝑛 𝑠𝑖𝑛(𝜃𝑖𝑛 + 𝛿𝑛 − 𝛿𝑖 ) (7)
 Load Flow Analysis
𝑔𝑖′ = 𝑃𝑖,𝑠𝑐ℎ − 𝑃𝑖,𝑐𝑎𝑙 = (𝑃𝑔𝑖 − 𝑃𝑑𝑖 ) − 𝑃𝑖,𝑐𝑎𝑙 = 0 (10)
𝑔𝑖′′ = 𝑄𝑖,𝑠𝑐ℎ − 𝑄𝑖,𝑐𝑎𝑙 = (𝑄𝑔𝑖 − 𝑄𝑑𝑖 ) − 𝑄𝑖,𝑐𝑎𝑙 = 0 (11)

Note:
❖ For each bus i, four potentially unknown quantities are 𝑃𝑖 , 𝑄𝑖 , voltage angle 𝛿𝑖
and 𝑉𝑖 .
❖ At most, there are two equations like (10) and (11) available for each node, and
so we must consider how the number of unknown quantities can be reduced to
agree with the number of available equations before beginning to solve the
power flow problem.
❖ The general practice in power flow studies is to identify three types of buses in
the network.
❖ At each bus i, two of the four quantities 𝑃𝑖 , 𝑄𝑖 , voltage angle 𝛿𝑖 and 𝑉𝑖 are
specified, and the remaining two are calculated.

𝑃𝑖 = σ𝑁
𝑛=1( 𝑌𝑖𝑛 𝑉𝑖 𝑉𝑛 𝑐𝑜𝑠(𝜃𝑖𝑛 + 𝛿𝑛 − 𝛿𝑖 ) (6)
𝑄𝑖 = − σ𝑁
𝑛=1( 𝑌𝑖𝑛 𝑉𝑖 𝑉𝑛 𝑠𝑖𝑛(𝜃𝑖𝑛 + 𝛿𝑛 − 𝛿𝑖 ) (7)
 Type of Buses in the Network
Specified quantities are chosen according
to the following discussion.

Load bus:
❑ Often called P-Q bus.
Fig. 3 One-line diagram of a four-
bus system
❑ Often 𝑃𝑑𝑖 →known. 𝑄𝑑𝑖 is calculated from
assumed 0.85 pf and MVA capacity of the
load.

❑ At bus i, 𝑃𝑖,𝑠𝑐ℎ = −𝑃𝑑𝑖 , and 𝑄𝑖,𝑠𝑐ℎ = −𝑄𝑑𝑖

are known, ∆𝑃𝑖 and ∆𝑄𝑖 are defined.

❑ Unknown quantities are 𝛿𝑖 and 𝑉𝑖

 Type of Buses in the Network
Slack bus:

❑ The voltage angle of this bus serves as

reference for the angles of all other
bus voltages
Fig. 3 One-line diagram of a four-
bus system
❑ Bus 1 at any system is considered as
slack bus.

❑ Usual practice is to set 𝛿𝑖 = 0

❑ Mismatches are not defined for slack bus. That mean, P1 and Q1 are not
scheduled at this bus. Why?
 Type of Buses in the Network
Voltage-controlled bus:
❑ PV bus, generator bus,

❑ Any bus at which the voltage

magnitude is kept constant.
Fig. 3 One-line diagram of a four-
bus system
❑ At each bus having generator
connected, MW generation can be
controlled by prime mover, and
voltage magnitude can be controlled
by excitation.

❑ Therefore, at each generator bus i, we may properly specify 𝑃𝑔𝑖 and 𝑉𝑖 .

❑ We have to calculate 𝛿𝑖 and 𝑄𝑖

 Type of Buses in the Network
Why are P1 and Q1 not scheduled at the slack bus?
𝑔𝑖′ = 𝑃𝑖,𝑠𝑐ℎ − 𝑃𝑖,𝑐𝑎𝑙 = (𝑃𝑔𝑖 − 𝑃𝑑𝑖 ) − 𝑃𝑖,𝑐𝑎𝑙 = 0 (10)
𝑔𝑖′′ = 𝑄𝑖,𝑠𝑐ℎ − 𝑄𝑖,𝑐𝑎𝑙 = (𝑄𝑔𝑖 − 𝑄𝑑𝑖 ) − 𝑄𝑖,𝑐𝑎𝑙 = 0 (11)

Consider a system having N buses. Equation (10) can be written for each bus.
Adding these equations for N buses, we shall get
𝑁 𝑁 𝑁

𝑃ณ𝐿 = 𝐼 2 𝑅 = ෍ 𝑃𝑖 = ෍ 𝑃𝑔𝑖 − ෍ 𝑃𝑑𝑖 (12)

𝑅𝑒𝑎𝑙 𝑃𝑜𝑤𝑒𝑟 𝑙𝑜𝑠𝑠 𝑖=1 𝑖=1 𝑖=1
𝑇𝑜𝑡𝑎𝑙 𝑔𝑒𝑛𝑒𝑟𝑎𝑡𝑖𝑜𝑛 𝑇𝑜𝑡𝑎𝑙 𝑑𝑒𝑚𝑎𝑛𝑑

❖ PL: Loss in the transmission lines and transformers in the network.

 Type of Buses in the Network
Why are P1 and Q1 not scheduled at the slack bus?
𝑁 𝑁 𝑁

𝑃ณ𝐿 = 𝐼 2 𝑅 = ෍ 𝑃𝑖 = ෍ 𝑃𝑔𝑖 − ෍ 𝑃𝑑𝑖 (12)

𝑅𝑒𝑎𝑙 𝑃𝑜𝑤𝑒𝑟 𝑙𝑜𝑠𝑠 𝑖=1 𝑖=1 𝑖=1
𝑇𝑜𝑡𝑎𝑙 𝑔𝑒𝑛𝑒𝑟𝑎𝑡𝑖𝑜𝑛 𝑇𝑜𝑡𝑎𝑙 𝑑𝑒𝑚𝑎𝑛𝑑

❖ Until the voltages of buses are not known, lines’ currents can be calculated.
So, PL is initially unknown, and it is not possible to prespecify all the
quantities in equation (12).

❖ Hence, in power flow problem formulation, we choose one bus, the slack bus,
at which Pg is not scheduled.

❖ When the power flow problem is solved, the difference (slack) between total
specified P going into the system at all other buses and the total output P and
I2R losses are assigned to the slack bus.
 Type of Buses in the Network
Why are P1 and Q1 not scheduled at the slack bus?
𝑔𝑖′ = 𝑃𝑖,𝑠𝑐ℎ − 𝑃𝑖,𝑐𝑎𝑙 = (𝑃𝑔𝑖 − 𝑃𝑑𝑖 ) − 𝑃𝑖,𝑐𝑎𝑙 = 0 (10)
𝑔𝑖′′ = 𝑄𝑖,𝑠𝑐ℎ − 𝑄𝑖,𝑐𝑎𝑙 = (𝑄𝑔𝑖 − 𝑄𝑑𝑖 ) − 𝑄𝑖,𝑐𝑎𝑙 = 0 (11)

❑ For reactive power, the difference between the total MVars supplied by the generators
at the buses and the Mvars received by the load can be expressed;
𝑁 𝑁 𝑁

෍ 𝑄𝑖 = ෍ 𝑄𝑔𝑖 − ෍ 𝑄𝑑𝑖 (13)

𝑖=1 𝑖=1 𝑖=1
𝑇𝑜𝑡𝑎𝑙 𝑔𝑒𝑛𝑒𝑟𝑎𝑡𝑖𝑜𝑛 𝑇𝑜𝑡𝑎𝑙 𝑑𝑒𝑚𝑎𝑛𝑑

❖ Individual Qi can be obtained from equation (7) and can be satisfied by equation (11).
❖ Total Qi in (13) accounts for Mvars associated with line charging, shunt capacitors and
reactors installed at the buses, and I2X loss in series reactance of the lines.

𝑃𝑖 = σ𝑁
𝑛=1( 𝑌𝑖𝑛 𝑉𝑖 𝑉𝑛 𝑐𝑜𝑠(𝜃𝑖𝑛 + 𝛿𝑛 − 𝛿𝑖 ) (6)
𝑄𝑖 = − σ𝑁
𝑛=1( 𝑌𝑖𝑛 𝑉𝑖 𝑉𝑛 𝑠𝑖𝑛(𝜃𝑖𝑛 + 𝛿𝑛 − 𝛿𝑖 ) (7)
 Type of Buses in the Network
❖ The unscheduled bus-voltage magnitudes and angles in power flow
study are called state variables or dependent variables which depend on
the specified quantities.

❖ Hence, it is necessary to solve the power flow problem to determine all

the state variables by solving an equal number of power flow equations.

❖ Therefore, in a power flow problem with N bus system, it is necessary

to find number of specified quantities, number of state variables and
number of state equations to be solved.

❖ Once the state variables are calculated, the complete state of the system
is known and all other quantities which depend on the state variables
can be determined.

❖ Quantities such as P1 and Q1 at the slack bus, Qi at each voltage-

controlled bus, and the power loss PL of the system are examples of
dependent functions.
 Type of Buses in the Network

Summary of power flow problem:

Contents

❑ Network Matrices

❑ Power Flow Problems

❑Solvers: Power Flow Solutions

✓ Gauss-Seidel Method

✓ Newton-Raphson Method
 Solvers: Power Flow Solutions
𝑃𝑖 = σ𝑁
𝑛=1( 𝑌𝑖𝑛 𝑉𝑖 𝑉𝑛 𝑐𝑜𝑠(𝜃𝑖𝑛 + 𝛿𝑛 − 𝛿𝑖 ) (6)
𝑄𝑖 = − σ𝑁
𝑛=1( 𝑌𝑖𝑛 𝑉𝑖 𝑉𝑛 𝑠𝑖𝑛(𝜃𝑖𝑛 + 𝛿𝑛 − 𝛿𝑖 ) (7)

❑ Pi and Qi in (6) and (7) are nonlinear functions of the state variables
𝑉𝑖 𝑎𝑛𝑑 𝛿𝑖 .

❑ Hence, load-flow solutions usually use iterative techniques such as Gauss-

Seidel and Newton-Raphson methods.

❑ The Gauss-Seidel method solves the load-flow equations in rectangular

coordinates until differences in bus voltages from one iteration to another
are sufficiently small.

❑ The Newton-Raphson method solves the polar form of the load-flow

equations until the ∆𝑃 𝑎𝑛𝑑 ∆𝑄 mismatches at all buses fall within
specified tolerances.

❑ Both methods are based on bus admittance equations.

Contents

❑ Network Matrices

❑ Power Flow Problems

❑Solvers: Power Flow Solutions

✓Gauss-Seidel Method
✓ Newton-Raphson Method
 Solvers: Power Flow Solutions
The Gauss-Seidel Method
❑ Consider a four-bus system as shown in Fig.
2. From equation (4), we know
𝑁

𝑃𝑖 − 𝑗𝑄𝑖 = 𝑉𝑖∗ 𝐼𝑖 = 𝑉𝑖∗ ෍ 𝑌𝑖𝑛 𝑉𝑛 (4)

𝑛=1
Fig. 2 One-line diagram of a four-
Where 𝑃𝑖 and 𝑄𝑖 entering the network at bus i.
bus system

❑ As usual consider bus 1 as a slack bus.

Therefore, computation starts with bus 2.

❑ If 𝑃2,𝑠𝑐ℎ and 𝑄2,𝑠𝑐ℎ are the scheduled real and reactive power entering the network at
bus 2, then set 𝑖 = 2 and 𝑁 = 4 in equation (4). So, we obtain

𝑃2,𝑠𝑐ℎ − 𝑗𝑄2,𝑠𝑐ℎ
= 𝑌21 𝑉1 + 𝑌22 𝑉2 + 𝑌23 𝑉3 + 𝑌24 𝑉4 (14)
𝑉2∗
 Solvers: Power Flow Solutions
The Gauss-Seidel Method
𝑃2,𝑠𝑐ℎ − 𝑗𝑄2,𝑠𝑐ℎ
= 𝑌21 𝑉1 + 𝑌22 𝑉2 + 𝑌23 𝑉3 + 𝑌24 𝑉4 (14)
𝑉2∗

Solving for 𝑉2 gives

1 𝑃2,𝑠𝑐ℎ −𝑗𝑄2,𝑠𝑐ℎ
𝑉2 = 𝑌 − (𝑌21 𝑉1 + 𝑌23 𝑉3 + 𝑌24 𝑉4 ) (15)
22 𝑉2∗ Fig. 2 One-line diagram of a four-
bus system
Similarly, at bus 3, we obtain
1 𝑃3,𝑠𝑐ℎ −𝑗𝑄3,𝑠𝑐ℎ
𝑉3 = 𝑌 − (𝑌31 𝑉1 + 𝑌32 𝑉2 + 𝑌34 𝑉4 ) (16)
33 𝑉3∗

And at bus 4, we obtain

1 𝑃4,𝑠𝑐ℎ −𝑗𝑄4,𝑠𝑐ℎ
𝑉4 = 𝑌 𝑉4∗
− (𝑌41 𝑉1 + 𝑌42 𝑉2 + 𝑌43 𝑉3 ) (17)
44

❖ As buses 2, 3, and 4 are load buses, they have scheduled real and reactive power.
❖ The specified voltage for the slack bus is 𝑉1 = |𝑉1 |∠𝛿1 .
 Solvers: Power Flow Solutions
The Gauss-Seidel Method

❑ Consider the initial voltage estimates of buses 2, 3,

(0) (0) (0)
and 4 are 𝑉2 , 𝑉3 and 𝑉4 , respectively.
❑ From equation (15) we obtain

(1) 1 𝑃2,𝑠𝑐ℎ −𝑗𝑄2,𝑠𝑐ℎ (0) (0)

𝑉2 =𝑌 (0)∗ − (𝑌21 𝑉1 + 𝑌23 𝑉3 + 𝑌24 𝑉4 ) (18)
22 𝑉2

(1) (0)
❖ See equation (18), the calculated value 𝑉2 and the estimated value 𝑉2 will not be
equal.
❖ After several iterations we may get the correct value of 𝑉2 with a good degree of
accuracy.
❖ 𝑉2 also depends on estimated voltages and scheduled powers at buses 3 and 4.
Therefore, the actual values of voltages are not yet known.
 Solvers: Power Flow Solutions
The Gauss-Seidel Method

❑ As the corrected voltage is found at each bus, it is

used to calculate the corrected voltage at the next
bus.
(1)
❑ Therefore, substituting 𝑉2 into equation (16) we
obtain

(1) 1 𝑃3,𝑠𝑐ℎ −𝑗𝑄3,𝑠𝑐ℎ (1) (0)

𝑉3 =𝑌 (0)∗ − (𝑌31 𝑉1 + 𝑌32 𝑉2 + 𝑌34 𝑉2 ) (19)
33 𝑉3

Similarly, from (17), we obtain

(1) 1 𝑃4,𝑠𝑐ℎ −𝑗𝑄4,𝑠𝑐ℎ (1) (1)

𝑉4 =𝑌 (0)∗ − (𝑌41 𝑉1 + 𝑌42 𝑉2 + 𝑌43 𝑉3 ) (20)
44 𝑉4
 Solvers: Power Flow Solutions
The Gauss-Seidel Method
❖ So, in the first iteration we obtain the state variables of the network.

❖ Then, the entire process is carried out again and again until the amount of correction
in voltage at every bus is less than a tolerance value.

❖ This process of solving the power-flow equations is known as the Gauss-Seidel

iterative method.

❖ For the initial estimates, it is usual practice to set the unknown voltages at all load
buses equal to = 1.0∠00 pu. Such initialization is called flat-start.

❑ For a system of N buses the general equation for the calculated voltage at any bus i
where P and Q are scheduled is
(𝑘) 1 𝑃𝑖,𝑠𝑐ℎ −𝑗𝑄𝑖,𝑠𝑐ℎ 𝑘 𝑘−1
𝑉𝑖 =𝑌 (𝑘−1)∗ − σ𝑖−1
𝑗=1 𝑌𝑖𝑗 𝑉𝑗 − σ𝑁
𝑗=𝑖+1 𝑌𝑖𝑗 𝑉𝑗 (21)
𝑖𝑖 𝑉𝑖

❖ we see that the values for the voltages on the right-hand side of this equation are the
most recently calculated values for the corresponding buses.
 Solvers: Power Flow Solutions
The Gauss-Seidel Method
Example

❑ Fig. 3 shows the one-line diagram of a

simple power system .
❑ Generators are connected at bus 1 and 4
while loads are indicated at all four buses .
Fig. 3 One-line diagram of a four-
❑ The line data of Table 9.2 give per-unit series bus system
impedances and line-charging susceptances for
the nominal-π equivalents of the four lines
identified by the buses at which they terminate .

❑ The bus data in Table 9.3 list values for P, Q, and V at each bus. The Q values of load
are calculated from the corresponding P values assuming a power factor of .0.85

❑ The net scheduled values Pi,sch and Qi,sch, are negative at the load buses 2 and .3
❑ Generated Qgi is not specified where voltage magnitude is constant .
❑ In the voltage column the values for the load buses are flat-start estimates .
 Solvers: Power Flow Solutions
The Gauss-Seidel Method
 Solvers: Power Flow Solutions
The Gauss-Seidel Method
Example

Solution:

Form Ybus from line data.

 Solvers: Power Flow Solutions
The Gauss-Seidel Method
Solution:

1  P2, sch − jQ2,sch ( 0) 

V 2
(1)
=  ( 0 )*
− (Y21V1 + Y23V3 + Y24V4 )
(0)

Y22  V2 
 − 1.7 + j1.0535 
1  − (1.00 + j 0.0)(−3.815629+ j19.078144) − (1.0 + j 0.0)(0.0 + j 0.0)
V2(1) = 1 + j 0.0
Y22  
 − (1.02 + j 0.0)(−5.169561+ j 25.847809) 

V 2
(1)
=
1
− 1.7 + j1.0535 + 9.088581− j 45.442909
Y22

7.388581− j 44.389409
V (1)
= = 0.983564− j 0.032316
8.985190− j 44.835953
2
 Solvers: Power Flow Solutions
The Gauss-Seidel Method
❑ To reduce the number of iteration, acceleration factor is used to converse desired
value.
❑ The difference between the newly calculated voltage and the best previous voltage
at the bus is multiplied by the appropriate acceleration factor to obtain a better
correction to be added to the previous value.
(1)
❑ For example, at bus 2, in the first iteration we have the accelerated value 𝑉2,𝑎𝑐𝑐
defined by the straight-line formula.

V2(,1acc
)
= (1 −  )V2( 0 ) + V2(1) = V2( 0 ) +  (V2(1) − V2( 0 ) ) (22)

for bus i

Vi (,acc
k)
= (1 −  )Vi (,acc
k −1)
+ Vi ( k ) = Vi (,acc
k −1)
+  (Vi ( k ) − Vi (,acc
k −1)
) (23)
 Solvers: Power Flow Solutions
The Gauss-Seidel Method
Vi (,acc
k)
= (1 −  )Vi (,acc
k −1)
+ Vi ( k ) = Vi (,acc
k −1)
+  (Vi ( k ) − Vi (,acc
k −1)
) (23)

❖ If α=1 ,then the Gauss-Siedel computed value of Vi is stored as the current

value.
❖ If 0<α <1, then the value to be stored is a weighted average of the Gauss-Seidel
value and the value stored from the previous iteration.
❖ If 1<α <2, then the value to be stored is essentially an extrapolated value.
❖ In power-flow studies α is generally set at about 1.6 and cannot exceed 2 if
convergence is to occur.
 Solvers: Power Flow Solutions
The Gauss-Seidel Method

❖ Since bus 4 is voltage controlled, we must treat it

differently as explained below.

❖ The acceleration factor for the real component of the correction may differ from
that for the imaginary component.

❖ For any system optimum values for acceleration factors exist, and poor choice of
factors may result in less rapid convergence or make convergence impossible .

❖ An acceleration factor of 1.6 for both the real and imaginary components is usually
a good choice but studies may be made to determine the best choice for a
particular system.
 Solvers: Power Flow Solutions
The Gauss-Seidel Method
Voltage-Controlled Bus
When voltage magnitude rather than reactive power is specified at bus i dnif ot deen ew ,
noitauqe gniwollof esU .tsrif rewop evitcaer
 N 
Qi = − Im Vi *  YijV j  )22(
 j =1 
which has the equivalent algorithmic expression


 ( k −1)  i −1 N 
( k −1) 
Q i
(k )
= − Im Vi *  YijV j +  YijV j  
(k )
)23(

  j =1 j =i 


Qi(k ) is evaluated by equation (23) for best previous voltage value at the buses, and this
(k )
value of Qi is used in the following equation to find a new value of V (k )
i

(𝑘) 1 𝑃𝑖,𝑠𝑐ℎ −𝑗𝑄𝑖,𝑠𝑐ℎ 𝑘 𝑘−1

𝑉𝑖 =𝑌 (𝑘−1)∗ − σ𝑖−1
𝑗=1 𝑌𝑖𝑗 𝑉𝑗 − σ𝑁
𝑗=𝑖+1 𝑌𝑖𝑗 𝑉𝑗 (21)
𝑖𝑖 𝑉𝑖
 Solvers: Power Flow Solutions
The Gauss-Seidel Method
Voltage-Controlled Bus

The new value of Vi (k ) is multiplied by the ratio of the specified constant magnitude
|Vi| to the magnitude of Vi (k ) found by above equation .
This result is the corrected complex voltage of the specified magnitude.

In the four-bus example if bus 4 is voltage controlled, Eq. (23) yields the calculated value


Q4(1) = − Im V4( 0 )* (Y41V1 + Y42V2(,1acc
)
+ Y43V3(,1acc
)
+ Y44V4( 0 )  )24(

(1) (1)
Then find V 4 using following equation where Q 4 is used instead of Q4,sch


1  4 , sch
P − jQ (1)

(1)  )25(
= − (Y41V1 + Y42V2 ,acc + Y43V3,acc 
4
V4(1) 
(1)

Y44 
 V4( 0 )*


 Solvers: Power Flow Solutions
The Gauss-Seidel Method
Voltage-Controlled Bus
Since |V4 |is specified, we correct the magnitude of V 4(1) as follows
V4(1)
V (1)
4 ,corr =| V4 | (1) )26(
| V4 |
(1)
and proceed to the next step with stored value V4 ,corr of bus 4 voltage having the specified
magnitude in the remaining calculations of the iteration.

Either voltage magnitude or reactive power must be specified at every bus except the
slack bus, where voltage is specified by magnitude and angle.
 Solvers: Power Flow Solutions
The Gauss-Seidel Method
Voltage-Controlled Bus

❖ A PV bus, |V| and Pg specified.

❖ Qg entering the network from the generation is then determined by solving LFA.
❖ From practical viewpoint Qg should be within defined limits given by the inequality
Qmin<=Qg<=Qmax.
❖ If calculated value of Qg is out of limit, Qg is set to the limit violated, the originally
specified voltage magnitude at the bus is relaxed, and the bus is then treated as a P-
Q bus for which a new voltage is calculated by the computer program .
❖ In subsequent iterations the program endeavors to sustain the originally specified
voltage at the bus while ensuring that Qg is within the permitted range of values.
This could well be possible since other changes may occur
Wish You
a clean
route
EEE 3201