Power System Operations and

Control (EPE 609)

Lecture Set (1)

Review of Power System Quantities,
and Load Flow Analysis
Prof. Muwaffaq I. Alomoush
Department of Electrical Power Engineering
Faculty of Engineering Technology
Yarmouk University

e-mail: [email protected]

First Semester 2023/2024

1
Terminology
𝑖, 𝑗 : Buses
𝑽𝑖 : Complex voltage at bus 𝑖 ; 𝑽𝑖 = 𝑽𝒊 ∠𝜹𝒊 = 𝑉𝑖 ∠𝛿𝑖 ; 𝑽𝑖 ≡ 𝑉𝑖
𝑉𝑖 : Voltage magnitude at bus 𝑖 ; 𝑉𝑖 ≡ 𝑽𝑖
𝛿𝑖 : Phase angle of the complex voltage at bus 𝑖
𝑺𝐺𝑖 : Complex power generated at bus 𝑖 , 𝑺𝐺𝑖 = 𝑃𝐺𝑖 + 𝑗𝑄𝐺𝑖
𝑃𝐺𝑖 : Real power generated at bus 𝑖
𝑄𝐺𝑖 : Reactive power generated at bus 𝑖
𝑰𝐺𝑖 : Complex current generated at bus 𝑖 (flowing from generator to bus)
𝑺𝐷𝑖 : Complex power of demand at bus 𝑖 , 𝑺𝐷𝑖 = 𝑃𝐷𝑖 + 𝑗𝑄𝐷𝑖
𝑃𝐷𝑖 : Real power of demand at bus 𝑖
𝑄𝐷𝑖 : Reactive power of demand at bus 𝑖

Prof. Muwaffaq I. Alomoush, Yarmouk University, Power Engineering Department, First Semester 2023/2024 2
Terminology: Cont.
𝑺𝑖 : Net injected complex power at bus 𝑖 ; 𝑺𝑖 = 𝑃𝑖 + 𝑗𝑄𝑖 = 𝑺𝐺𝑖 − 𝑺𝐷𝑖
𝑃𝑖 : Net injected real power at bus 𝑖 ; 𝑃𝑖 = 𝑃𝐺𝑖 − 𝑃𝐷𝑖
𝑄𝑖 : Net injected reactive power at bus 𝑖 ; 𝑄𝑖 = 𝑄𝐺𝑖 − 𝑄𝐷𝑖
𝑰𝐷𝑖 : Complex current of demand at bus 𝑖 (flowing from bus to demand)
𝑰𝑖 : Net injected complex current at bus 𝑖 ; 𝑰𝑖 = 𝑰𝐺𝑖 − 𝑰𝐷𝑖
𝑰𝑖𝑗 : Complex current flowing from bus 𝑖 to bus 𝑗 in the line 𝑖𝑗
𝑰𝑗𝑖 : Complex current from bus 𝑗 to bus 𝑖 in the line 𝑖𝑗 flowing, 𝑰𝑗𝑖 = −𝑰𝑖𝑗
𝒀𝑖𝑗 : The 𝑖𝑗𝑡ℎ element of the admittance matrix 𝒀, 𝒀𝑖𝑗 = 𝒀𝑖𝑗 ∠𝜃𝑖
= 𝑌𝑖𝑗 ∠𝜃𝑖 ; 𝒀𝑖𝑗 ≡ 𝑌𝑖𝑗
𝒀 : The admittance matrix ; 𝒀 = 𝑮 + 𝑗 𝑩
𝑮 : The conductance matrix
𝑩 : The susceptance matrix
𝒚𝑖𝑗 : The admittance of the line (branch) 𝑖𝑗 ; 𝒚𝑖𝑗 = 𝑔𝑖𝑗 + 𝑗𝑏𝑖𝑗
𝒛𝑖𝑗 : The impedance of the line (branch) 𝑖𝑗 ; 𝒛𝑖𝑗 = 𝑟𝑖𝑗 + 𝑗𝑥𝑖𝑗

Prof. Muwaffaq I. Alomoush, Yarmouk University, Power Engineering Department, First Semester 2023/2024 3
Recall:
𝑎 𝑏 1 𝑑 −𝑏
 If 𝑋 = , then 𝑋 −1 =
𝑐 𝑑 𝑎 𝑑 −(𝑏)(𝑐) −𝑐 𝑎

 If 𝒀𝑖𝑗 = 𝑌𝑖𝑗 ∠𝜃𝑖𝑗 ,

then 𝒀𝑖𝑗 = 𝑌𝑖𝑗 cos( 𝜃𝑖𝑗 ) + 𝑗 𝑌𝑖𝑗 sin( 𝜃𝑖𝑗 ) = 𝐺𝑖𝑗 +j 𝐵𝑖𝑗

 𝑗 = −1 and 𝑗 2 = −1

Prof. Muwaffaq I. Alomoush, Yarmouk University, Power Engineering Department, First Semester 2023/2024 4
Transmission Line
The transmission line parameters are 𝒓, 𝑳, 𝑪 𝐚𝐧𝐝 𝑮 . These
parameters are derived as per unit length of the transmission
line. They are not lumped and are uniformly distributed along the
length of the line. The transmission line models used in power
system analysis are developed using these distributed
parameters.
• 𝒓 represents the real power loss in the conductor.
• 𝑳 represents the magnetic field effect.
• 𝑪 represents the electric field effect.
• 𝒈 represents the real power loss caused by the leakage currents
and corona loss.

Prof. Muwaffaq I. Alomoush, Yarmouk University, Power Engineering Department, First Semester 2023/2024 5
Short Transmission Line Model
The transmission lines which have length less than
80 km are generally referred as short transmission
lines. For short length, the shunt capacitance of this
type of line is neglected and other parameters like
resistance and inductance of these short lines are
lumped.

Prof. Muwaffaq I. Alomoush, Yarmouk University, Power Engineering Department, First Semester 2023/2024 6
Medium Transmission Line Model

• The transmission lines above 80km and below 250km in length are
referred as medium length transmission lines.
• As the length of transmission line increases, the line charging current
becomes appreciable and the shunt capacitance must be considered.
• The modelling of a medium length transmission line is represented
using lumped shunt admittance along with the lumped impedance in
series to the circuit.
• A medium transmission line is represented as follows:

Prof. Muwaffaq I. Alomoush, Yarmouk University, Power Engineering Department, First Semester 2023/2024 7
Line Impedance and Line Admittance

𝒛𝑖𝑗 = 𝑟𝑖𝑗 + 𝑗𝑥𝑖𝑗

1 1 1 𝑟𝑖𝑗 − 𝑗𝑥𝑖𝑗 𝑟𝑖𝑗 − 𝑗𝑥𝑖𝑗
𝒚𝑖𝑗 = = = × = 2 2
𝒛𝑖𝑗 𝑟𝑖𝑗 + 𝑗𝑥𝑖𝑗 𝑟𝑖𝑗 + 𝑗𝑥𝑖𝑗 𝑟𝑖𝑗 − 𝑗𝑥𝑖𝑗 𝑟𝑖𝑗 + 𝑥𝑖𝑗
𝑟𝑖𝑗 −𝑥𝑖𝑗
= 2 2 +𝑗 2 2
𝑟𝑖𝑗 + 𝑥𝑖𝑗 𝑟𝑖𝑗 + 𝑥𝑖𝑗
𝑟𝑖𝑗 𝑥𝑖𝑗
𝒚𝑖𝑗 = 𝑔𝑖𝑗 + 𝑗 𝑏𝑖𝑗 ; 𝑔𝑖𝑗 = 2 +𝑥 2 and 𝑏𝑖𝑗 = − 2 +𝑥 2
𝑟𝑖𝑗 𝑖𝑗 𝑟𝑖𝑗 𝑖𝑗
Note that if 𝑟𝑖𝑗 is neglected (𝑟𝑖𝑗 = 0), then:
1 1
𝑔𝑖𝑗 = 0 , 𝑏𝑖𝑗 = − , 𝒚𝑖𝑗 = −𝑗
𝑥𝑖𝑗 𝑥𝑖𝑗

Prof. Muwaffaq I. Alomoush, Yarmouk University, Power Engineering Department, First Semester 2023/2024 8
Power System Quantities
 𝑽𝑖 = 𝑉𝑖 ∠𝛿𝑖 , 𝑽𝑗 = 𝑉𝑗 ∠𝛿𝑗
𝑽𝑖 −𝑽𝑗
 𝑰𝑖𝑗 = = (𝑽𝒊 − 𝑽𝒋 )𝒚𝑖𝑗 = (𝑉𝑖 ∠𝛿𝑖 − 𝑉𝑗 ∠𝛿𝑗 )(𝑔𝑖𝑗 + 𝑗𝑏𝑖𝑗 )
𝒛𝑖𝑗

= [𝑉𝑖 𝑐𝑜𝑠𝛿𝑖 + 𝑗𝑠𝑖𝑛𝛿𝑖 − 𝑉𝑗 𝑐𝑜𝑠𝛿𝑗 + 𝑗𝑠𝑖𝑛𝛿𝑗 ] 𝑔𝑖𝑗 + 𝑗 𝑏𝑖𝑗

𝑰𝑖𝑗 = 𝑔𝑖𝑗 𝑉𝑖 𝑐𝑜𝑠𝛿𝑖 − 𝑉𝑗 𝑐𝑜𝑠𝛿𝑗 − 𝑏𝑖𝑗 𝑉𝑖 𝑠𝑖𝑛𝛿𝑖 − 𝑉𝑗 𝑠𝑖𝑛𝛿𝑗
+ 𝑗 [𝑏𝑖𝑗 𝑉𝑖 𝑐𝑜𝑠𝛿𝑖 − 𝑉𝑗 𝑐𝑜𝑠𝛿𝑗 + 𝑔𝑖𝑗 𝑉𝑖 𝑠𝑖𝑛𝛿𝑖 − 𝑉𝑗 𝑠𝑖𝑛𝛿𝑗 ]
 Complex power flowing from bus to bus are given by:
𝑺𝑖𝑗 = 𝑽𝑖 𝑰∗𝑖𝑗 + 𝑽𝑖 𝑰∗𝑖𝑜 𝑺𝑖𝑗 = 𝑃𝑖𝑗 + 𝑗𝑄𝑖𝑗
 Complex power flowing from bus 𝑗 to bus 𝑖 (𝑺𝑗𝑖 ) is given by:
∗ ∗
𝑺𝑗𝑖 = 𝑽𝑗 𝑰𝑗𝑖 + 𝑽𝑗 𝑰𝑗𝑜
 Note that 𝑺𝑖𝑗 ≠ 𝑺𝑗𝑖 , and the complex power loss in the line 𝑖𝑗 is
given by: 𝑺𝑙𝑜𝑠𝑠𝑖𝑗 = 𝑺𝑖𝑗 + 𝑺𝑗𝑖

Prof. Muwaffaq I. Alomoush, Yarmouk University, Power Engineering Department, First Semester 2023/2024 9
Expressions of Line Power Flows

𝑺𝑖𝑗 = 𝑽𝑖 𝑰∗𝑖𝑜 + 𝑽𝑖 𝑰∗𝑖𝑗

∗ ∗
= 𝑉𝑖 ∠𝛿𝑖 𝑉𝑖 ∠𝛿𝑖 − 0 𝑔𝑖𝑜 + 𝑗𝑏𝑖𝑜 + 𝑉𝑖 ∠𝛿𝑖 𝑉𝑖 ∠𝛿𝑖 − 𝑉𝑗 ∠𝛿𝑗 𝑔𝑖𝑗 + 𝑗𝑏𝑖𝑗
𝑉𝑖 ∠𝛿𝑖 𝑉𝑖 ∠ −𝛿𝑖 𝑔𝑖𝑜 − 𝑗𝑏𝑖𝑜 + 𝑉𝑖 ∠𝛿𝑖 𝑉𝑖 ∠ −𝛿𝑖 −𝑉𝑗 ∠ −𝛿𝑗 𝑔𝑖𝑗 − 𝑗𝑏𝑖𝑗
= (𝑉𝑖2 𝑔𝑖𝑜 − 𝑗𝑉𝑖2 𝑏𝑖𝑜 + (𝑉𝑖2 − [𝑉𝑖 𝑉𝑖 cos (𝛿𝑖 −𝛿𝑗
+ 𝑗𝑉𝑖 𝑉𝑖 sin (𝛿𝑖 −𝛿𝑗 ] 𝑔𝑖𝑗 − 𝑗𝑏𝑖𝑗
𝑺𝑖𝑗 = 𝑉𝑖2 (𝑔𝑖𝑗 +𝑔𝑖𝑜 ) − 𝑉𝑖 𝑉𝑖 [𝑔𝑖𝑗 cos 𝛿𝑖 −𝛿𝑗 + 𝑏𝑖𝑗 sin 𝛿𝑖 −𝛿𝑗 ]
+𝑗 {−𝑉𝑖2 (𝑏𝑖𝑗 +𝑏𝑖𝑜 ) − 𝑉𝑖 𝑉𝑖 [𝑔𝑖𝑗 sin 𝛿𝑖 −𝛿𝑗 − 𝑏𝑖𝑗 cos 𝛿𝑖 −𝛿𝑗 ]}

Prof. Muwaffaq I. Alomoush, Yarmouk University, Power Engineering Department, First Semester 2023/2024 10
Expressions of Line Power Flows: Cont.

𝑺𝑖𝑗 = 𝑉𝑖2 (𝑔𝑖𝑗 +𝑔𝑖𝑜 ) − 𝑉𝑖 𝑉𝑖 [𝑔𝑖𝑗 cos 𝛿𝑖 −𝛿𝑗 + 𝑏𝑖𝑗 sin 𝛿𝑖 −𝛿𝑗 ]
+𝑗 {[−𝑉𝑖2 (𝑏𝑖𝑗 +𝑏𝑖𝑜 ) − 𝑉𝑖 𝑉𝑖 [𝑔𝑖𝑗 sin 𝛿𝑖 −𝛿𝑗 − 𝑏𝑖𝑗 cos 𝛿𝑖 −𝛿𝑗 ]}

But 𝑺𝑖𝑗 = 𝑃𝑖𝑗 + 𝑗𝑄𝑖𝑗 :

𝑃𝑖𝑗 = 𝑉𝑖2 (𝑔𝑖𝑗 +𝑔𝑖𝑜 ) − 𝑉𝑖 𝑉𝑖 [𝑔𝑖𝑗 cos 𝛿𝑖 −𝛿𝑗 + 𝑏𝑖𝑗 sin 𝛿𝑖 −𝛿𝑗 ]

𝑄𝑖𝑗 = −𝑉𝑖2 (𝑏𝑖𝑗 +𝑏𝑖𝑜 ) − 𝑉𝑖 𝑉𝑖 [𝑔𝑖𝑗 sin 𝛿𝑖 −𝛿𝑗 − 𝑏𝑖𝑗 cos 𝛿𝑖 −𝛿𝑗 ]

Prof. Muwaffaq I. Alomoush, Yarmouk University, Power Engineering Department, First Semester 2023/2024 11
Power System Matrices

Bus Admittance Matrix ( 𝒀𝒃𝒖𝒔 or 𝒀)

The first step in solving the power flow is to create what
is known as the bus admittance matrix, which is often
indicated by 𝒀𝒃𝒖𝒔 or 𝒀.

The 𝒀𝒃𝒖𝒔 gives the relationships between all the bus

injected currents, 𝑰 and all the bus voltages, 𝑽. The 𝒀𝒃𝒖𝒔
is developed by applying KCL at each bus in the system
to relate the bus current injections, the bus voltages, and
the branch impedances and admittances.
𝑰 = 𝒀 𝑽 or 𝑰 = 𝒀𝒃𝒖𝒔 𝑽

Prof. Muwaffaq I. Alomoush, Yarmouk University, Power Engineering Department, First Semester 2023/2024 12
Bus Admittance Matrix ( 𝒀𝒃𝒖𝒔 or 𝒀): Cont.

The admittance matrix has the following characteristics:

 Used in Load Flow Studies.
 Derived by KCL.
 Sparse (Majority of elements are zeros).
 Square
 𝒀 = 𝒀𝑻
 Easy to find (by inspection)
 Requires small storage in computer compared to
impedance matrix

13
Prof. Muwaffaq I. Alomoush, Yarmouk University, Power Engineering Department, First Semester 2023/2024
Net Injected Complex current: 𝑰𝑖
Next, we will derive the equations of net injected
complex currents at system buses. Let the bus 𝒊 is
connected to other buses by branches (lines) as
shown:

Prof. Muwaffaq I. Alomoush, Yarmouk University, Power Engineering Department, First Semester 2023/2024 14
Net Injected Complex current: Cont.

The net injected complex current at bus 𝑖 ; 𝑰𝑖 = 𝑰𝐺𝑖 − 𝑰𝐷𝑖

can be expressed as:

𝑰𝑖 = 𝑰𝑖𝑜 +𝑰𝑖1 +𝑰𝑖2 + ⋯ + 𝑰𝑖𝑛

𝑛 𝒏

𝑰𝑖 = 𝑰𝑖𝑜 + ෍ 𝑰𝑖𝑗 = 𝑽𝑖 𝒚𝑖𝑜 + ෍(𝑽𝑖 −𝑽𝑗 )𝒚𝑖𝑗

𝑗=1 𝑗=1
𝑗≠𝑖

𝑰𝑖 = (𝑽𝑖 −0)𝒚𝑖𝑜 + (𝑽𝑖 −𝑽1 )𝒚𝑖1 + (𝑽𝑖 −𝑽2 )𝒚𝑖2 + (𝑽𝑖 −𝑽3 )𝒚𝑖3 + ⋯ + (𝑽𝑖 −𝑽𝑛 )𝒚𝑖𝑛

𝑰𝑖 = 𝒚𝑖𝑜 + 𝒚𝑖2 + 𝒚𝑖3 + ⋯ + 𝒚𝑖𝑛 𝑽𝑖 − 𝒚𝑖1 𝑽1 − 𝒚𝑖2 𝑽2 − 𝒚𝑖3 𝑽3 − ⋯ − 𝒚𝑖𝑛 𝑽𝑛

Prof. Muwaffaq I. Alomoush, Yarmouk University, Power Engineering Department, First Semester 2023/2024 15
Net Injected Complex current: Cont

If we expand the last equation for all buses, we obtain:

𝑰1 = 𝒚1𝑜 + 𝒚12 + 𝒚13 + ⋯ + 𝒚1𝑛 𝑽1 − 𝒚12 𝑽2 − 𝒚13 𝑽3 − ⋯ − 𝒚1𝑛 𝑽𝑛

𝑰2 = −𝒚21 𝑽1 + 𝒚2𝑜 + 𝒚22 + 𝒚23 + ⋯ + 𝒚2𝑛 𝑽2 − 𝒚23 𝑽3 − ⋯ − 𝒚2𝑛 𝑽𝑛
𝑰3 = −𝒚31 𝑽1 − 𝒚32 𝑽3 + 𝒚3𝑜 + 𝒚32 + 𝒚33 + ⋯ + 𝒚3𝑛 𝑽3 − ⋯ − 𝒚3𝑛 𝑽𝑛
⋮

𝑰𝑛 = −𝒚𝑛1 𝑽1 − 𝒚𝑛2 𝑽2 − 𝒚𝑛3 𝑽3 + ⋯ + 𝒚𝑛𝑜 + 𝒚𝑛1 + 𝒚𝑛2 + 𝒚3𝑛 + ⋯ + 𝒚(𝒏−𝟏)(𝒏−𝟏 𝑽𝑛

Prof. Muwaffaq I. Alomoush, Yarmouk University, Power Engineering Department, First Semester 2023/2024 16
Net Injected Complex current: Cont

The net injected currents can be expressed in terms

of the bus admittance elements as:

𝑰1 = 𝒀11 𝑽1 +𝒀12 𝑽2 + ⋯ + 𝒀1𝑖 𝑽𝑖 + ⋯ + 𝒀1𝑛 𝑽𝑛

𝑰2 = 𝒀21 𝑽1 +𝒀22 𝑽2 + ⋯ + 𝒀2𝑖 𝑽𝑖 + ⋯ + 𝒀2𝑛 𝑽𝑛
𝑰3 = 𝒀31 𝑽1 +𝒀32 𝑽2 + ⋯ + 𝒀3𝑖 𝑽𝑖 + ⋯ + 𝒀3𝑛 𝑽𝑛
⋮
𝑰𝑛 = 𝒀𝑛1 𝑽1 +𝒀𝑛2 𝑽2 + ⋯ + 𝒀𝑛𝑖 𝑽𝑖 + ⋯ + 𝒀𝑛𝑛 𝑽𝑛

Prof. Muwaffaq I. Alomoush, Yarmouk University, Power Engineering Department, First Semester 2023/2024 17
Net Injected Complex current: Cont
In Matrix Form:
𝑰1 𝒀11 𝒀12 ⋯ 𝒀1𝑛 𝑽1 𝒀11 𝒀12 ⋯ 𝒀1𝑛
𝑰2 𝒀 𝒀22 ⋯ 𝒀1𝑛 𝑽2 𝒀 𝒀22 ⋯ 𝒀2𝑛
= 21 ; 𝒀 = 21
⋮ ⋮ ⋮ ⋱ ⋮ ⋮ ⋮ ⋮ ⋱ ⋮
𝑰𝑛 𝒀𝑛1 𝒀𝑛2 … 𝒀𝑛𝑛 𝑽𝑛 𝒀𝑛1 𝒀𝑛2 … 𝒀𝑛𝑛

𝒀𝑖𝑖 = 𝒚𝑖𝑜 + 𝒚𝑖1 + 𝒚𝑖2 + ⋯ + 𝒚𝑖𝑛

𝒀𝑖𝑗 = −𝒚𝑖𝑗 ; 𝑖 ≠ 𝑗
𝒀𝑖𝑗 = 𝒀𝑗𝑖 ; 𝑖 ≠ 𝑗
𝑰𝑖 = 𝒀𝑖1 𝑽𝑖 + 𝒀𝑖2 𝑽2 + ⋯ + 𝒚𝑖𝑖 𝑽𝑖 + ⋯ + 𝒀𝑖𝑛 𝑽𝑛
𝒏

𝑰𝑖 = ෍ 𝒀𝑖𝑗 𝑽𝑗
𝑗=1

Prof. Muwaffaq I. Alomoush, Yarmouk University, Power Engineering Department, First Semester 2023/2024 18
 𝑌11 ∠𝜃11 𝑌12 ∠𝜃12 ⋯ 𝑌1𝑛 ∠𝜃1𝑛
𝑌21 ∠𝜃21 𝑌22 ∠𝜃22 ⋯ 𝑌2𝑛 ∠𝜃2𝑛
𝒀= Polar Form
⋮ ⋮ ⋱ ⋮
𝑌𝑛1 ∠𝜃𝑛1 𝑌𝑛2 ∠𝜃𝑛2 … 𝑌𝑛𝑛 ∠𝜃𝑛𝑛

𝐺11 + 𝑗𝐵11 𝐺12 + 𝑗𝐵12 ⋯ 𝐺1𝑛 + 𝑗𝐵1𝑛

𝐺21 + 𝑗𝐵21 𝐺22 + 𝑗𝐵22 ⋯ 𝐺2𝑛 + 𝑗𝐵2𝑛
𝒀= Rectangular Form
⋮ ⋮ ⋱ ⋮
𝐺𝑛1 + 𝑗𝐵𝑛1 𝐺𝑛2 + 𝑗𝐵𝑛2 … 𝐺𝑛𝑛 + 𝑗𝐵𝑛𝑛

𝒀=𝑮 +𝒋 𝑩
𝐺11 𝐺12 ⋯ 𝐺1𝑛 𝐵11 𝐵12 ⋯ 𝐵1𝑛
𝐺 𝐺22 ⋯ 𝐺2𝑛 𝐵 𝐵22 ⋯ 𝐵2𝑛
𝒀 = 21 + 𝑗 21
⋮ ⋮ ⋱ ⋮ ⋮ ⋮ ⋱ ⋮
𝐺𝑛1 𝐺𝑛2 … 𝐺𝑛𝑛 𝐵𝑛1 𝐵𝑛2 … 𝐵𝑛𝑛
Prof. Muwaffaq I. Alomoush, Yarmouk University, Power Engineering Department, First Semester 2023/2024 19
Example: Nodal Analysis
For the three-bus system shown, find the nodal
equations, admittance matrix, conductance matrix, and
susceptance matrix.

Prof. Muwaffaq I. Alomoush, Yarmouk University, Power Engineering Department, First Semester 2023/2024 20
By Nodal Analysis (KCL):
(𝑽1 −𝟎) (𝑽1 −𝑽2 ) (𝑽1 −𝑽3 )
𝑰1 = + +
𝒛1𝑜 𝒛12 𝒛13
= (𝑽1 −𝟎)𝒚1𝑜 + (𝑽1 −𝑽2 )𝒚12 + (𝑽1 −𝑽3 )𝒚13
= 𝒚1𝑜 + 𝒚12 + 𝒚13 𝑽1 − 𝒚12 𝑽2 − 𝒚13 𝑽3
𝑰1 = 𝒀11 𝑽1 + 𝒀12 𝑽2 + 𝒀13 𝑽3
(𝑽2 −𝟎) (𝑽2 −𝑽1 ) (𝑽2 −𝑽3 )
𝑰2 = + +
𝒛2𝑜 𝒛21 𝒛23
= (𝑽2 −𝟎)𝒚2𝑜 + (𝑽2 −𝑽1 )𝒚21 + (𝑽2 −𝑽3 )𝒚23
= −𝒚21 𝑽1 + 𝒚2𝑜 + 𝒚21 + 𝒚23 𝑽2 − 𝒚23 𝑽3
𝑰2 = 𝒀21 𝑽1 + 𝒀22 𝑽2 + 𝒀23 𝑽3
(𝑽3 −𝟎) (𝑽3 −𝑽1 ) (𝑽3 −𝑽2 )
𝑰3 = + +
𝒛3𝑜 𝒛31 𝒛32
= (𝑽3 −𝟎)𝒚3𝑜 + (𝑽3 −𝑽1 )𝒚31 + (𝑽3 −𝑽2 )𝒚32
= −𝒚31 𝑽1 − 𝒚32 𝑽2 + 𝒚3𝑜 + 𝒚31 + 𝒚32 𝑽3
𝑰3 = 𝒀31 𝑽1 + 𝒀32 𝑽2 + 𝒀33 𝑽3
21
Note that 𝑰𝟐 = 𝟎
(No current source: No net injected current at bus 2)

𝑰1 𝒀11 𝒀12 𝒀13 𝑽1 𝒚1𝑜 + 𝒚12 + 𝒚13 −𝒚12 −𝒚13 𝑽1

𝑰2 = 𝒀21 𝒀22 𝒀23 𝑽2 = −𝒚21 𝒚2𝑜 + 𝒚21 + 𝒚23 −𝒚23 𝑽2
𝑰3 𝒀31 𝒀32 𝒀33 𝑽3 −𝒚31 −𝒚32 𝒚3𝑜 + 𝒚31 + 𝒚32 𝑽3

𝑰1 18 − 𝑗22 −12 + 𝑗16 −6 + 𝑗8 𝑽1

𝑰2 = −12 + 𝑗16 12 − 𝑗34 0 + 𝑗20 𝑽2
𝑰3 −6 + 𝑗8 0 + 𝑗20 6 − 𝑗28 𝑽3
18 − 𝑗22 −12 + 𝑗16 −6 + 𝑗8 18 −12 −6 −22 16 8
𝒀 = −12 + 𝑗16 12 − 𝑗34 0 + 𝑗20 = −12 12 0 + 𝑗 16 −34 20
−6 + 𝑗8 0 + 𝑗20 6 − 𝑗28 −6 0 6 8 20 −28
18 −12 −6 −22 16 8
𝑮 = −12 12 0 , 𝑩 = 16 −34 20
−6 0 6 8 20 −28
Prof. Muwaffaq I. Alomoush, Yarmouk University, Power Engineering Department, First Semester 2023/2024 22
The net injected complex power at bus 𝒊 : 𝑺𝒊

The net injected complex power at bus 𝑖 ; 𝑺𝑖 = 𝑺𝐺𝑖 − 𝑺𝐷𝑖 is given by:
∗
𝒏 𝒏

𝑺𝑖 = 𝑽𝑖 𝑰∗𝑖 = 𝑽𝑖 ෍ 𝒀𝑖𝑗 𝑽𝑗 = 𝑽𝑖 ෍ 𝒀∗𝑖𝑗 𝑽𝑗∗

𝑗=1 𝑗=1
𝒏

= 𝑉𝑖 ∠𝛿𝑖 ෍ 𝑌𝑖𝑗 ∠ −𝜃𝑖𝑗 𝑉𝑗 ∠ − 𝛿𝑗

𝑗=1
𝒏 𝒏

= ෍ 𝑉𝑖 ∠𝛿𝑖 𝑉𝑗 ∠ − 𝛿𝑗 𝑌𝑖𝑗 ∠ −𝜃𝑖𝑗 = ෍ 𝑉𝑖 𝑉𝑗 𝑌𝑖𝑗 ∠(𝛿𝑖 −𝛿𝑗 −𝜃𝑖𝑗 )

𝑗=1 𝑗=1

𝒏 𝒏

= ෍ 𝑉𝑖 𝑉𝑗 𝑌𝑖𝑗 𝑐𝑜𝑠(𝛿𝑖 −𝛿𝑗 −𝜃𝑖𝑗 ) + 𝑗 ෍ 𝑉𝑖 𝑉𝑗 𝑌𝑖𝑗 𝑠𝑖𝑛(𝛿𝑖 −𝛿𝑗 −𝜃𝑖𝑗 )

𝑗=1 𝑗=1

Prof. Muwaffaq I. Alomoush, Yarmouk University, Power Engineering Department, First Semester 2023/2024 23
Expressions of Net Injected Real Power and Reactive Power

The net injected real power at bus 𝑖 (𝑃𝑖 = 𝑃𝐺𝑖 − 𝑃𝐷𝑖 ) is given by:
𝒏

𝑃𝑖 = ෍ 𝑉𝑖 𝑉𝑗 𝑌𝑖𝑗 𝑐𝑜𝑠(𝛿𝑖 −𝛿𝑗 −𝜃𝑖𝑗 )

𝑗=1
𝒏

or 𝑃𝑖 = ෍ |𝑽𝑖 |𝑽𝑗 |𝒀𝑖𝑗 | 𝑐𝑜𝑠(𝛿𝑖 −𝛿𝑗 −𝜃𝑖𝑗 )

𝑗=1

The net injected real power at bus 𝑖 (𝑄𝑖 = 𝑄𝐺𝑖 − 𝑄𝐷𝑖 ) is given by:
𝒏

𝑄𝑖 = ෍ 𝑉𝑖 𝑉𝑗 𝑌𝑖𝑗 𝑠𝑖𝑛(𝛿𝑖 −𝛿𝑗 −𝜃𝑖𝑗 ) or 𝑄𝑖

𝑗=1
𝒏

= ෍ |𝑽𝑖 |𝑽𝑗 |𝒀𝑖𝑗 | 𝑠𝑖𝑛(𝛿𝑖 −𝛿𝑗 −𝜃𝑖𝑗 )

𝑗=1

Prof. Muwaffaq I. Alomoush, Yarmouk University, Power Engineering Department, First Semester 2023/2024 24
Example
For the three-bus system shown, write the expressions of 𝑃2 , 𝑃3 , 𝑄2 , and
𝑄3

Solution:
Number of buses is 𝑛 = 3
The net injected real power and net injected reactive power at bus 𝑖 is given by:
𝑛=3 𝑛=3

𝑃𝑖 = ෍ 𝑉𝑖 𝑉𝑗 𝑌𝑖𝑗 𝑐𝑜𝑠(𝛿𝑖 −𝛿𝑗 −𝜃𝑖𝑗 ) 𝑄𝑖 = ෍ 𝑉𝑖 𝑉𝑗 𝑌𝑖𝑗 𝑠𝑖𝑛(𝛿𝑖 −𝛿𝑗 −𝜃𝑖𝑗 )

𝑗=1 𝑗=1

Prof. Muwaffaq I. Alomoush, Yarmouk University, Power Engineering Department, First Semester 2023/2024 25
 𝑛=3

𝑃2 = ෍ 𝑉2 𝑉𝑗 𝑌2𝑗 𝑐𝑜𝑠(𝛿2 −𝛿𝑗 −𝜃2𝑗 ) = 𝑉2 𝑉1 𝑌21 𝑐𝑜𝑠(𝛿2 −𝛿1 −𝜃12 )

𝑗=1
+𝑉2 𝑉2 𝑌22 𝑐𝑜𝑠(𝛿2 −𝛿2 −𝜃22 ) +𝑉2 𝑉3 𝑌23 𝑐𝑜𝑠(𝛿2 −𝛿3 −𝜃23 )
𝑛=3

𝑃3 = ෍ 𝑉3 𝑉𝑗 𝑌3𝑗 𝑐𝑜𝑠(𝛿3 −𝛿𝑗 −𝜃3𝑗 ) = 𝑉3 𝑉1 𝑌31 𝑐𝑜𝑠(𝛿3 −𝛿1 −𝜃31 )

𝑗=1
+𝑉3 𝑉2 𝑌32 𝑐𝑜𝑠(𝛿3 −𝛿2 −𝜃32 ) +𝑉3 𝑉3 𝑌33 𝑐𝑜𝑠(𝛿3 −𝛿3 −𝜃33 )
𝑛=3

𝑄2 = ෍ 𝑉2 𝑉𝑗 𝑌2𝑗 𝑠𝑖𝑛(𝛿2 −𝛿𝑗 −𝜃2𝑗 ) = 𝑉2 𝑉1 𝑌21 𝑠𝑖𝑛(𝛿2 −𝛿1 −𝜃12 )

𝑗=1
+𝑉2 𝑉2 𝑌22 𝑠𝑖𝑛(𝛿2 −𝛿2 −𝜃22 ) +𝑉2 𝑉3 𝑌23 𝑠𝑖𝑛(𝛿2 −𝛿3 −𝜃23 )
𝑛=3

𝑄3 = ෍ 𝑉3 𝑉𝑗 𝑌3𝑗 𝑠𝑖𝑛(𝛿3 −𝛿𝑗 −𝜃3𝑗 ) = 𝑉3 𝑉1 𝑌31 𝑠𝑖𝑛(𝛿3 −𝛿1 −𝜃31 )

𝑗=1
+𝑉3 𝑉2 𝑌32 𝑠𝑖𝑛(𝛿3 −𝛿2 −𝜃32 ) + 𝑉3 𝑉3 𝑌33 𝑠𝑖𝑛(𝛿3 −𝛿3 −𝜃33 ) 26
Example:
For the three-bus system shown, if all values are given in p.u., find
the admittance, conductance, and susceptance matrices.

Prof. Muwaffaq I. Alomoush, Yarmouk University, Power Engineering Department, First Semester 2023/2024 27
𝒀11 = 𝒚1𝑜 + 𝒚12 + 𝒚13 𝒀12 = −𝒚12 = 𝒀12 𝒀13 = −𝒚13 = 𝒀31
𝒀22 = 𝒚2𝑜 + 𝒚22 + 𝒚23 𝒀23 = −𝒚23 = 𝒀32 𝒀33 = 𝒚3𝑜 + 𝒚31 + 𝒚32
9 − 𝑗13 −6 + 𝑗8 −3 + 𝑗4
𝒀 = −6 + 𝑗8 6 − 𝑗18 0 + 𝑗10
−3 + 𝑗4 0 + 𝑗10 3 − 𝑗16
9 −6 −3 −13 8 4
= −6 6 0 +𝑗 8 −18 10
−3 0 3 4 10 −16

9 −6 −3 −13 8 4
𝑮 = −6 6 0 , 𝑩= 8 −18 10
−3 0 3 4 10 −16 28
Prof. Muwaffaq I. Alomoush, Yarmouk University, Power Engineering Department, First Semester 2023/2024
Example: Advantages of Improving Power Factor

For the system shown, if we want to keep 𝑽𝐿 = 32 𝑘𝑉,

and the production cost of the generating unit is given by:
𝐶𝐺 𝑃𝐺 = 500 + 5.3𝑃𝐺 + 0.004 𝑃𝐺2 $/ℎ.
If the complex power of the generator and the load are
given by: 𝑺𝐺 = 𝑃𝐺 + 𝑗𝑄𝐺 and 𝑺𝐿 = 𝑃𝐿 + 𝑗𝑄𝐿

Prof. Muwaffaq I. Alomoush, Yarmouk University, Power Engineering Department, First Semester 2023/2024 29
Example: Cont.
Calculate:
 𝒁𝑒𝑞 of the load , 𝑝. 𝑓 of the load, 𝑰𝑙𝑖𝑛𝑒 ,
 Drop voltage in the line 𝑽𝑑𝑟𝑜𝑝,𝑙𝑖𝑛𝑒 , Source voltage 𝑽𝑠
 Complex power of the load (𝑺𝐿 )
 Complex power of the source (𝑺𝑆 )
 Complex power loss in line (𝑺𝐿𝑖𝑛𝑒 )
 𝑝. 𝑓 of the source.
 Cost to produce power

Resolve if a capacitor with 𝑍𝑐 = −𝑗57.14 is added in parallel with

the load. Compare results and comment.

Prof. Muwaffaq I. Alomoush, Yarmouk University, Power Engineering Department, First Semester 2023/2024 30
Solution:
𝒁𝑒𝑞 = 100 ∥ 10 + 𝑗20 = 20∠53.13𝑜
𝑽𝐿 32000∠0𝑜 𝑜
𝑰𝑙𝑖𝑛𝑒 = = == 960 − 𝑗1280 𝐴 = 1600∠ −53.13
𝒁𝑒𝑞 20∠53.13𝑜
𝑽𝑑𝑟𝑜𝑝,𝑙𝑖𝑛𝑒 = 𝑰𝑙𝑖𝑛𝑒 𝒁𝑙𝑖𝑛𝑒 = 7.36 + 𝑗 3.52 𝑘𝑉 = 8.158∠25.56𝑜 𝑘𝑉
𝑽𝑆 = 𝑽𝑑𝑟𝑜𝑝,𝑙𝑖𝑛𝑒 + 𝑽𝐿 = 39.36 + 𝑗 3.52 𝑘𝑉 = 39.517∠5.1104𝑜 𝑘𝑉
𝑺𝐿 = 𝑽𝐿 𝑰∗𝑙𝑖𝑛𝑒 = 32000∠0𝑜 × 1600∠53.13𝑜 = 30.72 𝑀𝑊 + 𝑗 40.96 𝑀𝑣𝑎𝑟
= 51.2 ∠53.13𝑜 𝑀𝑉𝐴
𝑺𝑆 = 𝑽𝑆 𝑰∗𝑙𝑖𝑛𝑒 = 39.517∠5.1104𝑜 × 1600∠ −53.13𝑜
= 33.28 𝑀𝑊 + 𝑗 53.13 𝑀𝑣𝑎𝑟 = 63.513 ∠58.24𝑜 𝑀𝑉𝐴
𝑺𝐿𝑖𝑛𝑒 = 𝑺𝑆 − 𝑺𝐿 = 2.564 𝑀𝑊 + 𝑗12.8 𝑀𝑣𝑎𝑟 = 13.053∠78.69𝑜 𝑀𝑉𝐴
𝑝. 𝑓𝑆 = cos 58.24𝑜 = 0.5264 (𝑙𝑎𝑔𝑔𝑖𝑛𝑔)
𝑝. 𝑓𝐿 = cos 53.13𝑜 = 0.60 (𝑙𝑎𝑔𝑔𝑖𝑛𝑔)
𝐶𝐺 𝑃𝐺 = 33.28 𝑀𝑊 = 500 + 5.3 33.28 + 0.004 33.28 2 = 680.81$/ℎ

Prof. Muwaffaq I. Alomoush, Yarmouk University, Power Engineering Department, First Semester 2023/2024 31
If a capacitor with 𝒁𝒄 = −𝒋𝟓𝟕. 𝟏𝟒 is added in parallel with the load:
Solution:
𝒁 = −𝑗57.14 ∥ 100 ∥
𝑒𝑞 10 + 𝑗20 = 26.667∠36.87𝑜 
𝑽𝐿 32000∠0𝑜 𝑜 𝑘𝐴
𝑰𝑙𝑖𝑛𝑒 = = = 960 − 𝑗720 𝐴 = 1.200∠ −36.87
𝒁𝑒𝑞 26.667∠36.87𝑜
𝑽𝑑𝑟𝑜𝑝,𝑙𝑖𝑛𝑒 = 𝑰𝑙𝑖𝑛𝑒 𝒁𝑙𝑖𝑛𝑒 = 4.56 + 𝑗4.08 𝑘𝑉 = 6.119∠41.82𝑜 𝑘𝑉

𝑽𝑆 = 𝑽𝑑𝑟𝑜𝑝,𝑙𝑖𝑛𝑒 + 𝑽𝐿 = 36.56 + 𝑗4.08 𝑘𝑉 = 36.8787∠6.3678𝑜 𝑘𝑉

𝑺𝐿 = 𝑽𝐿 𝑰∗𝑙𝑖𝑛𝑒 = 32000∠0𝑜 × 1200∠36.87𝑜 = 30.72 𝑀𝑊 + 𝑗23.039 𝑀𝑣𝑎𝑟

= 38.399 ∠36.87𝑜 𝑀𝑉𝐴

𝑺𝑆 = 𝑽𝑆 𝑰∗𝑙𝑖𝑛𝑒 = 36.8787∠6.3678𝑜 𝑘𝑉 × 1200∠36.87𝑜

= 32.16 𝑀𝑊 + 𝑗 30.24 𝑀𝑣𝑎𝑟 = 44.144 ∠43.237𝑜 𝑀𝑉𝐴

𝑺𝐿𝑖𝑛𝑒 = 𝑺𝑆 − 𝑺𝐿 = 1.44 𝑀𝑊 + 𝑗7.20 𝑀𝑣𝑎𝑟 = 7.342∠78.69𝑜 𝑀𝑉𝐴

𝑝. 𝑓𝑆 = cos 58.24𝑜 = 0.5264 (𝑙𝑎𝑔𝑔𝑖𝑛𝑔)
𝑝. 𝑓𝐿 = cos 43.237𝑜 = 0.7285 (𝑙𝑎𝑔𝑔𝑖𝑛𝑔)
𝐶𝐺 𝑃𝐺 = 32.16 𝑀𝑊 = 500 + 5.3 32.16 + 0.004 32.16 2 = 674.58 $/ℎ
Prof. Muwaffaq I. Alomoush, Yarmouk University, Power Engineering Department, First Semester 2023/2024 32
 Comparison between the two cases

quantity Before adding Capacitor Notes After adding Capacitor

𝒁𝑒𝑞 20∠53.13𝑜 26.667∠36.87𝑜 

𝑰𝑙𝑖𝑛𝑒 1600∠−53.13𝑜 Large current 1200∠−36.87𝑜 𝐴 (Smaller)

𝑽𝑑𝑟𝑜𝑝 ,𝑙𝑖𝑛𝑒 7.36 + 𝑗 3.52 𝑘𝑉 = 8.158∠25.56𝑜 Large drop 4.56 + 𝑗4.08 𝑘𝑉 = 6.119∠41.82𝑜 𝑘𝑉 (Less)

𝑽𝑆 39.517∠5.1104𝑜 Large Voltage (Size) 36.8787∠6.3678𝑜 𝑘𝑉 (Needs Smaller Size generator)

𝑺𝐿 30.72 𝑀𝑊 + 𝑗 40.96 𝑀𝑣𝑎𝑟 = 51.2 ∠53.13𝑜 𝑀𝑉𝐴 Large MVA (Cost) 30.72 𝑀𝑊 + 𝑗23.039 𝑀𝑣𝑎𝑟 = 38.399 ∠36.87𝑜 𝑀𝑉𝐴

𝑺𝑆 33.28 𝑀𝑊 + 𝑗 53.13 𝑀𝑣𝑎 = 63.513 ∠58.24𝑜 Large MVA (Size) 32.16 𝑀𝑊 + 𝑗 30.24 𝑀𝑣𝑎𝑟 = 44.144 ∠43.237𝑜

𝑺𝐿𝑖𝑛𝑒 2.564 𝑀𝑊 + 𝑗12.8 𝑀𝑣𝑎𝑟 Large losses (Cost) 2.564 𝑀𝑊 + 𝑗12.8 𝑀𝑣𝑎𝑟

𝑝. 𝑓𝑆 0.5264 (𝑙𝑎𝑔𝑔𝑖𝑛𝑔) Bad (Poor p.f.) 0.729 (𝑙𝑎𝑔𝑔𝑖𝑛𝑔)

𝑝. 𝑓𝐿 0.6000 (𝑙𝑎𝑔𝑔𝑖𝑛𝑔) Bad (Poor p.f.) 0.800 (𝑙𝑎𝑔𝑔𝑖𝑛𝑔)

𝐶𝐺 (𝑃𝐺 ) 680.8147 $/ℎ Larger cost 674.58 $/ℎ

Prof. Muwaffaq I. Alomoush, Yarmouk University, Power Engineering Department, First Semester 2023/2024 33
Newton-Raphson (NR) Method

Single-variable function
Solve 𝑓 𝑥 = 0 i.e., find the root (roots) of 𝑓 𝑥 , i.e.,
find 𝑥 that makes 𝑓(𝑥) = 0.

Solution
𝑓(𝑥 (𝑘) )
𝑥 (𝑘+1) = 𝑥 (𝑘) −
𝑑𝑓 (𝑘)
𝑑𝑥
Start by a certain value as an initial guess for the solution
𝑥 0 then continue iterations until a certain accuracy is
obtained such the |𝑥 𝑘+1 − 𝑥 𝑘 | ≤∈

Prof. Muwaffaq I. Alomoush, Yarmouk University, Power Engineering Department, First Semester 2023/2024 34
 Newton-Raphson Method Graphically

Prof. Muwaffaq I. Alomoush, Yarmouk University, Power Engineering Department, First Semester 2023/2024 35
Example:
Solve 𝑥 2 − 5𝑥 + 4 = 0 using NR. Assume 𝑥 (0) = 3 and ∈= 0.001

𝑓 (𝑥 ) = 𝑥 2 − 5𝑥 + 4
𝑑𝑓
= 2𝑥 − 5
𝑑𝑥

(𝑘+1) (𝑘)
𝑓(𝑥 (𝑘) )
𝑥 =𝑥 − (𝑘)
𝑑𝑓
𝑑𝑥
(𝑘)
(𝑘+1) (𝑘)
(𝑥 2 − 5𝑥 (𝑘) + 4)
𝑥 =𝑥 −
2𝑥 (𝑘) − 5
First iteration:
(0)
(1) (0)
(𝑥 2 − 5𝑥 (0) + 4) (32 − 5 × 3 + 4) −2
𝑥 =𝑥 − = 3 − = 3 − =5
2𝑥 (0) − 5 2×3−5 1
Check |𝑥 (1) − 𝑥 (0) | ≤∈ ∶ No

Second iteration:
(1)
(2) (1)
(𝑥 2 − 5𝑥 (1) + 4) (52 − 5 × 5 + 4) 4
𝑥 =𝑥 − = 5 − = 5 − = 4.2
2𝑥 (1) − 5 2×5−5 5
Check |𝑥 (2) − 𝑥 (1) | ≤∈ ∶ No
Continue until convergence or divergence happens.
36
Prof. Muwaffaq I. Alomoush, Yarmouk University, Power Engineering Department, First Semester 2023/2024
 multi-variable function
Solve
𝑓1 (𝑥1 , 𝑥2 , … , 𝑥𝑛 ) = 0
𝑓2 (𝑥1 , 𝑥2 , … , 𝑥𝑛 ) = 0 𝑛 equations with n unknowns (𝑥1 , 𝑥2 , … , 𝑥𝑛 )
⋮
𝑓𝑛 (𝑥1 , 𝑥2 , … , 𝑥𝑛 ) = 0 Similar to NR Load Flow equations:
Unknowns: 𝛿𝑖 and 𝑉𝑖 ≡ |𝑽𝑖 |
Solution:
(𝑘 ) −1
𝜕𝑓1 𝜕𝑓1 𝜕𝑓1
⋯
(𝑘+1) (𝑘 ) 𝜕𝑥1 𝜕𝑥2 𝜕𝑥𝑛 (𝑘 )
𝑥1 𝑥1 𝑓1
𝜕𝑓2 𝜕𝑓2 𝜕𝑓2
𝑥2 𝑥 ⋯ 𝑓2
⋮ = ⋮2 − 𝜕𝑥1 𝜕𝑥2 𝜕𝑥𝑛 ⋮
𝑥𝑛 𝑥𝑛 ⋮ ⋮ ⋱ ⋮ 𝑓𝑛
𝜕𝑓𝑛 𝜕𝑓𝑛 𝜕𝑓𝑛
…
𝜕𝑥1 𝜕𝑥2 𝜕𝑥𝑛
Jacobian Matrix ( 𝐽 )
In compact form:
−1
[𝒙](𝑘+1) = [𝒙](𝑘) − 𝑱(𝑘) 𝒇(𝑘)
−1
or [𝒙](𝑘+1) = [𝒙](𝑘) − [∆𝒙](𝑘) ; [∆𝒙](𝑘) = 𝑱(𝑘) 𝒇(𝑘)

Start by an initial guess for each of 𝑥1 0 , 𝑥2 0 , … , 𝑥𝑛 0 then iterate and stop when all
𝑥𝑖 𝑘+1 − 𝑥𝑖 𝑘 ≤∈ ; 𝑖 = 1, 2, … , 𝑛

Prof. Muwaffaq I. Alomoush, Yarmouk University, Power Engineering Department, First Semester 2023/2024 37
Load (Power) Flow Analysis
 Load flow studies are probably the most common of all
power system analysis calculations.
 They are used in planning studies to determine if and
when specific elements will become overloaded or its
values exceed limits.
 Major investment decisions begin with reinforcement
strategies based on load-flow analysis.
 The objective of any load-flow program is to find voltage
magnitude and phase angle at each bus, which are then
used to calculate other quantiles such as line current,
real and reactive power flowing in each element,
reactive power loading on each generator, line losses.

Prof. Muwaffaq I. Alomoush, Yarmouk University, Power Engineering Department, First Semester 2023/2024 38
Load (Power) Flow Analysis: Cont.
Given:
 Power system topology (number of buses, number of lines
(branches) and their connecting buses, number of generators
and their locations), line (branch) parameters,
 Real power generated and voltage magnitude at each
generator bus,
 Complex power of loads at each bus, voltage magnitude
 Phase-angle at one bus, which is the reference point for the
rest of the system.
Find:
 Voltage magnitude and phase angle at each bus (by writing a
set of nonlinear algebraic equations and solving them by
numerical iterative methods).


Prof. Muwaffaq I. Alomoush, Yarmouk University, Power Engineering Department, First Semester 2023/2024 39
Load (Power) Flow Analysis: Cont.
Information available from load-flow studies:
 All bus voltage magnitudes.
 Voltage phase angles w.r.t the slack bus.
 All bus active and reactive power injections.
 All line sending- and receiving-end complex power
flows.
 Individual line losses can be deduced by subtracting
receiving-end complex power from sending-end
complex power.
 Line losses and total system losses
 The most important information obtained from the
load-flow is the voltage profile of the system.

Prof. Muwaffaq I. Alomoush, Yarmouk University, Power Engineering Department, First Semester 2023/2024 40
Load (Power) Flow Analysis: Cont.

 The most important information obtained from the

load-flow is the voltage profile of the system.
 If │Vi│ varies greatly over the system, large reactive
flows will result; this, in turn, will lead to increased real
power losses and, in extreme cases, an increased
likelihood of voltage collapse.
 When a particular bus has an unacceptably low
voltage, the usual practice is to install capacitor banks
in order to provide reactive compensation to the load.
 Load-flow studies are used to determine how much
reactive compensation should be applied at a load
buses, to bring its voltage up to an appropriate level.
Prof. Muwaffaq I. Alomoush, Yarmouk University, Power Engineering Department, First Semester 2023/2024 41
Types of Buses
At each bus of the power system, there are four (4) quantities:
 Voltage magnitude: 𝑽𝑖 or 𝑉𝑖
 Voltage phase angle: 𝛿𝑖 .
 Real power: 𝑃𝑖 (Recall 𝑃𝑖 = 𝑃𝐺𝑖 −𝑃𝐷𝑖 )
 Reactive power: 𝑄𝑖 (Recall 𝑄𝑖 = 𝑄𝐺𝑖 −𝑄𝐷𝑖 )
Based on what is known and what is unknown of the four
quantities at each bus, power system buses are classified as three
types:
 Reference (or swing or slack) bus: this bus has a generator, and its
voltage magnitude ( 𝑉𝑖 𝑜𝑟 |𝑽𝑖 | ) and phase angle ( 𝛿𝑟𝑒𝑓 = 0 ) are
known. The generator at the reference bus compensates losses in
the system (transmission lines)
 Generator buses or PV buses or voltage-controlled buses: Knowns:
𝑃𝑖 and 𝑉𝑖 , Unknowns: 𝑄𝑖 and 𝛿𝑖
 Load buses or PQ buses: Knowns: 𝑃𝑖 and 𝑄𝑖 , Unknowns: 𝑉𝑖 and 𝛿𝑖

Prof. Muwaffaq I. Alomoush, Yarmouk University, Power Engineering Department, First Semester 2023/2024 42
Types of Buses: Cont.

Bus Type Other Names Known Unknown

Reference bus Swing, Slack 𝑉𝑖 , 𝛿𝑖 = 0𝑜 𝑃𝑖 , 𝑄𝑖
Generator bus PV , Voltage-Controlled 𝑃𝑖 , 𝑉𝑖 𝑄𝑖 , 𝛿𝑖
Load bus PQ 𝑃𝑖 , 𝑄𝑖 𝑉𝑖 , 𝛿𝑖

Prof. Muwaffaq I. Alomoush, Yarmouk University, Power Engineering Department, First Semester 2023/2024 43
Example:
For this system we have to write the proper equations to find the unknown
voltage magnitudes and unknown angles. (In NR: Write equations for the
known net injected real power and known net injected reactive power at
buses).
The unknowns: 𝑉2 , 𝑉3 , 𝛿2 , 𝛿3 , 𝛿4 (5 variables). Therefore, we need to write 5
nonlinear algebraic equations using the known 𝑃 and 𝑄 at buses.

44
Prof. Muwaffaq I. Alomoush, Yarmouk University, Power Engineering Department, First Semester 2023/2024
Newton-Raphson (NR) Load Flow

The net injected real power at bus 𝑖 (𝑃𝑖 = 𝑃𝐺𝑖 − 𝑃𝐷𝑖 ) and the net injected reactive at bus 𝑖 (𝑄𝑖 = 𝑄𝐺𝑖 −
𝑄𝐷𝑖 ) in Polar Form are given by:
𝒏 𝒏

𝑃𝑖 = ෍ 𝑉𝑖 𝑉𝑗 𝑌𝑖𝑗 𝑐𝑜𝑠(𝛿𝑖 − 𝛿𝑗 −𝜃𝑖𝑗 ) or 𝑃𝑖 = ෍ |𝑽𝑖 |𝑽𝑗 |𝒀𝑖𝑗 | 𝑐𝑜𝑠(𝛿𝑖 − 𝛿𝑗 −𝜃𝑖𝑗 ) Set of none-linear
𝑗 =1 𝑗 =1
algebraic equations
𝒏 𝒏 used in NR load
flow
𝑄𝑖 = ෍ 𝑉𝑖 𝑉𝑗 𝑌𝑖𝑗 𝑠𝑖𝑛(𝛿𝑖 − 𝛿𝑗 −𝜃𝑖𝑗 ) or 𝑄𝑖 = ෍ |𝑽𝑖 |𝑽𝑗 |𝒀𝑖𝑗 | 𝑠𝑖𝑛(𝛿𝑖 − 𝛿𝑗 −𝜃𝑖𝑗 )
𝑗 =1 𝑗 =1

In Rectangular Form, we can find 𝑃𝑖 and 𝑄𝑖 as follows:

𝒏

𝑃𝑖 = ෍ 𝑉𝑖 𝑉𝑗 [𝐺𝑖𝑗 𝑐𝑜𝑠(𝛿𝑖 − 𝛿𝑗 ) + 𝐵𝑖𝑗 𝑠𝑖𝑛(𝛿𝑖 − 𝛿𝑗 )]

𝑗 =1
𝒏

𝑄𝑖 = ෍ 𝑉𝑖 𝑉𝑗 𝐺𝑖𝑗 𝑠𝑖𝑛(𝛿𝑖 − 𝛿𝑗 − 𝐵𝑖𝑗 𝑐𝑜𝑠(𝛿𝑖 − 𝛿𝑗 )]

𝑗 =1

Prof. Muwaffaq I. Alomoush, Yarmouk University, Power Engineering Department, First Semester 2023/2024 45
Recall: In NR method:
(𝑘 ) −1
𝜕𝑓1 𝜕𝑓1 𝜕𝑓1
⋯
(𝑘+1) (𝑘 ) 𝜕𝑥1 𝜕𝑥2 𝜕𝑥𝑛 (𝑘 )
𝑥1 𝑥1 𝑓1
𝜕𝑓2 𝜕𝑓2 𝜕𝑓2
𝑥2 𝑥 ⋯ 𝑓2
⋮ = ⋮2 − 𝜕𝑥1 𝜕𝑥2 𝜕𝑥𝑛 ⋮
𝑥𝑛 𝑥𝑛 ⋮ ⋮ ⋱ ⋮ 𝑓𝑛
𝜕𝑓𝑛 𝜕𝑓𝑛 𝜕𝑓𝑛
…
𝜕𝑥1 𝜕𝑥2 𝜕𝑥𝑛
Jacobian Matrix ( 𝐽 )
In compact form:
−1 (𝑘)
[𝒙](𝑘+1) = [𝒙](𝑘) − 𝑱(𝑘) 𝒇
or
−1 (𝑘)
[𝒙](𝑘+1) = [𝒙](𝑘) − [∆𝒙](𝑘) ; [∆𝒙](𝑘) = 𝑱(𝑘) 𝒇
Start by an initial guess for each of 𝑥1 0 , 𝑥2 0 , … , 𝑥𝑛 0 then iterate and stop when all
𝑥𝑖 𝑘+1 − 𝑥𝑖 𝑘 ≤∈ ; 𝑖 = 1, 2, … , 𝑛

Prof. Muwaffaq I. Alomoush, Yarmouk University, Power Engineering Department, First Semester 2023/2024 46
Recall:

In power flow analysis:

Known:
 Real power injection at each bus except the reference bus
 Reactive power injection at each load bus
 Voltage magnitude at each generator bus: |𝑽𝑖 | or 𝑉𝑖
Unknown (to be found by solving the nonlinear algebraic equations)
 Voltage magnitude at all load buses.
 Phase angle at each bus except the reference bus.
Unknowns: 𝛿2 , 𝛿3 , 𝛿4 , 𝛿5 ,…, 𝛿𝑛 (assuming bus 1 is the reference bus)

Bus Type Known Unknown

Reference bus 𝑉𝑖 , 𝛿𝑖 = 0𝑜 𝑃𝑖 , 𝑄𝑖
Generator bus 𝑃𝑖 , 𝑉𝑖 𝑄𝑖 , 𝛿𝑖
Load bus 𝑃𝑖 , 𝑄𝑖 𝑉𝑖 , 𝛿𝑖
Prof. Muwaffaq I. Alomoush, Yarmouk University, Power Engineering Department, First Semester 2023/2024 47
In NR load flow, we will find unknown angles and unknown voltage magnitudes using
the equations of the known real power injections (𝑃𝑖 ) and the known reactive power
injections (𝑄𝑖 ).
Summary:
Use:
 Equations of known net injected real power at buses (at load buses and generator buses except
slack bus)
 Equations of known net injected reactive power at buses (at load buses)

To find:
 Unknown angles (at all buses except the slack bus)
 Unknown voltage magnitudes (at all load buses)

Example:

Known Unknown
(𝑃𝑖 and 𝑄𝑖 ) (𝛿𝑖 and 𝑉𝑖 )
𝑃2 𝛿2
𝑄2 𝑉2

We need two equations: Use equations of 𝑃2 and 𝑄2 to find 𝛿2 and 𝑉2

48
Example:

Known Unknown
(𝑃𝑖 and 𝑄𝑖 ) (𝛿𝑖 and 𝑉𝑖 )
𝑃2 𝛿2

We need one equation: Use equations of 𝑃2 to

find 𝛿2

Example

Known Unknown
(𝑃𝑖 and 𝑄𝑖 ) (𝛿𝑖 and 𝑉𝑖 )
𝑃2 𝛿2
𝑃3 𝛿3
𝑄2 𝑉2

49
We need three equations: Use equations of 𝑃2 , 𝑃3 and 𝑄2 to find 𝛿2 , 𝛿3 and 𝑉2
Example

Known Unknown
(𝑃𝑖 and 𝑄𝑖 ) (𝛿𝑖 and 𝑉𝑖 )
𝑃2 𝛿2
𝑃3 𝛿3
𝑃4 𝛿4
𝑄2 𝑉2
We need four equations: Use equations of 𝑃2 , 𝑃3 , 𝑃4 and 𝑄2 to find 𝛿2 , 𝛿3 , 𝛿3 and 𝑉2.
50
General Solution of NRLF: Introduction to Decoupled
Load Flow

We can express NR load flow (NRLF) solution by any of the following general iterative compact matrix
equations:

𝜕𝑃 𝜕𝑃 (𝑘) −1
(𝑘+1) (𝑘) (𝑘)
[𝛿 ] [𝛿 ] 𝜕𝛿 𝜕𝑉 [∆𝑃]
= +
[𝑉 ] [𝑉 ] 𝜕𝑄 𝜕𝑄 [∆𝑄 ]
𝜕𝛿 𝜕𝑉

𝜕𝑃 𝜕𝑃 (𝑘) −1
(𝑘) (𝑘) (𝑘) (𝑘+1) (𝑘)
[∆𝛿 ] 𝜕𝛿 𝜕𝑉 [∆𝑃] [∆𝛿 ] [𝛿 ] [𝛿 ]
Or = where, = =
[∆𝑉 ] 𝜕𝑄 𝜕𝑄 [∆𝑄 ] [∆𝑉 ] [𝑉 ] [𝑉 ]
𝜕𝛿 𝜕𝑉

(𝑘) (𝑘) −1 (𝑘) (𝑘) (𝑘+1) (𝑘)

[∆𝛿 ] [ 𝐽11 ] [𝐽12 ] [∆𝑃] [∆𝛿 ] [𝛿 ] [𝛿 ]
Or = ; = =
[∆𝑉 ] [ 𝐽21 ] [ 𝐽22 ] [∆𝑄 ] [∆𝑉 ] [𝑉 ] [𝑉 ]

(𝑘) (𝑘) (𝑘)

[∆𝑃] [ 𝐽11 ] [𝐽12 ] [∆𝛿 ]
Or =
[∆𝑄 ] [ 𝐽21 ] [ 𝐽22 ] [∆𝑉 ]

Prof. Muwaffaq I. Alomoush, Yarmouk University, Power Engineering Department, First Semester 2023/2024 51
 The last compact matrix equation can be expanded (detailed) to be in the following form:

(𝑘) (𝑘)
∆𝑃1 (𝑘) ∆𝛿1
𝜕𝑃1 𝜕𝑃1 𝜕𝑃1 𝜕𝑃1 𝜕𝑃1 𝜕𝑃1
⋯ ⋯
𝜕𝛿 1 𝜕𝛿 2 𝜕𝛿 2 𝜕𝑉1 𝜕𝑉2 𝜕𝑉2
Remove row ∆𝑃2 ∆𝛿2
𝜕𝑃2 𝜕𝑃2 ⋯ 𝜕𝑃2 𝜕𝑃2 𝜕𝑃2 ⋯ 𝜕𝑃2
and column of 𝜕𝛿 1 𝜕𝛿 2 𝜕𝛿 𝑛 𝜕𝑉1 𝜕𝑉2 𝜕𝑉𝑛
the reference ⋮ ⋮ ⋮ ⋮ ⋮ ⋮ ⋮ ⋮
𝜕𝑃𝑛 𝜕𝑃𝑛 𝜕𝑃𝑛 𝜕𝑃𝑛 𝜕𝑃𝑛 𝜕𝑃𝑛
bus ⋯ ⋯
∆𝑃2 𝜕𝛿 1 𝜕𝛿 2 𝜕𝛿 𝑛 𝜕𝑉1 𝜕𝑉2 𝜕𝑉𝑛 ∆𝛿2
=
∆𝑄1 𝜕𝑄1 𝜕𝑄1
⋯
𝜕𝑄1 𝜕𝑄1 𝜕𝑄1
⋯
𝜕𝑄1 ∆𝑉1
Remove rows 𝜕𝛿 1 𝜕𝛿 2 𝜕𝛿 2 𝜕𝑉1 𝜕𝑉2 𝜕𝑉2
and columns of ∆𝑄2 𝜕𝑄2 𝜕𝑄2 ⋯ 𝜕𝑄2 𝜕𝑄2 𝜕𝑄2 ⋯ 𝜕𝑄2
∆𝑉2
𝜕𝛿 1 𝜕𝛿 2 𝜕𝛿 𝑛 𝜕𝑉1 𝜕 𝑉2 𝜕𝑉𝑛
the reference
⋮ ⋮ ⋮ ⋮ ⋮ ⋮ ⋮ ⋮
bus and the 𝜕𝑄𝑛 𝜕𝑄𝑛 𝜕𝑄𝑛 𝜕𝑄𝑛 𝜕𝑄𝑛 𝜕𝑄𝑛
generator buses 𝜕𝛿 1 𝜕𝛿 2 ⋯ 𝜕𝛿 𝑛 𝜕𝑉1 𝜕𝑉2 ⋯ 𝜕𝑉𝑛
∆𝑄2 ∆𝑉2

Prof. Muwaffaq I. Alomoush, Yarmouk University, Power Engineering Department, First Semester 2023/2024 52
The NRLF solution can be written by the following coupled load flow
equations.

∆𝑷 𝒌 = 𝑱𝟏𝟏 𝒌 ∆𝜹 𝒌 + 𝑱𝟏𝟐 𝒌 ∆𝑽 𝒌

∆𝑸 (𝒌) = 𝑱𝟐𝟏 𝒌 ∆𝜹 (𝒌) + 𝑱𝟐𝟐 (𝒌) ∆𝑽 (𝒌)

It was found that the elements (values) of the off-diagonal sub-

matrices ( 𝐽12 and 𝐽21 ) are very small compared to the elements of
the diagonal sub-matrices ( 𝐽11 and 𝐽22 ) for the following reasons (
for real (typical) power systems operating on normal operating
conditions):
𝑥𝑖𝑗
(1) 𝑟𝑖𝑗 ≪ 𝑥𝑖𝑗 ( 𝑢𝑠𝑢𝑎𝑙𝑙𝑦: = 5 , 7 or more )
𝑟𝑖𝑗
(2) The difference between any two angles 𝛿𝑖 − 𝛿𝑗 is very small
(3) The power system has nearly a flat profile of voltages, i.e.,
𝑉𝑖 ≈ 𝑉𝑗 ≈ 1 𝑝. 𝑢.
Prof. Muwaffaq I. Alomoush, Yarmouk University, Power Engineering Department, First Semester 2023/2024 53
Based on the above observations, we can describe the
following NR coupled equations as follows

∆𝑷 𝒌 = 𝑱𝟏𝟏 𝒌 ∆𝜹 𝒌 + 𝑱𝟏𝟐 𝒌 ∆𝑽 𝒌 (A)

∆𝑸 (𝒌) = 𝑱𝟐𝟏 𝒌 ∆𝜹 (𝒌) + 𝑱𝟐𝟐 (𝒌) ∆𝑽 (𝒌) (B)

Equation (A): Real power changes in a power system is

affected by both changes in angles (frequency) and changes in
voltage magnitudes. Real power changes are mainly
affected by changes in angles (frequency) and weakly
affected by changes in voltage magnitudes.

Equation (B): Reactive power changes in a power system is

affected by both changes in angles (frequency) and changes in
voltage magnitudes. Reactive power changes are mainly
affected by changes in voltage magnitudes and weakly
affected by changes in angles (frequency).
Prof. Muwaffaq I. Alomoush, Yarmouk University, Power Engineering Department, First Semester 2023/2024 54
 Decouples Load Flow (DLF)
If the off-diagonal sub-matrices are neglected ( 𝐽12 = [0] and 𝐽21 = [0]),
(𝑘) (𝑘)
(𝑘)
∆𝑃 𝐽11 0 ∆𝛿 Decoupled Load Flow
then: =
∆𝑄 0 𝐽22 ∆𝑉 (DLF) equations

We will obtain the following two sets of Decoupled Load Flow (DLF) equations:
∆𝑷 𝒌 = 𝑱𝟏𝟏 𝒌 ∆𝜹 𝒌
∆𝑸 (𝒌) = 𝑱𝟐𝟐 (𝒌) ∆𝑽 (𝒌)
or ∆𝛿 𝑘 = 𝐽11 𝑘 −1
∆𝑃 𝑘

∆𝑉 (𝑘) = 𝐽 (𝑘) −1 ∆𝑄 (𝑘)

22
or 𝛿 𝑘+1
= 𝛿 + 𝐽11 𝑘 𝑘 −1 ∆𝑃 𝑘
−1
𝑉 (𝑘+1) = 𝑉 (𝑘) + 𝐽22 (𝑘) ∆𝑄 (𝑘)
In summary:
𝜕𝑃𝑖 𝜕𝑄𝑖
1. Calculations of and ; 𝑖, 𝑗 = 1,2, … , 𝑛 is the same for both NRLF and
𝜕𝛿𝑗 𝜕𝑉𝑗
DLF.
2. Calculations of ∆𝑃𝑖 and ∆𝑄𝑖 ; 𝑖 = 1,2, … , 𝑛 is the same for both NRLF and
55
DLF. Prof. Muwaffaq I. Alomoush, Yarmouk University, Power Engineering Department, First Semester 2023/2024
Decouples Load Flow (DLF): Cont.

DLF is the same as NRL (same equations and same procedure) with
the following exceptions:
 Make 𝐽12 = [0] and 𝐽21 = [0] in NRLF to obtain DLF, i.e., in
𝜕𝑃 𝜕𝑄
DLF we make 𝑖 = 0, 𝑖 = 0 ; 𝑖, 𝑗 = 1, 2, … , 𝑛.
𝜕𝑉𝑗 𝜕𝛿𝑗

 In NRLF you solve the whole system of equations, but in DLF the
whole system becomes two sets of equations.
 From previous discussion, DLF is an approximation of NRLF,
where we make 𝐽12 = [0] and 𝐽21 = [0] in NRLF to obtain DLF.

Prof. Muwaffaq I. Alomoush, Yarmouk University, Power Engineering Department, First Semester 2023/2024 56
The DLF is expressed by:
(𝑘) −1
𝜕𝑃1 𝜕𝑃1 𝜕𝑃1

Prof. Muwaffaq I. Alomoush, Yarmouk University, Power Engineering Department, First Semester 2023/2024
(𝑘+1) (𝑘) ⋯ (𝑘)
∆𝛿1 ∆𝛿1 𝜕𝛿1 𝜕𝛿2 𝜕𝛿𝑛 ∆𝑃1
0 0 ⋯ 0
𝜕𝑃2 𝜕𝑃2 𝜕𝑃2 ⋯
∆𝛿2 ∆𝛿2 ⋯ 0 0 0 ∆𝑃2
𝜕𝛿1 𝜕𝛿2 𝜕𝛿𝑛 ⋮ ⋮ ⋮
⋮ ⋮ ⋮ ⋮ ⋮ 0 0 0 ⋮
𝜕𝑃𝑛 𝜕𝑃𝑛 𝜕𝑃𝑛 ⋯
⋯
∆𝛿𝑛 ∆𝛿𝑛 𝜕𝛿1 𝜕𝛿2 𝜕𝛿𝑛 ∆𝑃𝑛
= +
∆𝑉1 ∆𝑉1 𝜕𝑄1 𝜕𝑄1 𝜕𝑄1 ∆𝑄1
⋯
0 0 ⋯ 0 𝜕𝑉1 𝜕𝑉2 𝜕𝑉2
∆𝑉2 ∆𝑉2 ⋯ 𝜕𝑄2 𝜕𝑄2 𝜕𝑄2 ∆𝑄2
0 0 0 ⋯
⋮ ⋮ ⋮ ⋮ ⋮ 𝜕𝑉1 𝜕𝑉2 𝜕𝑉𝑛 ⋮
0 0 0 ⋮ ⋮ ⋮
⋯ 𝜕𝑄𝑛 𝜕𝑄𝑛 𝜕𝑄𝑛
∆𝑉𝑛 ∆𝑉𝑛 ⋯ ∆𝑄𝑛
𝜕𝑉1 𝜕𝑉2 𝜕𝑉𝑛

The above one set of equations can be written as two separate decoupled equations:
(𝑘) −1
∆𝛿1 (𝑘+1)
∆𝛿1 (𝑘) 𝜕𝑃1 𝜕𝑃1 𝜕𝑃1 ∆𝑃1 (𝑘)
⋯
𝜕𝛿1 𝜕𝛿2 𝜕𝛿𝑛
∆𝛿2 ∆𝛿2 𝜕𝑃2 𝜕𝑃2 𝜕𝑃2 ∆𝑃2
⋯
= + 𝜕𝛿1 𝜕𝛿2 𝜕𝛿𝑛
⋮ ⋮ ⋮ ⋮ ⋮ ⋮
𝜕𝑃𝑛 𝜕𝑃𝑛 𝜕𝑃𝑛
∆𝛿𝑛 ∆𝛿𝑛 ⋯ ∆𝑃𝑛
𝜕𝛿1 𝜕𝛿2 𝜕𝛿𝑛

(𝑘) −1
∆𝑉1 (𝑘+1)
∆𝑉1 (𝑘) 𝜕𝑄1 𝜕𝑄1 𝜕𝑄1 ∆𝑄1 (𝑘)
⋯
𝜕𝑉1 𝜕𝑉2 𝜕𝑉𝑛
∆𝑉2 ∆𝑉2 𝜕𝑄2 𝜕𝑄2 𝜕𝑄2 ∆𝑄2
⋯
= + 𝜕𝑉1 𝜕𝑉2 𝜕𝑉𝑛
⋮ ⋮ ⋮ ⋮ ⋮ ⋮
𝜕𝑄𝑛 𝜕𝑄𝑛
⋯
𝜕𝑄𝑛 57
∆𝑉𝑛 ∆𝑉𝑛 ∆𝑄𝑛
𝜕𝑉1 𝜕𝑉2 𝜕𝑉𝑛
Example
For the four bus system shown, write the iterative matrix equations of NRLF and DLF.

Known Unknown
(𝑃𝑖 and 𝑄𝑖 ) (𝛿𝑖 and 𝑉𝑖 )
𝑃1 𝛿1
𝑃2 𝛿2
𝑃3 𝛿3
𝑃4 𝛿4
𝑄1 𝑉1
𝑄3 𝑉3

We need five equations: Use equations of 𝑃1 , 𝑃2 , 𝑃3 , 𝑃4 , 𝑄1 and 𝑄2 to find 𝛿1 , 𝛿2 , 𝛿3 , 𝛿4 , 𝑉1 and

𝑉3 . 58
Prof. Muwaffaq I. Alomoush, Yarmouk University, Power Engineering Department, First Semester 2023/2024
NRLF
(𝑘+1) (𝑘) (𝑘) −𝟏 (𝑘)
𝛿1 𝛿1 𝜕𝑃1 𝜕𝑃1 𝜕𝑃1 𝜕𝑃1 𝜕𝑃1 ∆𝑃1

Prof. Muwaffaq I. Alomoush, Yarmouk University, Power Engineering Department, First Semester 2023/2024
𝜕𝛿1 𝜕𝛿2 𝜕𝛿3 𝜕𝑉1 𝜕𝑉3
𝜕𝑃2 𝜕𝑃2 𝜕𝑃2 𝜕𝑃2 𝜕𝑃2
𝛿2 𝛿2 𝜕𝛿1 𝜕𝛿2 𝜕𝛿3 𝜕𝑉1 𝜕𝑉3 ∆𝑃2
𝜕𝑃3 𝜕𝑃3 𝜕𝑃3 𝜕𝑃3 𝜕𝑃3
= +
𝛿3 𝛿3 𝜕𝛿1 𝜕𝛿2 𝜕𝛿3 𝜕𝑉1 𝜕𝑉3 ∆𝑃3
𝑉1 𝑉1 𝜕𝑄1 𝜕𝑄1 𝜕𝑄1 𝜕𝑄1 𝜕𝑄1 ∆𝑄1
𝜕𝛿1 𝜕𝛿2 𝜕𝛿3 𝜕𝑉1 𝜕𝑉3
𝜕𝑄3 𝜕𝑄3 𝜕𝑄3 𝜕𝑄3 𝜕𝑄3
𝑉3 𝑉3 𝜕𝛿1 𝜕𝛿2 𝜕𝛿3 𝜕𝑉1 𝜕𝑉3 ∆𝑄3
DLF:
(𝑘+1) (𝑘) (𝑘) −𝟏 (𝑘)
𝛿1 𝛿1 𝜕𝑃1 𝜕𝑃1𝜕𝑃1 ∆𝑃1
𝜕𝛿1 𝜕𝛿2𝜕𝛿3
𝜕𝑃2 𝜕𝑃2𝜕𝑃2 0 0
𝛿2 𝛿2 0 0 ∆𝑃2
𝜕𝛿1 𝜕𝛿2𝜕𝛿3
0 0
𝜕𝑃3 𝜕𝑃3𝜕𝑃3
= +
𝛿3 𝛿3 𝜕𝛿1 𝜕𝛿2𝜕𝛿3 ∆𝑃3
𝑉1 𝑉1 𝜕𝑄1 𝜕𝑄1 ∆𝑄1
0 0 0 𝜕𝑉1 𝜕𝑉3
0 0 0 𝜕𝑄3 𝜕𝑄3
𝑉3 𝑉3 𝜕𝑉1 𝜕𝑉3 ∆𝑄3

or
(𝑘+1) (𝑘) (𝑘)
𝛿1 𝛿1 −1 ∆𝑃1
𝜕𝑃1 𝜕𝑃1 𝜕𝑃1 (𝑘)
(𝑘+1) (𝑘) −𝟏 (𝑘)
𝜕𝛿 1 𝜕𝛿 2 𝜕𝛿 3 𝑉1 𝑉1 𝜕𝑄1 𝜕𝑄1 (𝑘) ∆𝑄1
𝜕𝑃2 𝜕𝑃2 𝜕𝑃2 𝜕𝑉1 𝜕𝑉3
𝛿2 = 𝛿2 + 𝜕𝛿 1 𝜕𝛿 2 𝜕𝛿 3
∆𝑃2 , = + 𝜕𝑄3 𝜕𝑄3
𝜕𝑃3 𝜕𝑃3 𝜕𝑃3 𝑉3 𝑉3 𝜕𝑉1 𝜕𝑉3 ∆𝑄3
𝜕𝛿 1 𝜕𝛿 2 𝜕𝛿 3 59
𝛿3 𝛿3 ∆𝑃3
Important Notes:
 NRLF and DLF: Both methods are iterative
 NRLF requires larger storage requirements in computer than DLF.
 NRLF takes less number of iterations than DLF but time per iteration in NRLF is
longer than DLF. )Example: For a certain system, if NR takes 5 iterations in 3
seconds, DLF may take 12 iterations in 0.8 seconds.)
 Both NRLF and DLF require inversion of matrices in each iteration, but the sizes
of matrices are larger in NRLF than matrices in DLF.

Recall:
 GSLF takes a number of iterations that depends on the size of the power system.
 NRLF takes a number of iterations that does not depend on the size of the power
system.
 NRLF usually converges to a solution but GSLF may fail to converge (rare in
power systems).
 GSLF is easier to program in computer than NRLF.
 NRLF requires larger storage requirements in computer
 NRLF and its approximations are more common in real power system
applications.

Prof. Muwaffaq I. Alomoush, Yarmouk University, Power Engineering Department, First Semester 2023/2024 60
Fast Decouples Load Flow (FDLF)

−1 −1
∆𝛿1
(𝑘) 𝐵11 𝐵12 ⋯ 𝐵1𝑛 ∆𝑃1 /𝑉1 (𝑘)
𝛿1
(𝑘+1)
∆𝛿1
(𝑘) 𝐵11 𝐵12 ⋯ 𝐵1𝑛 ∆𝑃1 /𝑉1 (𝑘)
⋯ ⋯
∆𝛿2 𝐵 𝐵22 𝐵2𝑛 ∆𝑃2 /𝑉2 𝛿2 ∆𝛿2 𝐵 𝐵22 𝐵2𝑛 ∆𝑃2 /𝑉2
= − 21 or = − 21
⋮ ⋮ ⋮ ⋮ ⋮ ⋮ ⋮ ⋮ ⋮ ⋮ ⋮
∆𝛿𝑛 𝐵𝑛1 𝐵𝑛2 ⋯ 𝐵𝑛𝑛 ∆𝑃𝑛 /𝑉𝑛 𝛿𝑛 ∆𝛿𝑛 𝐵𝑛1 𝐵𝑛2 ⋯ 𝐵𝑛𝑛 ∆𝑃𝑛 /𝑉𝑛

Suceptance Matrix. Recall [𝒀] = [𝑮] + 𝑗 [ 𝑩]

−1 −1
∆𝑉1 (𝑘) 𝐵11 𝐵12 ⋯ 𝐵1𝑛 ∆𝑄1 /𝑉1 (𝑘)
𝑉1 (𝑘+1)
𝑉1 (𝑘) 𝐵11 𝐵12 ⋯ 𝐵1𝑛 ∆𝑄1 /𝑉1 (𝑘)
⋯ ⋯
∆𝑉2 𝐵 𝐵22 𝐵2𝑛 ∆𝑄2 /𝑉2 𝑉2 𝑉 𝐵 𝐵22 𝐵2𝑛 ∆𝑄2 /𝑉2
= − 21 or = 2 − 21
⋮ ⋮ ⋮ ⋮ ⋮ ⋮ ⋮ ⋮ ⋮ ⋮ ⋮
∆𝑉𝑛 𝐵𝑛1 𝐵𝑛2 𝑉𝑛 𝑉𝑛
⋯ 𝐵𝑛𝑛 ∆𝑄𝑛 /𝑉𝑛 𝐵𝑛1 𝐵𝑛2 ⋯ 𝐵𝑛𝑛 ∆𝑄𝑛 /𝑉𝑛

[∆𝛿 ](𝑘) = − [𝐵]−1 [∆𝑃/𝑉 ](𝑘) or [𝛿 ](𝑘+1) = [𝛿 ](𝑘) − [𝐵]−1 [∆𝑃/𝑉 ](𝑘) (1)

[∆𝑉 ](𝑘) = − [𝐵]−1 [∆𝑄/𝑉 ](𝑘) or [𝑉 ](𝑘+1) = [𝑉 ](𝑘) − [𝐵]−1 [∆𝑄/𝑉 ](𝑘) (2)

Prof. Muwaffaq I. Alomoush, Yarmouk University, Power Engineering Department, First Semester 2023/2024 61
The angle of the reference bus is known, therefore the column and the row of the reference bus are
removed from equation (1).

Also, the voltage magnitude of the reference bus and the generator buses are known, therefore the
columns and rows of the reference bus and generator buses are removed from equation (2).

After the removal of the buses, the above two equations (1) and (2) become:

[∆𝛿 ](𝑘) = − [ 𝐵′ ]−1 [∆𝑃/𝑉 ](𝑘) or [𝛿 ](𝑘+1) = [𝛿 ](𝑘) − [ 𝐵′ ]−1 [∆𝑃/𝑉 ](𝑘) (1)

[∆𝑉 ](𝑘) = − [𝐵′′]−1 [∆𝑄/𝑉 ](𝑘) or [𝑉 ](𝑘+1) = [𝑉 ](𝑘) − [𝐵′′]−1 [∆𝑄/𝑉 ](𝑘) (2)

[ 𝐵 ′ ] = [ 𝐵 ] − 𝑟𝑜𝑤 𝑎𝑛𝑑 𝑐𝑜𝑙𝑢𝑚 𝑜𝑓 𝑡ℎ𝑒 𝑟𝑒𝑓𝑟𝑒𝑛𝑐𝑒 𝑏𝑢𝑠

[ 𝐵 ′′ ] = [ 𝐵 ] − 𝑟𝑜𝑤 𝑎𝑛𝑑 𝑐𝑜𝑙𝑢𝑚 𝑜𝑓 𝑡ℎ𝑒 𝑟𝑒𝑓𝑟𝑒𝑛𝑐𝑒 𝑏𝑢𝑠 − 𝑟𝑜𝑤𝑠 𝑎𝑛𝑑 𝑐𝑜𝑙𝑢𝑚𝑠 𝑜𝑓 𝑔𝑒𝑛𝑒𝑟𝑎𝑡𝑜𝑟 𝑏𝑢𝑠𝑒𝑠

Prof. Muwaffaq I. Alomoush, Yarmouk University, Power Engineering Department, First Semester 2023/2024 62
Example:
For the three-bus system shown, if all values are given in p.u.,

Prof. Muwaffaq I. Alomoush, Yarmouk University, Power Engineering Department, First Semester 2023/2024
(1) Find [𝑩’] and [𝑩’’] used in FDLD if bus 1 is the reference bus and write the
expressions of the iterative solution of the FDLF.
(2) Resolve (1) if bus 3 is the reference bus.

9 − 𝑗13 −6 + 𝑗8 −3 + 𝑗4 9 −6 −3 −13 8 4 −13 8 4

𝒀 = −6 + 𝑗8 6 − 𝑗18 0 + 𝑗10 = −6 6 0 +𝑗 8 −18 10 ; 𝑩 = 8 −18 10
−3 + 𝑗4 0 + 𝑗10 3 − 𝑗16 −3 0 3 4 10 −16 4 10 −16 63
 −13 8 4
𝑩= 8 −18 10
4 10 −16
(1) Bus 1 is the reference bus, and bus 3 is a generator bus
−18 10
𝑩′ = , 𝑩′′ = [−18]
10 −16
(𝑘 ) −1 (𝑘 ) (𝑘+1 ) (𝑘 ) −1 (𝑘 )
∆𝛿2 −18 10 ∆𝑃2 /𝑉2 𝛿2 𝛿 −18 10 ∆𝑃2 /𝑉2
=− or = 2 −
∆𝛿3 10 −16 ∆𝑃3 /𝑉3 𝛿3 𝛿3 10 −16 ∆𝑃3 /𝑉3

[∆𝑉2 ](𝑘 ) = − [−18] −1 [∆𝑄2 /𝑉2 ](𝑘 ) or [𝑉2 ](𝑘+1) = [𝑉2 ](𝑘 ) − [−18]
−1
[∆𝑄2 /𝑉2 ](𝑘 )

(2) Bus 3 is the reference bus, and bus 1 is a generator bus

−13 8
𝑩′ = , 𝑩′′ = [−18]
8 −18
(𝑘 ) −1 (𝑘 ) (𝑘+1 ) (𝑘 ) −1 (𝑘 )
∆𝛿1 −13 8 ∆𝑃1 /𝑉1 𝛿2 𝛿 −13 8 ∆𝑃1 /𝑉1
=− or = 2 −
∆𝛿2 8 −18 ∆𝑃2 /𝑉2 𝛿3 𝛿3 8 −18 ∆𝑃2 /𝑉2

[∆𝑉2 ](𝑘 ) = − [−18] −1 [∆𝑄2 /𝑉2 ](𝑘 ) or [𝑉2 ](𝑘+1) = [𝑉2 ](𝑘 ) − [−18]
−1
[∆𝑄2 /𝑉2 ](𝑘 )

Prof. Muwaffaq I. Alomoush, Yarmouk University, Power Engineering Department, First Semester 2023/2024 64
Important Notes:

 Jacobian matrices and their inverses in NRLF and DLF are not constants but must be recalculated
every iteration, but they are constants in FDLF.
 𝐼𝑡𝑒𝑟𝑎𝑡𝑖𝑜𝑛𝑠(𝑁𝑅𝐿𝐹) < 𝐼𝑡𝑒𝑟𝑎𝑡𝑖𝑜𝑛𝑠(𝐷𝐿𝐹) < 𝐼𝑡𝑒𝑟𝑎𝑡𝑖𝑜𝑛𝑠(𝐹𝐷𝐿𝐹)
 𝑇𝑖𝑚𝑒 𝑃𝑒𝑟 𝐼𝑡𝑒𝑟𝑎𝑡𝑖𝑜𝑛𝑠(𝑁𝑅𝐿𝐹) ≫ 𝑖𝑚𝑒 𝑃𝑒𝑟 𝐼𝑡𝑒𝑟𝑎𝑡𝑖𝑜𝑛𝑠(𝐷𝐿𝐹) ≫ 𝑖𝑚𝑒 𝑃𝑒𝑟 𝐼𝑡𝑒𝑟𝑎𝑡𝑖𝑜𝑛𝑠(𝐹𝐷𝐿𝐹)

Example: To solve load flow for a typical power system using the three methods:
NRLF DLF FDLF
# of iterations 6 15 25
Run time (sec) 12 1.5 0.5
Time per iterations 2 0.1 0.02
1
Computer storage Large Medium ≈ 𝑁𝑅𝐿𝐹 small
2
Jacobian Matrices 𝐽(𝑘) change with iteration (𝑘) (𝑘)
𝐽11 , 𝐽22 change with iteration −𝐵′ , −𝐵′′ constants
Inverses of Jacobian matrices −1 (𝑘) −1 (𝑘) −1 −𝐵′
−1
, −𝐵′′−1 constants
𝐽(𝑘) change with iteration 𝐽11 , 𝐽22 change with iteration

Prof. Muwaffaq I. Alomoush, Yarmouk University, Power Engineering Department, First Semester 2023/2024 65
DC Load Flow (DCLF) or DC Power Flow

DC load flow (DCLF) is concerned only with real power and phase angles.
It does not consider reactive power and changes in voltage magnitudes. In
DCLF, we assume that “ the power system is lossless (has no losses, i.e.,
𝑟𝑖𝑗 = 0;, and the system has a flat profile of voltages, i.e., 𝑉𝑖 = 1; 𝑖 =
1, 2, … , 𝑛 ”.
Assumptions in DCLFF:
 𝑟𝑖𝑗 = 0 : neglect resistances in transmission lines.
 𝑉𝑖 = 𝑉𝑗 = 1 𝑝. 𝑢. : the system has a flat profile of voltages
 (𝛿𝑖 −𝛿𝑗 ) is very small: differences in phase angles are very small for a
system operating in normal conditions, therefore 𝑠𝑖𝑛(𝛿𝑖 −𝛿𝑗 ) ≅
(𝛿𝑖 −𝛿𝑗 ) (angles in rad).

Prof. Muwaffaq I. Alomoush, Yarmouk University, Power Engineering Department, First Semester 2023/2024 66
Derivation of DCLF equations

Prof. Muwaffaq I. Alomoush, Yarmouk University, Power Engineering Department, First Semester 2023/2024
The net injected real power at bus 𝑖 (𝑃𝑖 = 𝑃𝐺𝑖 − 𝑃𝐷𝑖 ) at bus 𝑖 in Polar
and Rectangular Forms are given by:
𝒏

𝑃𝑖 = ෍ 𝑉𝑖 𝑉𝑗 𝑌𝑖𝑗 𝑐𝑜𝑠(𝛿𝑖 −𝛿𝑗 −𝜃𝑖𝑗 )

𝑗=1
𝒏

𝑃𝑖 = ෍ 𝑉𝑖 𝑉𝑗 [𝐺𝑖𝑗 𝑐𝑜𝑠(𝛿𝑖 −𝛿𝑗 ) + 𝐵𝑖𝑗 𝑠𝑖𝑛(𝛿𝑖 −𝛿𝑗 )]

𝑗=1
Recall:
1 1 𝑟𝑖𝑗
 𝒚𝑖𝑗 = = = = 𝑔𝑖𝑗 + 𝑗 𝑏𝑖𝑗 ; 𝑔𝑖𝑗 = 2 +𝑥 2 and 𝑏𝑖𝑗 =
𝒛𝑖𝑗 𝑟𝑖𝑗 +𝑗𝑥𝑖𝑗 𝑟𝑖𝑗 𝑖𝑗
𝑥𝑖𝑗
− 2 +𝑥 2
𝑟𝑖𝑗 𝑖𝑗
1 1
 If 𝑟𝑖𝑗 is neglected, then: 𝑔𝑖𝑗 = 0 , 𝑏𝑖𝑗 = − , 𝒚𝑖𝑗 = −𝑗
𝑥𝑖𝑗 𝑥𝑖𝑗

If resistances are neglected (𝑟𝑖𝑗 = 0), then 𝑔𝑖𝑗 = 0 and 𝐺𝑖𝑗 = 0.

Therefore, 𝑮 = [0]. Also, for 𝑟𝑖𝑗 = 0, 𝜃𝑖𝑖 = −90𝑜 and 𝜃𝑖𝑗 = +90𝑜 ; 𝑖 ≠ 𝑗.
cos (𝜃𝑖𝑖 ) = cos −90𝑜 = 0 , sin 𝜃𝑖𝑗 = sin 90𝑜 = 1 ; 𝑖 ≠ 𝑗. 67
For 𝑉𝑖 = 𝑉𝑗 = 1 𝑝. 𝑢., and based on the above assumptions:
𝒏

𝑃𝑖 = ෍ 𝑉𝑖 𝑉𝑗 [𝐺𝑖𝑗 𝑐𝑜𝑠(𝛿𝑖 −𝛿𝑗 ) + 𝐵𝑖𝑗 𝑠𝑖𝑛(𝛿𝑖 −𝛿𝑗 )]

𝑗=1
𝒏 𝒏

= ෍(1) (1) [ 0 𝑐𝑜𝑠(𝛿𝑖 −𝛿𝑗 ) + 𝐵𝑖𝑗 𝑠𝑖𝑛(𝛿𝑖 −𝛿𝑗 )] = ෍ 𝐵𝑖𝑗 𝑠𝑖𝑛(𝛿𝑖 −𝛿𝑗 )
𝑗=1 𝑗=1

If we expand the last equation:

𝑃1 = 𝐵11 𝑠𝑖𝑛(𝛿1 −𝛿1 ) + 𝐵12 𝑠𝑖𝑛(𝛿1 −𝛿2 ) + 𝐵13 𝑠𝑖𝑛(𝛿1 −𝛿3 ) + ⋯ + 𝐵1𝑛 𝑠𝑖𝑛(𝛿1 −𝛿𝑛 )
𝑃2 = 𝐵21 𝑠𝑖𝑛(𝛿2 −𝛿1 ) + 𝐵22 𝑠𝑖𝑛(𝛿2 −𝛿2 ) + 𝐵23 𝑠𝑖𝑛(𝛿2 −𝛿3 ) + ⋯ + 𝐵2𝑛 𝑠𝑖𝑛(𝛿2 −𝛿𝑛 )
𝑃3 = 𝐵31 𝑠𝑖𝑛(𝛿3 −𝛿1 ) + 𝐵32 𝑠𝑖𝑛(𝛿3 −𝛿2 ) + 𝐵33 𝑠𝑖𝑛(𝛿3 −𝛿3 ) + ⋯ + 𝐵3𝑛 𝑠𝑖𝑛(𝛿3 −𝛿𝑛 )
⋮
𝑃𝑛 = 𝐵𝑛1 𝑠𝑖𝑛(𝛿𝑛 −𝛿1 ) + 𝐵𝑛2 𝑠𝑖𝑛(𝛿𝑛 −𝛿2 ) + 𝐵𝑛3 𝑠𝑖𝑛(𝛿𝑛 −𝛿3 ) + ⋯ + 𝐵𝑛𝑛 𝑠𝑖𝑛(𝛿𝑛 −𝛿𝑛 )

If (𝛿𝑖 −𝛿𝑗 ) is very small, then 𝑠𝑖𝑛(𝛿𝑖 −𝛿𝑗 ) ≅ (𝛿𝑖 −𝛿𝑗 )

Prof. Muwaffaq I. Alomoush, Yarmouk University, Power Engineering Department, First Semester 2023/2024 68
𝑃1 = 𝐵11 (𝛿1 −𝛿1 ) + 𝐵12 (𝛿1 −𝛿2 ) + 𝐵13 (𝛿1 −𝛿3 ) + ⋯ + 𝐵1𝑛 (𝛿1 −𝛿𝑛 )

Prof. Muwaffaq I. Alomoush, Yarmouk University, Power Engineering Department, First Semester 2023/2024
𝑃2 = 𝐵21 (𝛿2 −𝛿1 ) + 𝐵22 (𝛿2 −𝛿2 ) + 𝐵23 (𝛿2 −𝛿3 ) + ⋯ + 𝐵2𝑛 (𝛿2 −𝛿𝑛 )
𝑃3 = 𝐵31 (𝛿3 −𝛿1 ) + 𝐵32 (𝛿3 −𝛿2 ) + 𝐵33 (𝛿3 −𝛿3 ) + ⋯ + 𝐵3𝑛 (𝛿3 −𝛿𝑛 )
⋮
𝑃𝑛 = 𝐵𝑛1 (𝛿𝑛 −𝛿1 ) + 𝐵𝑛2 (𝛿𝑛 −𝛿2 ) + 𝐵𝑛3 (𝛿𝑛 −𝛿3 ) + ⋯ + 𝐵𝑛𝑛 (𝛿𝑛 −𝛿𝑛 )
1
But 𝐵𝑖𝑗 = ; 𝑖≠𝑗
𝑥𝑖𝑗

(𝛿1 −𝛿2 ) (𝛿1 −𝛿3 ) (𝛿1 −𝛿𝑛 )

𝑃1 = + +⋯+
𝑥12 𝑥13 𝑥1𝑛
(𝛿2 −𝛿1 ) (𝛿2 −𝛿3 ) (𝛿2 −𝛿𝑛 )
𝑃2 = + +⋯+
𝑥21 𝑥23 𝑥2𝑛
(𝛿3 −𝛿1 ) (𝛿3 −𝛿2 ) (𝛿3 −𝛿𝑛 )
𝑃3 = + +⋯+
𝑥31 𝑥32 𝑥3𝑛
⋮
(𝛿𝑛 −𝛿1 ) (𝛿𝑛 −𝛿2 ) (𝛿𝑛 −𝛿𝑛−1 )
𝑃𝑛 = + + ⋯+
𝑥𝑛1 𝑥𝑛2 𝑥𝑛 𝑛−1 69
 1 1 1 1 1 1
𝑃1 = + + ⋯+ 𝛿1 − 𝛿2 − 𝛿3 − ⋯ − 𝛿𝑛
𝑥12 𝑥13 𝑥1𝑛 𝑥12 𝑥12 𝑥1𝑛
1 1 1 1 1 1
𝑃2 = − 𝛿1 + + + ⋯+ 𝛿2 − 𝛿3 − ⋯ − 𝛿𝑛
𝑥21 𝑥21 𝑥23 𝑥2𝑛 𝑥23 𝑥2𝑛
1 1 1 1 1 1
𝑃3 = − 𝛿1 − 𝛿2 + + +⋯+ 𝛿3 − ⋯ − 𝛿𝑛
𝑥31 𝑥32 𝑥31 𝑥32 𝑥3𝑛 𝑥3𝑛
⋮
1 1 1 1 1 1
𝑃𝑛 = − 𝛿1 − 𝛿2 − 𝛿3 − ⋯ − + + ⋯+ 𝛿𝑛
𝑥𝑛1 𝑥𝑛2 𝑥3𝑛 𝑥𝑛1 𝑥𝑛2 𝑥𝑛−1

Prof. Muwaffaq I. Alomoush, Yarmouk University, Power Engineering Department, First Semester 2023/2024 70
In Matrix Form (done for 4-bus system):
1 1 1 1 1 1
+ + − −
𝑥12 𝑥13 𝑥14 𝑥12 𝑥13 𝑥14
𝑃1 1 1 1 1 1 1 𝛿1
− + + − −
𝑃2 𝑥21 𝑥21 𝑥23 𝑥24 𝑥23 𝑥24 𝛿2
=
𝑃3 1 1 1 1 1 1 𝛿3
𝑃4 − − + + − 𝛿4
𝑥31 𝑥32 𝑥31 𝑥32 𝑥34 𝑥34
1 1 1 1 1 1
− − − + +
𝑥41 𝑥42 𝑥43 𝑥41 𝑥42 𝑥43

This is – 𝑩 = −sucespatnce Matrix (found from 𝐘 after making 𝑟𝑖𝑗 = 0)

𝑃1 𝐵11 𝐵12 𝐵13 𝐵14 𝛿1

𝑃2 𝐵21 𝐵22 𝐵23 𝐵24 𝛿2
=−
𝑃3 𝐵31 𝐵32 𝐵33 𝐵34 𝛿3
𝑃4 𝐵41 𝐵42 𝐵43 𝐵44 𝛿4

−𝟏
𝜹𝟏 𝑩𝟏𝟏 𝑩𝟏𝟐 𝑩𝟏𝟑 𝑩𝟏𝟒 𝑷𝟏 𝑷𝟏 𝑷𝑮𝟏 𝑷𝑫𝟏
𝜹𝟐 𝑩𝟐𝟏 𝑩𝟐𝟐 𝑩𝟐𝟑 𝑩𝟐𝟒 𝑷𝟐 𝑷𝟐 𝑷𝑮𝟐 𝑷𝑫𝟐
=− ; = −
𝜹𝟑 𝑩𝟑𝟏 𝑩𝟑𝟐 𝑩𝟑𝟑 𝑩𝟑𝟒 𝑷𝟑 𝑷𝟑 𝑷𝑮𝟑 𝑷𝑫𝟑
𝜹𝟒 𝑩𝟒𝟏 𝑩𝟒𝟐 𝑩𝟒𝟑 𝑩𝟒𝟒 𝑷𝟒 𝑷𝟒 𝑷𝑮𝟒 𝑷𝑫𝟒 71
Note that the angle of the reference bus is known, therefore we
must remove the row and the column of the reference bus from
the above matrix solution of the DCLF. Therefore, the DCLF
solution can be expressed as:

𝜹 = − 𝑩′ −𝟏 𝑷 where 𝑷 = 𝑷𝑮 − 𝑷𝑫

𝑩′ = 𝑩 − 𝒓𝒐𝒘 𝒂𝒏𝒅 𝒄𝒐𝒍𝒖𝒎 𝒐𝒇 𝒕𝒉𝒆 𝒓𝒆𝒇𝒓𝒆𝒏𝒄𝒆 𝒃𝒖𝒔

• In the above equation, phase angles are in radians.

• After finding angles, you can find line real power flows 𝑃𝑖𝑗 as
follows:
(𝛿𝑖 −𝛿𝑗 )
𝑃𝑖𝑗 = = −𝑃𝑗𝑖
𝑥𝑖𝑗

72
Prof. Muwaffaq I. Alomoush, Yarmouk University, Power Engineering Department, First Semester 2023/2024
Example: DCLF x13=0.05

For the system shown, all values are given in p.u., PG1=3 PG3=5

and bus (2) is the reference bus. For this system, use
1 3
DC Load Flow to find the phase angles 1 and 3 in
radians, PG2 and P12 in p.u. PD1=4
x12=0.2
PD3=3

−𝑗25 𝑗5 𝑗20 0 0 0 −25 5 20 2

𝒀 = 𝑗5 −𝑗5 0 = 0 0 0 +𝑗 5 −5 0 PG2 PD2= 4
𝑗20 0 −𝑗20 0 0 0 20 0 −20
−25 5 20
𝑩= 5 −5 0
20 0 −20
Bus 2 is the reference bus, therefore, we remove row and column of the reference bus:
−25 20 25 −20 1 20 20 0.20 0.20
𝑩′ = ; −𝑩′ = , −[𝑩′ ]−1 = ( )( ) ( )( ) =
20 −20 −20 20 25 20 − −20 −20 20 25 0.20 0.25
−𝟏
𝛿1 𝐵 𝐵13 𝑃1 𝑃1 𝑃 𝑃 3 4 −1
= − 11 ; = 𝐺1 − 𝐷1 = − =
𝛿3 𝐵31 𝐵33 𝑃3 𝑃3 𝑃𝐺3 𝑃𝐷3 5 3 2

𝛿1 0.20 0.20 −1 0.2 𝑟𝑎𝑑

= =
𝛿3 0.20 0.25 2 0.3 𝑟𝑎𝑑

Prof. Muwaffaq I. Alomoush, Yarmouk University, Power Engineering Department, First Semester 2023/2024 73
 𝛿2 −𝛿1 0 − 0.2
𝑃21 = = = −1 𝑝. 𝑢.
𝑥21 0.2
𝑃𝐺2 = 𝑃𝐷2 + 𝑃21 = 4 + −1 = 3 𝑝. 𝑢.
𝑃12 = − 𝑃12 = 1 𝑝. 𝑢.

Note that the DC load flow solution can be found by inspection, as follows:
Because the system is lossless:
𝑃𝐺1 + 𝑃𝐺2 + 𝑃𝐺3 = 𝑃𝐷1 + 𝑃𝐷2 +𝑃𝐷3
3 + 𝑃𝐺2 + 5 = 4 + 4 + 3
𝑃𝐺2 = 3 𝑝. 𝑢.
𝑃𝐺2 = 𝑃𝐷2 + 𝑃21
3 = 4 + 𝑃21
𝑃21 = −1 𝑝. 𝑢.
𝛿2 −𝛿1
𝑃21 =
𝑥21
0−𝛿1
−1 =
0.2
𝛿1 = 0.2 𝑟𝑎𝑑 74
Prof. Muwaffaq I. Alomoush, Yarmouk University, Power Engineering Department, First Semester 2023/2024
Example: DCLF
PG1=4 PG2=3
SD1=6+j0.5 SD2=6+j2
For the system shown, all values are given in p.u., 𝒛12 = 𝒛13 = 1 2

𝒛23 = 0.02 + 𝑗 0.1 p.u., and bus (3) is the reference bus. For this
system, use DC Load Flow to find the phase angles 1 and 2 in
radians. 3
Solution: PG3
SD3=9+j4

Ignore resistances:
−𝑗20 𝑗10 𝑗10 0 0 0 −20 10 10
𝒀 = 𝑗10 −𝑗20 𝑗10 = 0 0 0 + 𝑗 10 −20 10
𝑗10 𝑗10 −𝑗20 0 0 0 10 10 −20
−20 10 10
𝑩 = 10 −20 10
10 10 −20
Bus 3 is the reference bus, therefore, we remove row and column of the reference bus:
−20 10 20 −10 −1 1 20 10 2/30 1/30
𝑩′ = ;−𝑩′ = , − 𝑩′ = (20)(20)−(−10)(−10) =
10 −20 −10 20 10 20 1/30 2/30
−𝟏
𝛿1 𝐵 𝐵12 𝑃1 𝑃1 𝑃 𝑃 4 6 −2
= − 11 ; = 𝐺1 − 𝐷1 = − =
𝛿2 𝐵21 𝐵22 𝑃3 𝑃2 𝑃𝐺2 𝑃𝐷2 3 6 −3

𝛿1 2/30 1/30 −2 −7/30 −0.23333 𝑟𝑎𝑑

= = =
𝛿3 1/30 2/30 −3 −8/30 −0.26667 𝑟𝑎𝑑

Prof. Muwaffaq I. Alomoush, Yarmouk University, Power Engineering Department, First Semester 2023/2024 75
Example: DCLF

Prof. Muwaffaq I. Alomoush, Yarmouk University, Power Engineering Department, First Semester 2023/2024
For the two-bus system shown, if all values are
given in p.u. and let bus 1 is the reference
bus.Use DCLF method to find 𝛿2 , 𝑃𝐺1 , 𝑃12 , and 𝑃21 .

Solution:
Neglect resistance, and consider only active power.
𝒛12 = 𝑗0.05
1 1 1
𝒚12 = = = = 20∠ − 90𝑜
𝒛12 𝑗0.05 0.05∠90𝑜
𝒀11 𝒀12 𝒚12 −𝒚12
𝒀= = −𝒚 𝒚21
𝒀21 𝒀22 12

−𝑗20 𝑗20
=
𝑗20 −𝑗20
0 0 −20 20
𝒀= +𝑗
0 0 20 −20
−20 20
𝑩=
20 −20
Remove first row and first column of B, because bus 1 is the reference bus whose angle is known.
𝑩′ = [−20]
1
[𝛿2 ] = −[ 𝑩′ ]−𝟏 [𝑃2 ] = (−3) = −0.15 𝑟𝑎𝑑 ; 𝑃2 = 𝑃𝐺2 − 𝑃𝐷2 = 2 − 5 = −3
20

(𝛿1 −𝛿2 ) 0 − (−0.15)

𝑃12 = = = 3 𝑝. 𝑢.
𝑥12 0.05
𝑃21 = −𝑃12 = −3 𝑝. 𝑢.
76
𝑃𝐺1 = 𝑃𝐷1 + 𝑃12 = 4 + 3 = 7 𝑝. 𝑢. ( or use 𝑃𝐺1 + 𝑃𝐺2 = 𝑃𝐷1 + 𝑃𝐷2 )
Example: DCLF
Resolve the previous example if bus 2 is the
reference bus with 𝑃𝐺1 = 7 𝑝. 𝑢., and find
𝛿1 , 𝑃𝐺2 , 𝑃12 , and 𝑃21 .

Answers:
𝛿1 = +0.15 𝑟𝑎𝑑 , 𝑃𝐺2 = 2 𝑝. 𝑢. , 𝑃12 = 3 𝑝. 𝑢. , and 𝑃21 = −3 𝑝. 𝑢..

Prof. Muwaffaq I. Alomoush, Yarmouk University, Power Engineering Department, First Semester 2023/2024 77
Example: DCLF
For the three-bus system shown, if all values are given in p.u. and let bus 1 is the reference

Prof. Muwaffaq I. Alomoush, Yarmouk University, Power Engineering Department, First Semester 2023/2024
bus. Use DCLF method to find 𝛿2 , 𝛿3 , 𝑃𝐺1 , 𝑃12 , 𝑃21 , 𝑃13 , 𝑃31 .

−𝑗40 𝑗20 𝑗20 0 0 0 −40 20 20 −40 20 20

𝒀 = 𝑗20 −𝑗20 0 = 0 0 0 + 𝑗 20 −20 0 , 𝑩 = 20 −20 0
𝑗20 0 −𝑗20 0 0 0 20 0 −20 20 0 −20
Bus 1 is the reference bus, therefore, we remove the first row and the first column from 𝑩:
−20 0 20 0 1 20 0 0.05 0
𝑩′ = ; −𝑩′ = , −[𝑩′ ]−1 = ( )( ) ( )( ) =
0 −20 0 20 20 20 − 0 0 0 25 0 0.05
𝛿2 𝐵22 𝐵23 −𝟏 𝑃2 𝑃 𝑃 𝑃 0 6 −6
=− ; 2 = 𝐺2 − 𝐷2 = − =
𝛿3 𝐵32 𝐵33 𝑃3 𝑃3 𝑃𝐺3 𝑃𝐷3 3 1 2
𝛿2 0.05 0 −6 −0.3 𝑟𝑎𝑑
= =
𝛿3 0 0.05 2 0.1 𝑟𝑎𝑑
(𝛿1 −𝛿2 ) 0 − (−0.3) (𝛿1 −𝛿3 ) 0 − (0.1)
𝑃12 = = = 6 𝑝. 𝑢. , 𝑃13 = = = −2 𝑝. 𝑢.
𝑥12 0.05 𝑥13 0.05
𝑃21 = −𝑃12 = −6 𝑝. 𝑢. , 𝑃31 = −𝑃13 = +2 𝑝. 𝑢.
𝑃𝐺1 = 𝑃𝐷1 + 𝑃12 + 𝑃13 = 5 + 6 + (−2) = 9 𝑝. 𝑢.
or
𝑃𝐺1 + 𝑃𝐺2 + 𝑃𝐺2 = 𝑃𝐷1 + 𝑃𝐷2 +𝑃𝐷3 78
𝑃𝐺1 + 0 + 3 = 5 + 6 + 1 → 𝑃𝐺1 = 9 𝑝. 𝑢.