EEE 305: Power System I

(Part-05)

Instructor:
Dr. Md. Nasim Ahmed Dewan
Professor, Department of EEE, BUET

Class Routine:
Sat-9:00, Sun-9:00, Tue-10:00
The Newton-Raphson Power-Flow Solution
Solution:
• Slack bus has no rows or no columns in the jacobian

• A 6X6 matrix would be necessary for remaining 3 buses, if P and Q

were specified for those buses

• Bus (4) is voltage controlled, |𝑉4 | is constant; 𝑄𝑠𝑐ℎ is not specified

• The jacobian will be a 5X5 matrix

• To calculate 𝑃3,𝑐𝑎𝑙𝑐 based on the estimated and specified voltages of

Table 9.3, we need polar form of the off-diagonal entries of Table 9.4
𝑌31 = −5.169561 + 𝑗25.847809 = 26.359695∠101.30993°
𝑌34 = −3.023705 + 𝑗15.118528 = 15.417934∠101.30993°

Also, we need the diagonal entry in rectangular form:

𝑌33 = 8.193267 − 𝑗40.863838
From Eq. (38) we have for i = 3,
𝑃𝑖 = 𝑉𝑖 2 𝐺𝑖𝑖 + σ𝑁
𝑛=1 𝑉𝑖 𝑉𝑛 𝑌𝑖𝑛 cos(𝜃𝑖𝑛 + 𝛿𝑛 − 𝛿𝑖 ) (38)
𝑛≠𝑖

(0)
𝑃3, 𝑐𝑎𝑙𝑐 = 𝑉3 2 𝐺33 + 𝑉3 𝑉1 𝑌31 cos 𝜃31 + 𝛿1 − 𝛿3
+ 𝑉3 𝑉4 𝑌34 cos 𝜃34 + 𝛿4 − 𝛿3 ∵ 𝑌32 = 0

(0) (0)
From Table 9.3 , 𝛿1 = 0, 𝛿3 = 0 and 𝛿4 =0
∴ 𝑃3, 𝑐𝑎𝑙𝑐 = 𝑉3 2 𝐺33 + 𝑉3 𝑉1 𝑌31 cos 𝜃31 + 𝑉3 𝑉4 𝑌34 cos 𝜃34
= 1.0 2 8.193267 + 1.0 × 1.0 × 26.359695 cos(101.30993°)
+ 1.0 × 1.02 × 15.417934 cos(101.30993°)
= −0.06047 pu

200
From Table 9.3 , 𝑃3,𝑠𝑐ℎ = 0 − 100 = −2 pu
(0) (0)
So, Δ𝑃3,𝑐𝑎𝑙𝑐 = 𝑃3,𝑠𝑐ℎ − 𝑃3, 𝑐𝑎𝑙𝑐 = −2 − −0.06047 = −1.3953 pu
Jacobian elements:
2nd row, 3rd column:
From Eq. (52) we have,
𝜕𝑃𝑖
= − 𝑉𝑖 𝑉𝑗 𝑉𝑖𝑗 sin 𝜃𝑖𝑗 + 𝛿𝑗 − 𝛿𝑖 (52)
𝜕𝛿𝑗
𝜕𝑃3
= − 𝑉3 𝑉4 𝑌34 sin 𝜃34 + 𝛿4 − 𝛿3
𝜕𝛿4
= − 1.0 × 1.02 × 15.417934 sin 101.30993° + 0° − 0°
= −15.420898 pu

2nd row, 2nd column:

From Eq. (9.53) we have,
𝜕𝑃𝑖 𝜕𝑃𝑖
= σ𝑁 𝑁
𝑛=1 𝑉𝑖 𝑉𝑛 𝑌𝑖𝑛 sin(𝜃𝑖𝑛 + 𝛿𝑛 − 𝛿𝑖 ) = − σ𝑛=1 (53)
𝜕𝛿𝑖 𝜕𝛿𝑛
𝑛≠𝑖 𝑛≠𝑖
𝜕𝑃3 𝜕𝑃3 𝜕𝑃3 𝜕𝑃3
=− − − = 𝑉3 𝑉1 𝑌31 sin 𝜃31 + 𝛿1 − 𝛿3 − 0 − −15.420898
𝜕𝛿3 𝜕𝛿1 𝜕𝛿2 𝜕𝛿4
= 1.0 × 1.0 × 26.359695 sin 101.30993° + 0° − 0° + 15.420898
= 41.268707 pu
5th row, 5th column:
From Eq. (9.63) we have,
𝜕𝑄𝑖 𝜕𝑃𝑖
𝑉𝑖 =− − 2 𝑉𝑖 2 𝐵𝑖𝑖 = 𝑄𝑖 − 𝑉𝑖 2 𝐵𝑖𝑖 (63)
𝜕 𝑉𝑖 𝜕𝛿𝑖

𝜕𝑄3 𝜕𝑃3
𝑉3 =− − 2 𝑉3 2 𝐵33
𝜕 𝑉3 𝜕𝛿3
2
= −41.268707 − 2 1.0 −40.863838 = 40.458969 pu

Calculating all the elements of jacobian and the power mismatches

(considering 3 decimal places) we have,
 The Newton-Raphson Power-Flow Solution
• This system of equations yields values for the voltage corrections of the
first iteration which are needed to update the state variables according
to Eqs. (49) and (50)
(𝑘+1) (𝑘) (𝑘)
𝛿𝑖 = 𝛿𝑖 + Δ𝛿𝑖 (49)
(𝑘+1) (𝑘) (𝑘) 𝑘 Δ 𝑉𝑖 𝑘
𝑉𝑖 = 𝑉𝑖 + Δ 𝑉𝑖 = 𝑉𝑖 1+ (50)
𝑉𝑖 𝑘

• At the end of first iteration the set of updated voltages at the buses is:
Bus no. i (1) (2) (3) (4)
𝛿𝑖 0 -0.93094 -1.78790 -1.54383
|𝑉𝑖 | 1.00 0.98335 0.97095 1.02

• These updated voltages are then used to re-calculate the jacobian and
mismatches of the second iteration, and so on
 The Newton-Raphson Power-Flow Solution
• The iterative procedure continues until either the mismatches Δ𝑃𝑖 and
Δ𝑄𝑖 become less than their stipulated allowable values or all Δ𝛿𝑖 and
Δ|𝑉𝑖 | become less than the chosen precision index

• When the solution is complete, we can use Eqs. (38) and (39) to
calculate real and reactive power, P1 and Q1, at the slack bus, and the
reactive power Q4 at voltage controlled bus (4)
𝑃𝑖 = 𝑉𝑖 2 𝐺𝑖𝑖 + σ𝑁
𝑛=1 𝑉𝑖 𝑉𝑛 𝑌𝑖𝑛 cos(𝜃𝑖𝑛 + 𝛿𝑛 − 𝛿𝑖 ) (38)
𝑛≠𝑖
𝑄𝑖 = − 𝑉𝑖 2 𝐵𝑖𝑖 − σ𝑁
𝑛=1 𝑌𝑖𝑛 𝑉𝑖 𝑉𝑛 sin 𝜃𝑖𝑛 + 𝛿𝑛 − 𝛿𝑖 (39)
𝑛≠𝑖

• Line flows can also be computed from the differences in bus voltages
and the known parameters of the lines
 The Newton-Raphson Power-Flow Solution

Fig. 9.4

Fig. 9.5
 The Newton-Raphson Power-Flow Solution
• The number of iteration required by the Newton-Raphson method using bus
admittances is practically independent of the number buses

• The time for the Gauss-Seidel method (employing bus admittances) increases
almost directly with the number of buses

• On the other hand, computing the elements of the jacobian is time-consuming,

and the time per iteration is considerably longer for the Newton-Raphson
method

• When sparse matrix techniques are employed, the advantage of shorter

computer time for solution of the same accuracy is in favor of the Newton-
Raphson method for all but very small system
Power-Flow Studies in System Design and Operation
• Electric utility companies use very elaborate programs for power-flow
studies
– aimed at evaluating adequacy of complex, interconnected network

• Important information is obtained concerning

– the design and operation of systems that have yet not been built
– and the effects of changes on existing systems

• A power-flow study for system operating under actual or projected

normal operating condition is called a ‘base case’.

• The results from the base case constitute a benchmark for comparison
of changes in network flows and voltages under abnormal or
‘contingency’ conditions
Power-Flow Studies in System Design and Operation
• The transmission planning engineers can discover system weaknesses
such low voltages, line overloads, or loading conditions deemed
excessive
– These weaknesses can be removed by making design studies involving changes
and/or additions to the base case system

• The system model is then subjected to computer based contingency

testing to discover whether weaknesses arises under contingency
conditions involving abnormal generation schedules or load levels

• Interaction between the system designer and the computer based

power-flow program continues until system performance satisfy local
and regional planning or operating criteria

• A typical power-flow program is capable of handling system of more

than 2000 buses, 3000 lines, and 500 transformers.
– Off course, programs can be expanded to even greater size provided the computer
facilities are sufficiently large
Power-Flow Studies in System Design and Operation
• Data suppled to the computer must include—
– the numerical value of the lines and bus data (such as Table 9.2 and 9.3)
– an indication of whether a bus is the slack bus, a regulated bus where voltage
magnitude is held constant by generation of reactive power Q, or a load bus with
fixed P and Q

• Where voltage are not to be held constant the quantities given in the
tables are interpreted as initial estimates

• Limits of P and Q generation usually must be specified as well as the

limits of line kilovoltamperes

• Unless otherwise specified, the programs usually assume a base of 100

MVA

• Total line-charging megavars specified for each line account for shunt
capacitance and,
Power-Flow Studies in System Design and Operation
• 𝑀𝑣𝑎𝑟𝑠 𝑐ℎ𝑔 = 3 𝑉 𝐼𝑐ℎ𝑔 × 10−3 = 𝜔𝐶𝑛 𝑉 2 (69)
Where,
𝑉 = rated line-to-line voltage in kV,
𝐶𝑛 = Line-to-neutral capacitance (farad), for the entire length of the line
𝐼𝑐ℎ𝑔 = 𝑗𝜔𝐶𝑛 𝑉𝑎𝑛

• The program creates a nominal- representation of the line by dividing

the capacitance computed from the given value of charging megavars
equally between the two ends of the line

• It is evident from Eq. (69) that,

Line charging megavars (pu) = shunt susceptance (pu) at 1.0 pu voltage

• For a long line the computer could be programmed to compute the

equivalent- for capacitance distributed evenly along the line
Power-Flow Studies in System Design and Operation
• The computer printout/results consists of a number of tabulations.

• Most important information is the table that lists—

– each bus number and name
– bus-voltage magnitude (pu) and phase angle
– generation and load at each bus in MWs and Mvars
– Mvars of static capacitors/reactors on the bus
– Accompanying the bus information, is the flow of MWs and Mvars from that bus over each
transmission line connected to that bus
– The totals of system generation and loads are listed in MWs and Mvars
Power-Flow Studies in System Design and Operation
• A system may be divided into areas, or one study may include the
systems of several companies with each designated as a different areas

• The computer program will examine the flow between areas, and
deviations from the prescribed flow will be overcome by causing the
appropriate change in generation of a selected generator in each area

• In actual system operation, interchange of power between areas is

monitored to determine whether a given area is producing that
amount of power which will result in the desired interchange

• A listing (may be obtained) of all buses where the per-unit voltage

magnitude is above 1.05 or below 0.95, or other limits that may be
specified

• A list of line loadings in MVAs can be obtained

Power-Flow Studies in System Design and Operation
• The printout will also list the total MW ( 𝐼 2 𝑅) losses and Mvar ( 𝐼 2 𝑋)
requirements of the system, and both P and Q mismatch at each bus
– Mismatch is an indication of the preciseness of the solution and is the difference
between P (and also Q) entering and leaving each bus

• The numerical results in the printout of Fig. 9.4 are from a Newton-
Raphson power-flow study of the system described in Example 9.5
– The system line data and bus data are provided in Table 9.2 and 9.3
– 3 Newton-Raphson iterations were required

• Similar studies employing the Gauss-Seidel procedure required may

more iterations,
– this is a common observation in comparing the two iterative methods

• Printout shows that,

𝐼 2 𝑅 losses of the system = 504.81-500 = 4.81 MW
Power-Flow Studies in System Design and Operation
• The information can be displayed on a one-line diagram (Fig.9.4)
– showing the entire system
– or a portion of the system, e.g., bus (3) of Fig. 9.5
Power-Flow Studies in System Design and Operation
• Transmission design engineers and system operators usually can call for
video display of such selected power-flow results from computer-
interactive terminal or workstations

• The MW loss in any of the lines can be found by comparing the value of
P at the two ends of the line
98.12 MW flow from bus (1) into line (1)-(3)
97.09 MW flow into bus (3) from the same line
∴ 𝐼 2 𝑅 loss in all 3 phases = (98.12-97.09)=1.03 MW

• The Mvars flow between bus (1) and (3) is slightly complicated because
of the charging Mvars

• The computer considers the distributed capacitance of the line to be

concentrated, half at one end of the line and half at the other end
Power-Flow Studies in System Design and Operation
In the Table 9.3 data given for the line (1)-(3) is 7.75 Mvars
– Computer recognizes this value when voltage = 1.0 per unit
2
– Charging Mvars ∝ |𝑉| ∵ 𝑀𝑣𝑎𝑟𝑠 𝑐ℎ𝑔 = 𝜔𝐶𝑛 𝑉 (69)
At bus (1): 𝑉 = 1.0 pu
At bus (3): 𝑉 = 0.969 pu

So, effective charging at these buses—

7.75 2
Bus (1): × 1.0 = 3.875 Mvars
2
7.75 2
Bus (3): × 0.969 = 3.638 Mvars
2

• Figure shows 61.21 Mvars going from bus (1)

into the line to bus (3) and 63.57 Mvars received
at bus (3). The increase in Mvars is due to line charging

Fig. 9.5
Power-Flow Studies in System Design and Operation
• In three phase flow of MWs and Mvars in the line is shown the single
line diagram of the Fig. 9.6

Fig. 9.6
 Fig. 9.6

Solution:
Total MVA through R and X of all three phases is,
At bus (1) end:
𝑆 = 98.12 + 𝑗65.085 = 117.744∠33.56° MVA
|𝑆| 117.744×103
∴ 𝐼 = = = 295.56 A ∵ Base voltage = 230 kV
3 𝑉𝐿 3×230×1.0

At bus (3) end:

𝑆 = 97.09 + 𝑗59.932 = 114.098∠31.69° MVA
|𝑆| 114.098×103
∴ 𝐼 = = = 295.56 A
3 𝑉𝐿 3×230×0.969
𝑉1 −𝑉3
Note: 𝐼 could also be calculated using 𝐼 =
𝑅+𝑗𝑋
 Fig. 9.6

Now,
𝑘𝑉 2 2302
𝑍𝑏𝑎𝑠 = = = 529 Ω
MVA 100

𝐼2 𝑅 𝑙𝑜𝑠𝑠 = 3 × 295.56 2
× 0.00744 × 529 × 10−6
= 1.03 MW

𝐼2 𝑋 𝑜𝑓 𝑙𝑖𝑛𝑒 = 3 × 295.56 2
× 0.03720 × 529 × 10−6
= 5.157 MVar

Comparison with data in Fig.9.6:

Decrease of MW in line = 98.12 - 97.09 = 1.03 MW
Decrease of Mvar in line = 65.085 - 59.932 = 5.153 MW
 The (Fast) Decoupled Power-Flow Method
• In the strictest use of the Newton-Raphson method, the jacobian is
calculated and triangularized in each iteration in order to update the
LU factors

• In practice, the jacobian is often recalculated only every few iterations

– this speeds up the overall solution process

• The final solution is determined by the allowable power mismatches

and voltage tolerances at the buses

• When solving large-scale power transmission systems, an alternative

strategy for improving computational efficiency and reducing computer
storage requirements is the ‘decoupled power-flow method’
– which makes use of an approximate version of the Newton-Raphson procedure
 The (Fast) Decoupled Power-Flow Method
The decoupled approach is based on two observations:
– Change in 𝛿 at a bus primarily affects the flow of P in the in the transmission lines
and leaves the flow of Q relatively unchanged
– Change in |𝑉| at a bus primarily affects the flow of Q in the transmission lines and
leaves the flow of P relatively unchanged

The fist observation states that,

𝜕𝑃𝑖 𝜕𝑄𝑖
≫
𝜕𝛿𝑗 𝜕𝛿𝑗
𝜕𝑄𝑖
We will consider here ≈0
𝜕𝛿𝑗

The second observation states that,

𝜕𝑄𝑖 𝜕𝑃𝑖
≫
𝜕|𝑉𝑗 | 𝜕|𝑉𝑗 |
𝜕𝑃𝑖
We will consider here ≈0
𝜕|𝑉𝑗 |
 The (Fast) Decoupled Power-Flow Method
• Incorporation these approximation into the jacobian of Eq. (45) makes
the elements of the submatrices J12 and J21 zero

𝜕𝑃2 𝜕𝑃2 𝜕𝑃2 𝜕𝑃2

… |𝑉2 | … |𝑉4 |
𝜕𝛿2 𝜕𝛿4 𝜕|𝑉2 | 𝜕|𝑉4 | Δ𝛿2
⋮ 𝐉11 ⋮ ⋮ 𝐉12 ⋮ ⋮ Δ𝑃2
𝜕𝑃4 𝜕𝑃4 𝜕𝑃4 𝜕𝑃4 Δ𝛿4 ⋮
… |𝑉2 | … |𝑉4 | Δ𝑃4
𝜕𝛿2 𝜕𝛿4 𝜕|𝑉2 | 𝜕|𝑉4 | Δ 𝑉2
𝜕𝑄2 𝜕𝑄2 𝜕𝑄2 𝜕𝑄2
=
Δ𝑄2
(45)
… |𝑉2 | … |𝑉4 | 𝑉2
𝜕𝛿2 𝜕𝛿4 𝜕|𝑉2 | 𝜕|𝑉4 | ⋮ ⋮
⋮ 𝐉21 ⋮ ⋮ 𝐉22 ⋮ Δ 𝑉4 Δ𝑄4
𝜕𝑄4 𝜕𝑄4 𝜕𝑄4 𝜕𝑄4 𝑉4
… |𝑉2 | … |𝑉4 |
𝜕𝛿2 𝜕𝛿4 𝜕|𝑉2 | 𝜕|𝑉4 |

We are then left with two separate systems of equations,

𝜕𝑃2 𝜕𝑃2
… Δ𝛿2 Δ𝑃2
𝜕𝛿2 𝜕𝛿4
⋮ ⋮ ⋮ = ⋮ (77)
𝜕𝑃4 𝜕𝑃4 Δ𝛿4 Δ𝑃4
…
𝜕𝛿2 𝜕𝛿4
 The (Fast) Decoupled Power-Flow Method
𝜕𝑄2 𝜕𝑄2 Δ 𝑉2
|𝑉2 | … |𝑉4 | Δ𝑄2
𝜕|𝑉2 | 𝜕|𝑉4 | 𝑉2
and, ⋮ ⋮ ⋮ = ⋮ (78)
𝜕𝑄 𝜕𝑄 Δ 𝑉4 Δ𝑄4
|𝑉2 | 4 … |𝑉4 | 4
𝜕|𝑉2 | 𝜕|𝑉4 | 𝑉4

• These equations are ‘decoupled’ in the sense that the voltage-angle

corrections Δ𝛿 are calculated using only real power mismatch Δ𝑃,
while the voltage magnitude corrections are calculated using only Δ𝑄
mismatches

• However, J11 and J22 are still interdependent because—

– The elements of J11 depend on the voltage magnitudes being solved in Eq. (78),
– The elements of J22 depend on the angles of Eq. (77)
 The (Fast) Decoupled Power-Flow Method
• Two sets of equations can solved alternately
– Using in one set, the most recent solutions from the other set

• This scheme would still require evaluation and factoring of the two
coefficient matrices at each iteration

• To avoid such computations, we introduce further simplifications, which

are justified by the physics of transmission line power flow

In a well designed and properly operated power transmission system:

• The angular differences 𝛿𝑖 − 𝛿𝑗 between typical buses of the system are
usually so small that,
cos 𝛿𝑖 − 𝛿𝑗 ≈ 1; sin 𝛿𝑖 − 𝛿𝑗 ≈ 𝛿𝑖 − 𝛿𝑗 (79)

• The line susceptances 𝐵𝑖𝑗 are may times larger than the line conductances
𝐺𝑖𝑗 so that
𝐺𝑖𝑗 sin 𝛿𝑖 − 𝛿𝑗 ≪ 𝐵𝑖𝑗 cos 𝛿𝑖 − 𝛿𝑗 (80)
 The (Fast) Decoupled Power-Flow Method
• The reactive power 𝑄𝑖 injected into any bus (i) during normal operation
is much less than the reactive power which would flow if all lines from
that bus were short circuited to reference, i.e.,
𝑄𝑖 ≪ 𝑉𝑖 2 𝐵𝑖𝑖 (81)

Now from Eq. (62) we have the off-diagonal elements of J11 and J22 as,
𝜕𝑃𝑖 𝜕𝑄𝑖
= 𝑉𝑗 = − 𝑉𝑖 𝑉𝑗 𝑌𝑖𝑗 sin 𝜃𝑖𝑗 + 𝛿𝑗 − 𝛿𝑖 (82)
𝜕𝛿𝑗 𝜕 𝑉𝑗

Using the identity sin(𝛼 + 𝛽) = sin 𝛼 cos 𝛽 + cos 𝛼 sin 𝛽 in Eq. 82 we

cam write,
𝜕𝑃𝑖 𝜕𝑄𝑖
= 𝑉𝑗 = − 𝑉𝑖 𝑉𝑗 𝑌𝑖𝑗 sin 𝜃𝑖𝑗 cos 𝛿𝑗 − 𝛿𝑖 + cos 𝜃𝑖𝑗 sin 𝛿𝑗 − 𝛿𝑖
𝜕𝛿𝑗 𝜕 𝑉𝑗
= − 𝑉𝑖 𝑉𝑗 𝐵𝑖𝑗 cos 𝛿𝑗 − 𝛿𝑖 + 𝐺𝑖𝑗 sin 𝛿𝑗 − 𝛿𝑖 (83)

where, 𝑌𝑖𝑗 sin 𝜃𝑖𝑗 = 𝐵𝑖𝑗 and 𝑌𝑖𝑗 cos 𝜃𝑖𝑗 = 𝐺𝑖𝑗
 The (Fast) Decoupled Power-Flow Method
The approximation listed above gives,
𝜕𝑃𝑖 𝜕𝑄𝑖
= 𝑉𝑗 ≈ − 𝑉𝑖 𝑉𝑗 𝐵𝑖𝑗 (84)
𝜕𝛿𝑗 𝜕 𝑉𝑗

For the diagonal elements of J11 and J22 consider Eqs. (54) and (63),
𝜕𝑃𝑖
= −𝑄𝑖 − 𝑉𝑖 2 𝐵𝑖𝑖 (54)
𝜕𝛿𝑖
𝜕𝑄𝑖 𝜕𝑃𝑖
𝑉𝑖 =− − 2 𝑉𝑖 2 𝐵𝑖𝑖 = 𝑄𝑖 − 𝑉𝑖 2 𝐵𝑖𝑖 (63)
𝜕 𝑉𝑖 𝜕𝛿𝑖
Applying the inequality 𝑄𝑖 ≪ 𝑉𝑖 2 𝐵𝑖𝑖 we have,
𝜕𝑃𝑖 𝜕𝑄𝑖
≈ 𝑉𝑖 ≈ − 𝑉𝑖 2 𝐵𝑖𝑖 (85)
𝜕𝛿𝑖 𝜕 𝑉𝑖

By substituting the approximations of Eqs. (84) and (85) in J11 and J22, we
obtain,
 The (Fast) Decoupled Power-Flow Method
− 𝑉2 𝑉2 𝐵22 − 𝑉2 𝑉3 𝐵23 − 𝑉2 𝑉4 𝐵24 Δ𝛿2 Δ𝑃2
− 𝑉3 𝑉2 𝐵32 − 𝑉3 𝑉3 𝐵33 − 𝑉3 𝑉4 𝐵34 Δ𝛿3 = Δ𝑃3 (86)
− 𝑉4 𝑉2 𝐵42 − 𝑉4 𝑉3 𝐵43 − 𝑉4 𝑉4 𝐵44 Δ𝛿4 Δ𝑃4
Δ 𝑉2
− 𝑉2 𝑉2 𝐵22 − 𝑉2 𝑉3 𝐵23 − 𝑉2 𝑉4 𝐵24 𝑉2
Δ𝑄2
Δ 𝑉3
and − 𝑉3 𝑉2 𝐵32 − 𝑉3 𝑉3 𝐵33 − 𝑉3 𝑉4 𝐵34 𝑉3
= Δ𝑄3 (87)
− 𝑉4 𝑉2 𝐵42 − 𝑉4 𝑉3 𝐵43 − 𝑉4 𝑉4 𝐵44 Δ 𝑉4 Δ𝑄4
𝑉4

Now, multiplying the first row with correction vector in Eq. (87),
− 𝑉2 𝐵22 Δ 𝑉2 − 𝑉2 𝐵23 Δ 𝑉3 − 𝑉2 𝐵24 Δ 𝑉4 = Δ𝑄2
Δ𝑄2
⟹ −𝐵22 Δ 𝑉2 − 𝐵23 Δ 𝑉3 − 𝐵24 Δ 𝑉4 = |𝑉2 |
(88)

• The coefficients of Eq. (88) are constants equal to the negative of the susceptance in
the row of Ybus corresponding to bus (2)

• Each other row of Eq. (87) can be similarly treated by representing the reactive
Δ𝑄
mismatch at bus(i) by quantity |𝑉 |𝑖
𝑖
 The (Fast) Decoupled Power-Flow Method
So, Eq. (87) can be re-written as,

Δ𝑄2
Δ 𝑉2 𝑉2
−𝐵22 −𝐵23 −𝐵24
Δ𝑄3
−𝐵32 −𝐵33 −𝐵34 Δ 𝑉3 = 𝑉3
(91)
−𝐵42 −𝐵43 −𝐵44 Δ 𝑉4 Δ𝑄4
𝑉4
• So, all the entries in the coefficient matrix of Eq. (91) is now constants given by
the know susceptance of Ybus

Again, from Eq. (86) we have from first row,

− 𝑉2 𝑉2 𝐵22 Δ𝛿2 − 𝑉2 𝑉3 𝐵23 Δ𝛿3 − 𝑉2 𝑉4 𝐵24 Δ𝛿4 = Δ𝑃2
Δ𝑃2
⟹ − 𝑉2 𝐵22 Δ𝛿2 − 𝑉3 𝐵23 Δ𝛿3 − 𝑉4 𝐵24 Δ𝛿4 = (89)
𝑉2

Putting |𝑉2 | , |𝑉3 | and |𝑉4 | approximately equal to 1.0 per unit in the left hand
side of Eq. (89) we have,
Δ𝑃2
−𝐵22 Δ𝛿2 − 𝐵23 Δ𝛿3 − 𝐵24 Δ𝛿4 =
𝑉2
 The (Fast) Decoupled Power-Flow Method
Δ𝑃2
• Here, represents the real power mismatch
𝑉2

Writing for all buses in matrix form we have,

Δ𝑃2
Δ𝛿2 𝑉2
−𝐵22 −𝐵23 −𝐵24
Δ𝑃3
−𝐵32 −𝐵33 −𝐵34 Δ𝛿3 = 𝑉3
(90)
−𝐵42 −𝐵43 −𝐵44 Δ𝛿4 Δ𝑃4
𝑉4
• Matrix B is generally symmetrical with non-zero elements, which are
constants, real numbers, exactly to equal to the negative of the
susceptance of Ybus

• Matrix B is easily formed, and triangular factors, once computed at the

beginning of the solution, do not have to be recomputed
– Which leads to very fast iteration
 The (Fast) Decoupled Power-Flow Method
• At voltage controlled buses Q is not specified and Δ|𝑉| is zero
– The rows and columns corresponding to such buses are omitted from Eq. (91)

One typical solution strategy is to:

Δ𝑃
1. Calculate the initial mismatches ,
𝑉
2. Solve Eq. (90) for Δ𝛿 ,
ΔQ
3. Update the angles 𝛿 and use them to calculate mismatches ,
𝑉
4. Solve Eq. (91) for Δ|𝑉| and update the magnitude |𝑉| , and
5. Return to Eq. (90) to repeat the iteration until all mismatches are within
specified tolerances.

• Using this decoupled version of the Newton-Raphson method, faster

power-flow solutions may be found within a specified degree of
precision
Solution:
The B matrix can be read directly from Table 9.4 and the mismatches
corresponding to the initial estimates are already calculated in Example
9.5, so that Eq. (90) becomes,
 The (Fast) Decoupled Power-Flow Method
44.835953 0 −25.847809 Δ𝛿2 −1.59661
0 40.863838 −15.118528 Δ𝛿3 = −1.93953
−25.847809 −15.118528 40.863838 Δ𝛿4 2.21286

Solving this equation gives the angle corrections in radians,

(0) (0) (0)
Δ𝛿2 = −0.02057 ; Δ𝛿3 = −0.03781 ; Δ𝛿4 = 0.02609

(0) (0) (0)

Therefore (as 𝛿2 = 𝛿3 = 𝛿4 = 0) ,
(1) (0) (0)
𝛿2 = 𝛿2 + Δ𝛿2 = −0.02057 ;
(1) (0) (0)
𝛿3 = 𝛿3 + Δ𝛿3 = −0.03781 ;
(1) (0) (0)
𝛿4 = 𝛿4 + Δ𝛿4 = 0.02609

Using these values to calculate the reactive power mismatches,

𝚫𝑸𝟐 𝟏
= 𝑸𝟐, 𝒔𝒄𝒉 − 𝑸𝟐, 𝒄𝒂𝒍𝒄
𝑽𝟐 𝑽𝟐
1 − 𝑉2 2 𝐵22 − 𝑌21 𝑉2 𝑉1 sin 𝜃21 + 𝛿1 − 𝛿2
= 𝑄2, 𝑠𝑐ℎ −
𝑉2 − 𝑌24 𝑉2 𝑉4 sin 𝜃24 + 𝛿4 − 𝛿2
 The (Fast) Decoupled Power-Flow Method
2
𝜋
1 −1.0535 + 1.0 −44.835953 + 19.455965 sin 101.30993 × + 0 + 0.02057
= 180
1.0 𝜋
+26.359695 × 1.02 sin 101.30993 × + 0.02609 + 0.02057
180
= −0.80370 pu

𝚫𝑸𝟑 𝟏
= 𝑸𝟑, 𝒔𝒄𝒉 − 𝑸𝟑, 𝒄𝒂𝒍𝒄
𝑽𝟑 𝑽𝟑
1 − 𝑉3 2 𝐵33 − 𝑌31 𝑉3 𝑉1 sin 𝜃31 + 𝛿1 − 𝛿3
= 𝑄3, 𝑠𝑐ℎ −
𝑉3 − 𝑌34 𝑉3 𝑉4 sin 𝜃34 + 𝛿4 − 𝛿3

2
𝜋
1 −1.2394 + 1.0 −40.863838 + 26.359695 sin 101.30993 × + 0 + 0.03781
= 180
1.0 𝜋
+15.417934 × 1.02 sin 101.30993 × + 0.02609 + 0.03781
180
= −1.27684 pu

A reactive mismatch calculation is not required for bus (4) as it is voltage controlled.
Accordingly, Eq. (91) becomes,
44.835953 0 Δ|𝑉2 | −0.80370
=
0 40.86388 Δ|𝑉3 | −1.27684
 The (Fast) Decoupled Power-Flow Method
Solving this equation we have,
(0) (0)
Δ 𝑉2 = −0.01793 ; Δ 𝑉3 = −0.03125 ;

(0) (0)
Therefore (as 𝑉2 = 𝑉3 = 1.0) ,
(1) (0) (0)
𝑉2 = 𝑉2 + Δ 𝑉2 = 1.0 − 0.01793 = 0.98207 pu
(1) (0) (0)
𝑉3 = 𝑉3 + Δ 𝑉3 = 1.0 − 0.03125 = 0.96875 pu

• The above calculation completes the first iternation

• Updated mismatches for second iteration of Eq. (90) are calculated

using the new values of voltage magnitudes

• Repeating the procedure over a number of iterations yield the same

solution as tabulated in Fig. 9.4 obtained by Newton-Raphson method
 The (Fast) Decoupled Power-Flow Method
• While the decoupled power-flow usually takes more iterations to
converge, it is usually significantly faster than the Newton-Raphson
algorithm since the jacobian does not need to be recomputed each
iteration

• Since the mismatch equations themselves have not been modified, the
solution obtained by the decoupled method is the same as that found
with the Newton-Raphson method

• In some situations, in which only an approximate power-flow solution

is needed, the (fast) decoupled approach can be used with a fixed
number of iterations (typically one) to give an extremely fast, though,
approximate solution
 Control of Power Flow
Book: Power System Analaysis and Design, 5e—Glover, Sarma and Overbye

The following means are used to control system power flows:

1. Prime mover and excitation control of generators
2. Switching of shunt capacitor banks, shunt reactors, and static var systems
3. Control of tap-changing and regulating transformers

A simple model of a generator under balanced steady-state conditions is

Thevenin equivalent shown in Fig. 6.7

𝐸𝑔 = 𝐸𝑔 ∠𝛿 𝑉𝑡 = 𝑉𝑡 ∠0° Fig. 6.7

 Control of Power Flow
Here,
𝑉𝑡 = Generator terminal voltage
𝐸𝑔 = Excitation voltage (Generated voltage)
𝛿 = power angle
𝑋𝑔 = positive-sequence synchronous reactance

From the figure, the generator current is,

|𝐸𝑔 |𝑒 𝑗𝛿 −|𝑉𝑡 |
𝐼= (1)
𝑗𝑋𝑔
And the complex power delivered by the generator,
∗
𝐸𝑔 𝑒 −𝑗𝛿 − 𝑉𝑡
𝑆 = 𝑃 + 𝑗𝑄 = 𝑉𝑡 𝐼 = 𝑉𝑡
−𝑗𝑋𝑔
𝑉𝑡 𝐸𝑔 cos 𝛿−𝑗 𝐸𝑔 sin 𝛿−|𝑉𝑡 | 𝑉𝑡 𝐸𝑔 𝑗 cos 𝛿+sin 𝛿 −𝑗 𝑉𝑡 2
= =
−𝑗𝑋𝑔 𝑋𝑔
𝑉𝑡 𝐸𝑔 sin 𝛿
𝑃 = 𝑅𝑒 𝑆 = (3)
𝑋𝑔
𝑉𝑡
𝑄 = 𝐼𝑚 𝑆 = 𝐸𝑔 cos 𝛿 − |𝑉𝑡 | (4)
𝑋𝑔
 Control of Power Flow
• Eq. (3) shows that the real power P increases when the power angle 𝛿
increases

• From an operational standpoint, when the prime mover increases the

power input to the generator while the excitation voltage is held
constant—
– the rotor speed, 𝜔𝑚 ↑ → 𝛿 ↑ → 𝑃 ↑
– There is also a decrease in Q given by Eq. (4), however, when 𝛿 is less than 15°, the
increase in P is much more larger than the decrease in Q

• From the power flow standpoint, an increase in prime mover power

corresponds to an increase in P at the constant-voltage bus to which
the generator is connected
– The power-flow program computes the increase in 𝛿 along with the small change in
Q

• Eq.(4) shows that reactive power output Q increase when the

excitation voltage |𝐸𝑔 | increases
 Control of Power Flow
• From an operational standpoint, when the generator exciter output
increases while holding the prime mover power constant—
– the rotor current ↑ → |𝐸𝑔 | ↑ → 𝑄 ↑
– There is also a small decrease in 𝛿 required to hold P constant Eq. (3)

• From the power flow standpoint, an increase in generator excitation

corresponds to an increase in voltage magnitude at the constant-
voltage bus to which the generator is connected
– The power-flow program computes the increase in reactive power Q supplied by the
generator along with the small change in 𝛿
 Control of Power Flow
• Fig. 6.8 shows the effect of adding a shunt capacitor bank to a power
system bus
– The system is modeled by its Thevenin equivalent

Fig. 6.8

• Before the capacitor bank is connected, the switch SW is open and the
bus voltage equals Eth

• After the bank is connected, SW is closed, and capacitor current IC

leads the bus voltage Vt by 90

• From the phasor diagram, 𝑉𝑡 > |𝐸𝑔 | when SW is closed

 Control of Power Flow
• From the power flow standpoint, the addition of a shunt capacitor
bank to a load bus corresponds to the addition of negative reactive
loads, since a capacitor absorbs negative reactive power
– The power-flow program computes the increase in bus voltage magnitude along with
the small change in 𝛿

• Similarly, the addition of a shunt reactor corresponds to the addition of

a positive reactive load
– The power-flow program computes the decrease in bus voltage magnitude

• Tab changing and voltage-magnitude-regulating transformers are used

to control bus voltage as well as reactive power flows on lines to which
they are connected

• Similarly, phase-angle-regulating transformers are used to control bus

voltage angles as well as real power flows on lines to which they are
connected
 (a) (b)
Figure: Voltage regulating transformers: (a) magnitude control, (b) phase control

• Both tap changing and regulating transformers are modeled by a

transformer with an off-nominal turns ratio

• From the power-flow standpoint, a change in tap setting or voltage

regulation corresponds to a change in the off-nominal turn ratio
– The power flow program computes the change in Ybus, bus voltage magnitudes and
angles, and branch flows
 Control of Power Flow
• Besides the above controls, the power-flow program can be used to
investigate the effect of switching in or out lines, transformers, loads,
and generators

• Proposed system changes to meet future load growth, including new

transmission, new transformers, and new generation can also be
investigated

• Power-flow design studies are normally conducted by trial and error

– Using engineering judgement, adjustments in generation levels and controls are
made until the desired equipment loadings and voltage profile are obtained