Newton-Raphson

Power Flow

Newton-Raphson Algorithm

The second major power flow solution method is
the Newton-Raphson algorithm.

Key idea behind Newton-Raphson is to use
sequential linearization

General form of problem: Find an x such that

f(x)

=

0

Newton-Raphson Method

(v)

1. For each guess of x,

(v) ( v )

x

, define

Ax

=

x-

x

2. Represent

f x) by a Taylor series about

f

(x)

(v)

(v)

df

(x

)

(v)

f x)

=

f

(x

1d
+

2

) +

2

f(x

2

dx

dx

(v)

)

Ax

(v)

Ax

(

+

2

+higher order terms

)

Newton-Raphson Method

3. Approximate

f x) by neglecting all terms

except the first two

(v)

f x) = 0

(v)

~ f(x ) +

df(x )

dx

(v)

Ax

(v)

4. Use this linear approximation to solve for

Ax

1

(v)

(v)

Ax =

df

(x )

dx

(v)

f(x )

5. Solve for a new estimate of

(v+1) ( v ) ( v )



x

=

x

+Ax

Example

2

Use Newton-Raphson to solve

f

(x)

=

x

-2

=

0

The equation we must iteratively solve is

1

(v)

(v)

Ax =

(v)

Ax =

(v+1)

df(x

dx

1

(v)

2x

(v)

)

(v)

f(x )

(v) 2

((x ) -2)

(v)

x = x +A x

(v+1) ( v )

x = x

1

(v)

(v) 2

((x ) -2)

2x

Example Solution

(v+1) ( v )

x = x

(0)

1

(v)

2x

(v) 2

((x ) -2)

Guess x

=1. Iteratively solving we get

(v)

(v)

(v)

v

x

f

(x

)

Ax

0

1

1

0.5

1

1.5

0.25

0.08333

3

3

2

1.41667

6.95310

2.45410

6

3

1.41422

6.02410

Comments

When close to the solution the error decreases

quite quickly -- method has quadratic convergence

Stopping criteria is when |f(x(v)) | < c

Results are dependent upon the initial guess.
What if we had guessed x(0) = 0, or x (0) = -1?

Multi-Variable Newton-Raphson

Next we generalize to the case where x is an n-

dimension vector, and

x

f(x) is an n-dimension function
f(x)

1

1

x

f

(x)

2

x = f ( x )

x

2

=

f ( x )

n

n

Again define the solution x so

f x)

=

0 and

Ax=

xx

Multi-Variable Case, contd

The Taylor series expansion is written for each f
c f(x) c f(x)

i

(x)

f(x) = f

1

(x)+ A x

1

+ A x +

1 1

cx
cf

(x)

1

1

2

cx

2

1

cx

Ax

n

n

+higher order terms

f

n

cf

(x) = f ( x )+

n

n

(x) c f

Ax +

1

(x)

n

Ax +

2

cf

n

cx

(x)

1

cx

2

cx

Ax

n

n

+higher order terms

Multi-Variable Case, contd

This can be written more compactly in matrix form

cf

1

(x) c f(x) c f(x)

1 1

cx

cx

cx

1
f

1

f

(x)
c f

(x)

2

2

(x) c f ( x )

2

n

Ax

1

cf ( x )

2

Ax

2

f x)=

+ c x c x

2

cx

f

n

(x)
c f

n

1 2

(x) c f ( x )

n

n

Ax

n

cf ( x )

n

cx

cx

cx

1

2

n

+

higher order terms

Jacobian Matrix

The n by n matrix of partial derivatives is known

as the Jacobian matrix,
c f(x)

J(x)
cf(x) c f(x)

1

1

1

cx

cx

cx

1

cf

2 n

(x) c f ( x ) c f ( x )

2 2 2

J(x) = c x

cx c x

cf

n

1 2 n

(x) c f ( x ) c f ( x )

n n

cx

cx

cx

1

2

n

Multi-Variable N-R Procedure

Derivation of N-R method is similar to the scalar case

f x)
f x)
Ax

(v+1)

=
=
~

f(x) +J(x x+) higher order terms
0~f(x) +J(x)Ax

1

J(x) f ( x )

(v) ( v )

x

=

x

+Ax

(v+1)

(v)

(v)

1

(v)

x

=

x

J(x

)

f(x

)

(v)

Iterate until

f(x

)

<c

Example

Solve forx=

f(x)=2x

x
x

2

+

1

such that

2

2

x 8=0

f (x)=0 where

1

f ( x )=x

1 2

2 2

x +x

x 4=0

2

1

2

1

2

First symbolically determine the Jacobian

cf

(x)

cf

(x)

1

cx
J(x)=

1

1

cx

2

cf

2

(x) c f ( x )

2

cx

1

cx

2

Solution

J(x) =

4x
2x+

1

x x

2x

2

2x

1

Then
A x 4 x

2 1 2

1

2x f(x)

1

Ax

1

=
2 x+

2 1

x x2x f ( x )

2 1 2

(0)

Arbitrarily guessx =

1 4

1 2 2

1

1

1

2 5 2.1

(1)

x =
1

=

3 1 3 1.3

Solution, contd

(2)

x =

2.1

1.3

8.40 2.60


5.50 0.50

1

2.51 1.8284

=
1.45 1.2122

Each iteration we check
specified tolerancec

0.1556

f(x) to see if it is below our

(2)

f(x ) =
0.0900

If

c

= 0.2 then we would be done. Otherwise we'd

continue iterating.

NR Application to Power Flow

We first need to rewrite complex power equations
as equations with real coefficients

S

i

*

= V I i =

n

V Y V k

i
k=1

*

n

* *

=V Y V k

i
k=1

These can be derived by defining

Y

=

G

+

jB

ik ik

V = V

ik

jui

e =V Z u

u

i
ik

i i i

= u u

i k

j

Recall e

u =

cosu +

jsinu

Power Balance Equations

S

i

= P+jQ

i i

n

= V V

n n

* *

=V Y V =

i

k=1 k = 1

(cosu +jsinu

juik

V V e

i k

)(G jB

(G jB )

ik ik

)

i

k

ik

ik

ik

ik

k=1

Resolving into the real and imaginary parts

n

P

i

Q

= V V

i k
k=1

n

= V V

(G cos u

ik ik

(G sin u

+B sin u

ik

B cos u

)=P P

ik Gi Di

)=Q Q

i i

k=1

k ik ik

ik ik Gi Di

NR Power Flow

In the Newton-Raphson power flow we use Newton's
method to determine the voltage magnitude and angle
at each bus in the power system.

We need to solve the power balance equations

n

P

i

Q

= V V

i k
k=1

n

= V V

(G cos u

ik

(G sin u

+B sin u

ik ik

B cos u

)=P P

ik Gi Di

)=Q Q

i i

k=1

k ik ik

ik ik Gi Di

Power Flow Variables

Assume the slack bus is the first bus (with a fixed

voltage angle/magnitude). We then need to determine
the voltage angle/magnitude at the other buses.

x =

u

u

V

2

n

f(x)

P(x) P +P

2 G 2 D 2

P(x) P +P

n Gn Dn

=
Q(x) Q + Q

2 2

V Q

G2 D 2

(x) Q + Q

n

n

Gn

Dn

N-R Power Flow Solution

The power flow is solved using the same procedure
discussed last time:

(v)

Setv = 0; make an initial guess of

(v)

Whilef(x )>c Do

(v+1) ( v ) ( v ) 1 ( v )

x,

x

x

=

x

J(x

)

f(x

)

v

=

v

+1

End While

Power Flow Jacobian Matrix

The most difficult part of the algorithm is determining

and inverting the n by n Jacobian matrix,

J(x)

cf

1

(x) c f(x) c f(x)

1 1

cx

cx

cx

1

cf

2 n

(x) c f(x) c f(x)

2 2 2

J(x) = c x

cx c x

cf

n

1 2 n

(x) c f ( x ) c f ( x )

n n

cx

cx

cx

1

2

n

Power Flow Jacobian Matrix,

Jacobian elements are calculated by differentiating

each function, f

i

(x), with respect to each variable.

For example, if f

n

i

(x) is the bus i real power equation

f(x) = V

V ( G cos u

+B sin u

)P +P

i

cf

i

k=1

n

(x)
=

i k ik

V V ( G

ik ik

sinu +B

ik Gi Di

cosu )

cu

i

i k = 1

k=i

k ik ik ik ik

cf(x)

i

= V

V ( G sin u

B cos u

) (j=i)

cu

i j
j

ik ik ik ik

Two Bus, Example

For the two bus power system shown below, use the Newton-

Raphson power flow to determine the voltage magnitude and angle
at bus two. Assume that bus one is the slack and SBase = 100 MVA.

u

x =

2

Y

bus

j10 j 10
=

V

2

j10 j10

Two Bus Example, contd

General power balance equations

n

P

i

Q

= V V

i k
k=1

n

= V V

(G cos u

ik ik

(G sin u

+B sin u

ik

B cos u

)=P P

ik Gi Di

)=Q Q

i i k ik ik ik

k=1

Bus two power balance equations
P=VV(10sinu) +2.0=

ik Gi Di

0

2

Q

2

2 1 2

= V V ( 10cos u

2 1

2

) +V (10) + 1.0=0

2 2

Two Bus Example, contd

P

2

Q

(x)= V

2

(x)= V

(10sinu ) + 2.0

2

(10 cosu ) +V

= 0

2

(10)+1.0=0

2

2

2

2

Now calculate the power flow Jacobian

cP

2

(x) c P(x)

2

cu
J(x)=

2

cV

2

cQ

2

(x) c Q(x)

2

cu
10V

2

cosu

cV

2

10 sinu

=

2

10V

2

2 2

sinu 10cos u +20V

2 2 2

First Iteration

(0)

Setv = 0, guessx =

Calculate

0

1

V

2

(0)

f(x )=

(10sinu ) + 2.0

2 2.0

=

2

V

2

(10cosu

) +V (10) + 1.0 1.0

2 2

(0)

J(x )=

10V cos u

2 2

10V sin u

10 sinu
10cosu+

10 0

2

=
20V 0 10

(1)

Solve x =

2

0

1

2

10 0


0 10

2 2

1

2.0 0.2

=
1.0 0.9

Next Iterations

(1)

f(x )=

(1)

J(x )=

0.9(10sin(0.2))+2.0

2

0.9(10 cos(0.2))+0.9 10 + 1.0

8.82 1.986

1.788 8.199

1

0.212

=
0.279

(2)

x =

0.2
0.9

8.82 1.986

1.788 8.199

0.212 0.233

=
0.279 0.8586

(2)

f(x )=

(3)

f(x )=

0.0145

0.0190

0.0000906

0.236

(3)

x =
0.8554

Done! V =

0.8554Z13.52

2

0.0001175

Two Bus Solved Values

Once the voltage angle and magnitude at bus 2 are known we can
calculate all the other system values, such as the line flows and the
generator reactive power output

PV Buses

Since the voltage magnitude at PV buses is fixed there
is no need to explicitly include these voltages in x or
write the reactive power balance equations

the reactive power output of the generator varies to
maintain the fixed terminal voltage (within limits)

optionally these variations/equations can be

included by just writing the explicit voltage constraint
for the generator bus

|Vi | - Vi setpoint = 0

Three Bus PV Case Example

For this three bus case we have

u

x = u

2

3

P

2

f(x) = P

3

(x) P +P

G2 D 2

(x) P +P =0

G3 D 3

V

2

!"

Q(x) + Q

2 D 2