Unit-3

Load flow studies

Importance of Load Flow Studies

Load flow studies help utilities address several pressing concerns in planning and operations:

 Equipment loading: Checks for overloads on transmission lines, transformers, or any other
equipment.

 Voltage profiles: Ensures voltage levels stay within statutory limits across the system.

 System losses: Assesses total technical losses happening in the network.

 Contingencies: Identifies network vulnerabilities in loss of equipment or lines.

 Expansion needs: Determines if upgrades/augmentation is needed for load growth.

 Outage management: Schedules maintenance while accounting for constraints.

Objectives and Purposes of Load Flow Study

Generally, load flow analysis is aimed at the following technical objectives:

Or Importance of load flow analysis in power system:

 Bus Voltage Magnitudes Calculation

 Active and Reactive Power Flows Determination

 Power Loss Evaluation

 Voltage Profile Examination

 Equipment Loading Assessment

 Transfer Capability Estimation

It serves various purposes like transmission planning, tariff determination, market studies, stability
analysis, contingency screening, and corrective action planning
Bus Classification

Bus Type Specified Variables Unspecified Variables

Load Bus or PQ bus P,Q |V|, δ

Generator bus or voltage controlled Bus or P-V bus P, |V| Q, δ

Slack bus or reference bus or swing bus -|V|, δ P, Q

Slack/Swing Bus - Balances generated and demanded power (normally the largest generator bus).

Voltage Controlled Bus/PV Bus - In this type of bus, voltage magnitude |Vi| and active power Pi are
known, and the phase angle δi and Reactive power Qi are to be determined. Terminal voltage is
maintained constant by generator excitation controls.

PQ Bus/Load Bus - Active and reactive (PDi and QDi) are known, demand is treated as constants and is
not controlled. The |Vi| and δi are unknowns.

Where |V| = Magnitude of voltage

δ = Phase angle of the voltage

P = Active power

Q = Reactive power

Methods Of Load Flow Analysis

Various numerical techniques are employed to solve the nonlinear power flow equations including:

Gauss Seidel method load flow analysis:

 Successive approximation approach

 Slow convergence for large systems

 Prone to numerical instabilities

Newton Raphson method for load flow analysis:

 Most widely used technique

 Faster convergence within 3-5 iterations

 Solves set of equations using matrix inversion

 Requires Jacobian matrix formulation

Fast Decoupled Load Flow (FDLF) Method

 Simpler and suitable for online applications

 Separates real and reactive power equations

 Fast but less accurate than Newton Raphson

Comparison and selection depend on system size, stability requirements, and computation resources.
Modern tools allow automatic switching between methods.

Advantages Of Load Flow Analysis

 Assess power system security, capacity, and vulnerabilities

 Ensure compliance with reliability standards

 Identify needs for system augmentation or flexibility

 Schedule maintenance safely considering all constraints

 Determine optimal generation dispatch and topology

 Minimize transmission losses through optimal scheduling

 Facilitate integration of renewable and distributed energy

Application of Load Flow Analysis

 Transmission planning for new lines, equipment sizing

 Distribution planning for load growth, network reinforcement

 Tariff determination and infrastructure charging rules

 Transmission congestion management through optimal dispatch

 Stability and short circuit studies requiring power flow inputs

 Operational studies including outage analysis

Network model formulation
Consider an ith bus of an ‘n’ bus power system as shown in Fig.

Where, Si = Net injected complex bus power

SGi = complex power injected in ith bus

SDi = complex power absorbed in ith bus

Generator (G) - replaced by (+ )ve complex power source

Transmission line (TL)- replaced by π model

Load - replaced by (- )ve complex power source

SGi = PGi + jQGi

SDi = PDi + jQDi

Si = SGi - SDi = (PGi + jQGi) – (PDi + jQDi) = (PGi - PDi) +j(QGi - PDi)
Applying KCL at ith bus

[I bus]nx1 = [ Ybus ] nxn [ Vbus ] nx1

If we consider 3 bus power system,

 I1  Y11 Y12 Y13  V1 

 I   Y Y Y  V 
 2   21 22 23   2 
 I 3  Y31 Y32 Y33  V3 

Yii = Short circuit driving point admittance or self admittance

Yik = Short circuit transfer admittance
Formation of Y bus or Nodal Admittance matrix

Let us consider the single line diagram of a four bus system as shown below

S1= Complex power injected to bus 1 = SG1 - SD1

S2= Complex power injected to bus 2 = SG2 - SD2

S3= Complex power injected to bus 3 = SG3 - SD3

S4= Complex power injected to bus 4 = SG4 - SD4

 Y12 ' Y13' Y12' Y24 ' Y13' Y34 ' Y24' Y34'
Y10   , Y20   , Y30   , Y40  
2 2 2 2 2 2 2 2
Applying KCL at node 1,

I1  Y10V1  Y13 (V1  V3 )  Y12 (V1  V2 )

I1  (Y10  Y13  Y12 )V1  Y13V3  Y12V2 ……………….(1)

Applying KCL at node 2,

I 2  Y20V2  Y24 (V2  V4 )  Y23 (V2  V3 )  Y12 (V2  V1 )

I 2  (Y20  Y24  Y23  Y12 )V2  Y24V4  Y23V3  Y12V1 ………………..(2)

Applying KCL at node 3,

I3  Y30V3  Y34 (V3  V4 )  Y23 (V3  V2 )  Y13 (V3  V1 )

I3  (Y30  Y34  Y23  Y13 )V3  Y34V4  Y23V2  Y13V1 ………………(3)

Applying KCL at node 4,

I 4  Y40V4  Y34 (V4  V3 )  Y24 (V4  V2 )

I 4  (Y40  Y34  Y24 )V4  Y34V3  Y24V2 ……………………..(4)

Now from equation 1,2,3 and 4,

 I1  (Y10  Y13  Y12 ) Y12 Y13 0  V1 

I    Y (Y  Y  Y  Y )  Y  Y  V 
 
2 12 20 24 23 12 23 24  2
 I3   Y13 Y23 (Y30  Y34  Y23  Y13 ) Y34  V3 
    
I4   0 Y24 Y34 (Y40  Y34  Y24 )  V4 

Note- Y bus matrix is known as sparse matrix

Problem1,

For the 3φ system above ,the series and shunt impedance of line L1 is 14.3+j97 Ω , -j3274 Ω, line L2 is

7.13+j48.60 Ω , -j6547 Ω and line L3 is 9.38+j64 Ω , -j4976 Ω. Find Ybus matrix

Solution,

Line Zs (Ω) ys (℧) x 10-3 Zp (Ω) yp (℧) x 10-3

L1 14.3+j97 1.48-j10.08 -j3274 j0.3054

L2 7.13+j48.60 2.96-j20.14 -j6547 j0.1527

L3 9.38+j64 2.24-j15.30 -j4976 j0.2010

1 14.3 97
ys ( L1)   j  (1.48  j10.08) X 103
14.3  j97 14.3  97
2 2
14.3  97
2 2

Similarly , ys(L2)= 2.96-j20.14 and , ys(L3)= 2.24-j15.30

1 1
y p ( L1)   j  j 0.3054 X 103
 j3274 3274
-3 -3
Similarly , yp(L2)= j0.1527 x 10 and , yp(L3)= j0.2010 x 10

Y11= ys (L1) + ys (L2) + yp (L1) + yp (L2)

=[(1.48-j10.08)+( 2.96-j20.14)+ j0.3054+ j0.1527] x 10-3

=(4.44-j29.26) x 10-3 ℧
Y22= ys (L1) + ys (L3) + yp (L1) + yp (L3)

=[(1.48-j10.08)+( 2.24-j15.30)+ j0.3054+ j0.2010] x 10-3

=(3.72-j24.87) x 10-3 ℧

Y33= ys (L2) + ys (L3) + yp (L2) + yp (L3)

=[(2.96-j20.14)+( 2.24-j15.30)+ j0.1527+ j0.2010] x 10-3

=(5.2-j35.09 ) x 10-3 ℧

Y12=Y21=-ys(L1)=(-1.48+j10.08) x 10-3 ℧

Y13=Y31=-ys(L2)=(-2.96+j20.14) x 10-3 ℧

Y23=Y32=-ys(L3)=(-2.24+j15.30) x 10-3 ℧

 4.44  j 29.26 1.48  j10.08 2.96  j 20.14 

Ybus   1.48  j10.08 3.72  j 24.87 2.24  j15.30  X 103 
 2.96  j 20.14 2.24  j15.30 5.2  j35.09 

Problem2,

Transmission line shown in the following figure is a 230 kv ,200 km line. Resistance , series
impedance and shunt impedance is 0.074 Ω/km , 0.457 Ω/km and -0.277 M Ω/km. Find its bus
admittance matrix.

Solution,

Given, R = 0.074 Ω/km , series impedance (zs) = 0.457 Ω/km

shunt impedance (zp) = -0.277 M Ω/km

300
MVAbase   100MVA / phase
3
 230
kVbase   132.79kV / phase
3
Base Admittance,

MVAbase 100
Ybase    0.00567108 / phase
 kVbase  132.79 
2 2

1 1
Series Admittance (ys)  
zs 200(0.074  j 0.457)

=0.00172633-j0.0106614 ℧/phase

0.00172633  j 0.0106614
= PU ℧/phase
0.00567108

=0.3044093-j1.879959 PU ℧/phase

length
2 100
Shunt Admittance (yp)   ℧/phase
shunt impedance( z p ) 0.277 X 106

=j0.00036101 ℧/phase

j 0.00036101
 PU ℧/phase
0.00567108
=j0.0636582 PU ℧/phase

[Note-while calculating ys consider full length and for yp consider half of the length.]

Y11=Y22=ys+yp

= 0.3044093-j1.879959+ j0.0636582

=0.3044093-j1.8163008 PU ℧/phase

Y12 = Y21 = -ys = - 0.3044093 + j1.879959 PU ℧/phase

 0.3044093  j1.8163008  0.3044093  j1.879959 

Ybus    PU  / phase
  0.3044093  j1.879959 0.3044093  j1.8163008 
Load flow problem

The complex power injected by the source into the ith bus of a power system is

Si = Pi + j Qi = Vi Ii* i=1, 2,….,n

Since it is convenient to work with Ii instead of Ii* of the above equation,

Si = Pi - j Qi = Vi* Ii i=1, 2,….,n …………….(1)

Now from the relation I=VY we can write,

n
I i   YikVk  i=1, 2,….,n
k 1

Now substituting the value of Ii in equation (1),

 n 
Si = Pi - j Qi = Vi* Ii = 
Vi *

 k 1
 Yik Vk  

i=1, 2,….,n

Let, Angle associated with Vi is δi , Vk is δk and Yik is θik

So in polar form , Vi = |vi| ∠ δi , Vk =|vk| ∠ δk , Yik =|Yik| ∠ θik

When phasor are in multiplication their angle are added.
n
Si  vi . Yik . vk    i  ik   k  [-δi is written because of conjugate]
k 1

Pi (Real power)= vi . Y
k 1
ik . vk cos ik   k   i  …………(2)

Qi (Reactive power)=  vi . Yik . vk sin ik   k   i  …………(3)

k 1

 If a system have n number of buses then there will be 2n number of equation

 At each buses there are 4 number of variables-|vi|, δi , Pi , Qi

Equation (2 ) and (3) is known as static load flow equation(SLFE)

transposing all the variables on one side, these equations can be written in the vector form

f(x,y)=0

f = vector function of dimension 2n

x= vector of dependent variables (2n unspecified variables)

y= vector of independent variables (2n specified variables)

Some of the independent variables in y can be used to manipulate some of the state variables.
These adjustable independent variables are called control parameters. Vector y can then be
partitioned into a vector u of control parameters and a vector p of fixed parameters.

u 
y  u= control variable
 p
P= fixed variable

For SLFE solution to have practical significance, all the state and control variables must lie
within specified practical limits These limits, which are dictated by specifications of power
system hardware and operating constraints, are described below:

(i)Voltage magnitude |V i | must satisfy the inequality

| Vi |min | Vi || Vi |max

The power system equipment is designed to operate at fixed voltages with allowable variations
of ± (5 – 10)% of the rated values.

(ii)Certain of the δ i s (state variables) must satisfy the inequality constraint

|  i   k ||  i   k |max
This constraint limits the maximum permissible power angle of transmission line connecting
buses i and k and is imposed by considerations of system stability.

(iii)Owing to physical limitations of P and/or Q generation sources, P Gi and Q Gi are

constrained as follows:

PGi ,min  PGi  PGi ,max

QGi ,min  QGi  QGi ,max
It is, of course, obvious that the total generation of real and reactive power must equal the total
load demand plus losses, i.e.

[NB-From 2n no. of equation specify 2n no of unknowns and using any iterative algorithm find
remaining 2n no of unknowns]

Gauss-Siedel method
It is a iterative tool that help us to solve non linear equation.
Gauss-Siedel method apply to power flow equation
The complex power injected to the ith bus ,
Si = Pi + jQi =Vi Ii*
Si* = Pi – jQi =Vi* Ii
Pi – jQi
Ii  …………….(1)
Vi*
The total current entering to the ith bus of n bus system,
Ii = Yi1V1 + Yi2V2 +……+YiiVi +…..+ YinVn …………….(2)
From equation (1) and (2) we can write,
Pi – jQi n
  YikVk
Vi* K 1

n
Pi – jQi
Or,  YikVk 
K 1 Vi*

 
1  Pi – jQi n
Vi    YikVk  for i = 1 , 2 ….n ……….(3)
Yii  Vi *
K 1

 K i 

1  P2 – jQ2 
V2   *
 Y V
21 1  Y V
23 3  .........  Y2n n 
V
Y22  V2 

1  P3 – jQ3 
V3   *
 Y V
31 1  Y V
32 2  .........  Y3n n 
V
Y33  V3 
Iteration should continue until,

|V(r+1) | - |V(r) | < €

€= Tolerance indicating the convergence.

Acceleration factor
An acceleration factor is a value that can be used to speed up the convergence and reduce the number
of required alteration.

It is found in practice that the process of convergence in GS method is slow. That means it requir a
large no. of iteration before the solution is obtained. The process of convergence may be speed up if the
voltage correction during two consecutive iteration is modify to ,

Viacc  r 1  Vi ( r )   Vi ( r 1)  Vi ( r ) 

Where,

Viacc 
r 1
=Accelerated new valueof voltage of (r+1) iteration.

Vi (r) = Value of voltage of ‘r’ iteration .

Vi (r+1) = Value of voltage of ‘r+1’ iteration .

α = Acceleration factor , it is a real number.

R = Iteration count.

Advantage of GS method
1) It can be easily programmed.

2) The solution technique is very simple.

3) computer memory requirement is smaller.

4) It takes less computational time for iteration.

Limitation of GS method
1) Larger number of iteration is required.

2) Slow convergence rate.

3) Inefficient for large system network.

Example-1,

Generators are connected to all four buses while loads are connected at bus 2 and 3.

Bus P Q V Remark Line R X

1 ..... ….. 1.05∠00 slack 1-2 0.05 0.15
2 0.5 -0.2 ….. PQ 1-3 0.10 0.30
3 -1 0.5 ….. PQ 1-4 0.20 0.40
4 0.3 -0.1 ….. PQ 2-4 0.10 0.30

3-4 0.05 0.15

Solve the problem upto one iteration.

Solution,

As we know,

Bus V δ P Q

Slack ✓ ✓ ? ?

PV ✓ ? ✓ ?

PQ ? ? ✓ ✓

Z12 = 0.05+j0.15

1 1 0.05 0.15
Y12     j  2  j6
z12 0.05  j 0.15 0.052  0.152 0.052  0.152
Similaly,

Y13 = 1-j3 , Y14= 1-j2 , Y24= 1-j3 , Y34= 2-j6

And,

Y11 =Y12+ Y14 + Y13 =(2-j6 )+( 1-j2 )+( 1-j3 ) =4-j11

Y22 = Y21+ Y24 =

Now,

Y11 Y12 Y13 Y14   4  j11 2  j 6 1  j3 1  j 2 

Y Y24   2  j 6 3  j 9 
Y22 Y23  
Ybus   21 Ybus 
0 1 j 3 
Y31 Y32 Y33 Y34   1  j 3 0 3  j 9 2  j 6 
   
Y41 Y42 Y43 Y44   1  j 2 1  j3 2  j 6 4  j11 

Initial value (0th iteration)

V10 = 1.05+j0

V20 = 1+j0 , P2 = 0.5 , Q2 = -0.2

V30 = 1+j0 , P3 = -1 , Q3 = 0.5

V40 = 1+j0 , P4 = 0.3 , Q4 = -0.1

1st iteration

V11 = 1.05+j0

1  P2  jQ2 0
V21    Y V
21 1
1
 Y V
23 3
0
 Y24 4 
V
Y22  (V20 )* 
 1  0.5  j 0.2 
         
3  j9  1  j 0 
( 2 j 6)(1.05 j 0) 0 ( 1 j 3)(1 j 0)


1
 0.5  j 0.2  2.1  j6.3  1  j3
3  j9

=1.03+j0.00566 PU

1  P3  jQ3 0
V31   0 *
 Y V
31 1
1
 Y V
32 2
1
 Y34 4 
V
Y33  (V3 ) 

1  1  j 0.5 
         
3  j9  1  j 0 
( 1 j 3)(1.05 j 0) 0 ( 2 j 6)(1 j 0)


1
  1  j 0.5  1.05  j3.15  2  j 6
3  j9

=1.033-j0.1166 PU

1  P4  jQ4 1
V41    Y41V1
1
 Y42V2
1
 Y43V3 
Y44  (V40 )* 

1  0.3  j 0.1 
   (1  j 2)(1.05  j 0)  (1  j3)(1.03  j 0.00566 )  (2  j 6)(1.033  j 0.1166 ) 
4  j11  1  j 0 

=1.035-j0.0203 PU

Example-2,

Bus-1 slack bus V=0.5∠00 pu

Bus-2 PV bus |V|=1 pu PG=3 pu

Bus-3 PQ bus PL=4 pu QL=2 pu

Carry out one iteration of load flow solution by Gauss Seidel method.
Solution,

As we know,

Bus V δ P Q

Slack ✓ ✓ ? ?

PV ✓ ? ✓ ?

PQ ? ? ✓ ✓

1 1
Y12     j 2.5
z12 j 0.4

1 1
Y13     j3.33
z13 j 0.3

1 1
Y23     j5
z23 j 0.2

  j 5.83 j 2.5 j 3.33 

Ybus   j 2.5  j 7.5 j 5 
 j 3.33 j5  j8.33

P2 = PG2 - PL2 = 3 – 0 = 3 pu

P3 = PG3 - PL3 = 0 – 4 = - 4 pu

Q2 = - Imaginary{(V20)* [Y21V10+Y22 V20 + Y23 V30]} Note:

Q2 = - {(1-j0) [(j2.5)(1.05+j0)+(-j7.5) (1-j0) + (j5)(1+j0)]} P2-jQ2

=-{j2.625 – j7.5 + j5} Q2=- (Imaginary term)

=j0.125

Q2 = -0.125 pu

Q3 = QG3 – QL3 = 0-2 = 2 pu

0Th iteration,

V10 = 1.05+j0

V20 = 1 + j0 , P2 =3 , Q2 = -0.125 pu

V30 = 1 + j0 , P3 =-4 , Q3 = -2

1 st iteration,

V11 = 1.05+j0

1  P2  jQ2 0
V21    Y21V1
1
 Y23V3 
Y22  (V20 )* 

1  3  j 0.125 
    
 j 7.5  1  j 0 
( j 2.5)(1.05 j 0) ( j 5)(1 j 0)

=1 + j0.4

=1.077∠21.80

V21 = 1∠? =1∠21.80 = 0.928+j0.371

1  P3  jQ3 1
V31   0 *
 Y V
31 1
1
 Y32 2 
V
Y33  (V3 ) 

1  4  j 2 
    
 j8.33  1  j 0 
( j 3.33)(1.05 j 0) ( j 5)(0.928 j 0.371)

= 0.73704-j0.25724

=0.78064∠-19.240 pu

Newton Rasphson method

Problem-1,
Perform an iteration of Newton-Raphson load flow method and determine the power flow
solution for the given system.

Take base MVA as 100. Assume V₁=1.05∠0 pu

 R(pu) X(pu) Half line charging admittance (Yp/2) pu
Line From To

1 1 2 0.0839 0.5183 0.0636

Bus PL QL

1 90 20

2 30 10

Solution,

Step-1, (Form Ybus matrix)

1
Y11   j 0.0636 = 0.3044-j1.2164
0.0839  j 0.5183

Y22 = 0.3044-j1.2164

1
Y12 = Y21 = - Y12 =  = - 0.3044 + j1.88
0.0839  j0.5183

 0.3044  j1.2164  0.3044  j1.88

Ybus   
  0.3044  j1.88 0.3044  j1.2164 

1.842  1.405 1.9041.731 

Ybus   
 1.9041.731 1.842  1.405
Step-2 (Assume initial value)

δ = 0 , V = 1 pu
Step-3 ,

1st bus is slack bus so we need not calculate anything.

For bus 2 , we have to calculate P2,calculated = ? , Q2,calculated = ? , ΔP2 = ? , ΔQ2 = ?

Using Static Load Flow Equation,

Pi (Real power)= vi . Y
k 1
ik . vk cos ik   k   i  …………(1)

Qi (Reactive power)=  vi . Y
k 1
ik . vk sin ik   k   i  ………….(2)

P2,calculated = |V2||Y21||V1|cos(θ21+ δ1- δ2) + |V2||Y22||V2|cos(θ22+ δ2- δ2)

=1 x 1.904 x 1.05 x cos(1.731+0-0) + 1 x 1.842 x 1xcos (-1.405)

=-0.3129+0.3039
=- 0.015 pu
30
P2,specified = PG2 -PL = 0  = - 0.3 pu
100

ΔP2 = P2,specified - P2,calculated = - 0.3- (- 0.015) = - 0.285

Q2,calculated = - |V2| |Y21||V1|sin(θ21+ δ1- δ2) - |V2||Y22||V2|sin(θ22+ δ2- δ2)

=1 x 1.904 x 1.05 x sin(1.731) + 1 x 1.842 x 1 x sin (-1.405)

=-1.97-(-1.8167)
=- 0.153 pu
10
Q2,specified = QG2 -QL = 0  = - 0.1 pu
100

ΔQ2 = Q2,specified - Q2,calculated = - 0.1 – (-0.153) = 0.053

Step-4 (Form Jacobian matrix)

 P2 P2 
   | V2 |    2   P2 
 2  
 Q2 Q2    | V2 |  Q2 
   | V2 | 
 2
By derivating equation (1 and (2),
n

Pi = vi . Yk 1
ik . vk cos ik   k   i 

Pi n
 vi . Yik . vk sin ik   k   i 
 i k 1
i

Put i=2 , n=2

P2
 v2 v1 Y21 sin 21  1   2 
 2

=1.0 x 1.05 x 1.904 x sin(1.731)

= 1.973

Pi n
 2 vi Yii cos ii   vk Yik cos ik   k   i 
 | Vi | k 1
i

Put i=2 , k=1 to 2 , n= 2

P2
 2 v2 Y22 cos  22  v1 Y21 cos  21  1   2 
 | V2 |
=[2 x 1.0 x 1.842 x cos (-1.405)]+[1.05 x 1.904 x cos(1.731)]

= 0.60799 + (-0.31891)

=0.289

Qi n
 vi . Yik . vk cos ik   k   i 
 i k 1
i
Put i=2 , n=2

Q2
 v2 v1 Y21 cos  21  1   2 
 2
=1 x 1.05 x 1.904 x cos (1.731)

=- 0.3189

Qi n
 2 vi Yii sin(ii )   vk Yik sin ik   k   i 
 | Vi | k 1
i

Put i=2 , k=2

Q2
 2 v2 Y22 sin( 22 )  v1 Y21 sin 21  1   2 
 | V2 |

=[-2 x 1 x 1.842 x sin (-1.405)]-[1.05 x 1.904 x sin (1.731)]

=3.633 – 1.9732

= 1.6598

 1.973 0.289    2   0.285

 0.3189 1.6598   | V |   0.053 
  2   
1
  2   1.973 0.289   0.285

  | V |  0.3189 1.6598  0.053 
 2     

0.4929 0.085  0.285

  
0.0947 0.5859   0.053 

 0.145 
 
0.00406 
X = X0 + ΔX

     0  0.145   0.145 

  2  2      
V2   V2  1  0.00406 1.00406
V2 = 1.00406 ∠-0.145 pu
Advantages of Newton-Raphson method

 Less number of iteration.

 Less computation time.

 More accurate

 Not sensitive to factors like slack bus selection, regulation transformer etc.

 Number of iteration is almost independent of system size.

 Applicable for large power system networks.

Disadvantages of Newton-Raphson method

 More calculation involved in iteration.

 Per iteration time is large.
 Large computer memory required.
 Programming is difficult.

comparison of load flow methods.