Download as pptx, pdf, or txt
Download as pptx, pdf, or txt
You are on page 1of 80

Introduction to Optimal

Power Flow
Prof. Dr. Ashik Ahmed
Dept. of EEE
Islamic University of Technology
Boardbazar, Gazipur-1704
Outline
• Introduction to the optimal power flow problem
 Motivation
 Basic formulation
 Challenges and solution techniques
• Security constrained optimal power flow
 DC formulation
 Solution by relaxation
 Examples
Motivation
l
• Recall the economic dispatch problem F  min  Ci  g i 
i 1

Subject to: g
i 1
i  d  dloss  0

gimin  gi  gimax ; i  1... l

Minimizing the cost of electricity generation required to supply the load


and losses within operation limits of the available system generators.
• There is definitely more to power system operations planning than
minimizing the cost of generation!
Motivation- Continued
There are many good reasons to subject the generation dispatch to the
entire set of power flow equations
• Losses are modelled and minimized exactly.
• It is possible to further impose limits on
 Generator reactive and apparent power outputs
 Voltage magnitudes at generation and load buses.
 Line flows either in terms of active (MW) and apparent (MVA) power as well
as current (A).
• It is possible to model operation under contingent conditions –
preventive dispatch. That is, impose security constraints.
Motivation- Continued
There are many good reasons to subject the generation dispatch to the
entire set of power flow equations
• Include all other network control variables as part of the optimization
goal.
 Generator voltages
 Tap changing transformer tap positions
 Quadrature booster (phase-shifting transformer) tap positions
 Switched capacitor settings
 Static VAR and synchronous condenser reactive power injections
 Load shedding
 DC line flows
OPF Definition and Applications
Power flow problems in which some quantities are optimized subject to
network constraints are called Optimal Power Flow (OPF) problem
• Calculate the optimal generation dispatch and control variables minimizing
operation cost whilst meeting all transmission system limits
• Provide a preventive dispatch when security constraints are included
• Provide sets of corrective measures in emergency situations
• Provide periodical voltage/ reactive power optimization
• Determine the maximum stress a system can withstand in planning studies
(e.g. maximum loadability)
• Perform power system economic analyses- determining bus incremental costs
Control & State Variables and Parameters

We distinguish two basic categories in OPF problems


• Variables
 Control Variables (denoted by the vector u)
 Offer degrees of freedom to optimized power system operation
 Can be modified through human or machine intervention
 Examples: generation set-points, tap changer positions, reactive output from SVCs
 State Variables (denoted by the vector x)
 Arise in modelling the physical response of the power system to control inputs
 Cannot be acted upon directly by human or machine intervention
 Examples: bus voltages (magnitude and angle) and line flows
• Parameters (denoted by the vector y)
 Characterize the technical and economic structure of the power system
 Assume known and constant
 Examples: generation costs, transmission line parameters, etc.
Basic Formulation
min f  u, x, y 
u

Subject to:
G  u, x, y   0
H  u, x, y   0

f(.) is the objective function


 Any Examples?
G(.) represents the constraints imposed by the network power flow
H(.) represents all other constraints imposed on control and state variables
• Any examples of constraints on control variables?
• Any examples of constraints on state variables?
Other Choices for OPF: Minimum Losses
min (Pg  Pd )
Pg ,V pv ,t

Subject to Pg  Pd  G p (t, Vpv , Vpq ,  , y)  0


Qg  Qd  Gq (t, Vpv , Vpq ,  , y)  0
H (Pg , Q g , t, Vpv , Vpq ,  , y)  0

• Equivalent to minimizing the net system active power injection


• Pg , Vpv and t are control variables
• Vpq and θ are state variables
• Pd , Qd are parameters
• H(.) includes restrictions on voltage, line flows, active/reactive power generations
Other Choices for OPF: Maximize Loading
min (  )
 , Pg ,V pv ,t

Subject to Pg   Pd  G p (t, Vpv , Vpq ,  , y)  0


Qg  Qd  Gq (t, Vpv , Vpq ,  , y)  0
H (Pg , Qg , t, Vpv , Vpq ,  , y)  0

• Maximizing the system loading for a given demand pattern


 The loading coefficient λ is increased until load flow is attained without any
limits being violated
 The active/reactive power dispatch and tap/phase changer settings are
adjusted accordingly
Other Choices for OPF: DC OPF
min f  Pg , y 
Pg

Subject to Pg  Pd  B /  0
H (Pg ,  , y)  0

• Minimizing the generation dispatch cost subject to


 DC power flow equations (no reactive power balance)
 Other line flow and generator dispatch constraints
• Can be easily extended to a security constrained version as will be seen later
Other Choices for OPF: SCOPF
min f(Pg (0), y(0))
Pg (k),V (k),t

Subject to: (for k = 0,1,…K)


Pg (k)  Pd (k)  G p (t, V(k),  (k), y(k))  0
Qg (k)  Qd (k)  Gq (t, V(k), (k), y(k))  0
H (Pg (k), Qg (k), V(k),  (k), y(k))  0

• Minimizing system pre-contingency generation cost f(Pg (0), y(0)) using active/reactive
dispatch and tap/phase changers such that
 Active and reactive power will balance in both pre (k = 0) and post (k = 1… K)
contingency conditions
 All generation and network limits are met under pre and post contingency conditions
• Contingency conditions are modelled through the parameters y(k)
DC SCOPF
• Study the SCOPF in linear form only
 Simple, can be worked out by hand
 Neglects reactive power balance and bus voltage magnitudes
 Widely used in the industries
 Easy to implement and solve using off-the-shelf (commercial) software
Formulation of DC SCOPF

Subject to: (for k = 0,1,…K)


Pg (k)  Pd (k)  B / (k) (k)  0
C (k) (k)  f max  0
C (k) (k)  f min  0
H (Pg (k), k)  0
 r (k)  0
• Minimize the pre-contingency generation dispatch cost subject to
 Active power balance at every bus
 Line flow limits (sending and receiving)
 All other generation dispatch constraints
 Fix the reference bus angle to zero
For all pre and post contingency scenarios (k = 0,1, … K )
Challeneges of the OPF
Solving practical OPF problems is difficult because:
• Active power systems are very large
 Dimensionality
• OPF problems are intrinsically nonlinear and nonconvex because of
AC power flow equations
 Multiple optimal solutions
 Slow solution convergence, even divergence
• Many control variables are not continuous-valued (e.g. tap changer
position or breaker states are integer or binary in nature)
 Require mixed-integer solution technique
 An optimal solution can be guaranteed only by exhaustive enumeration
Solution Methods
• Gradient method
• Newton’s method
• Linear programming and interior point method (LP-IP)
• Nature Based Optimization
• Relaxation Method (A variant of Gauss-Siedel Method)
Solution Methods: Relaxation
The concept of relaxation is critical in facilitating the solution of large scale OPF
problems
Definition (Relaxation)
• We say we are solving a relaxation of an OPF if we solve the same OPF while
ignoring some of the complicating aspects of the original problem
• If problem A/ is a relaxation of the minimization problem A then
Feasible space (A)  Feasible space (A )
/

Minimum (A)  Minimum (A / )

• Examples of relaxation include:


 Ignore some of the problem’s original constraints
 Letting discrete-valued control variables assume any value of subsets on the real line.
Solution methods: Successive relaxation
• The main idea is to solve a relaxed OPF and then add violated
constraints one-by-one
• Results in having to solve sequences of smaller-scale problems which
are likely much simpler to solve than the full problem
• Choosing a good initial relaxation is not straightforward
 A “tight” relaxation may require few iterations, but requires carrying many
constraints which may be non-binding
 A “loose” relaxation may require many more iterations
 Striking a proper balance is the key to acceptable solution speed
 Expert knowledge of the problem is useful in formulating good relaxations
• Quite useful with security-constrained OPFs
Solution methods: Successive relaxation
Basic procedure
• Solve an economic dispatch ignoring network constraints and
contingent states
• Calculate the line flows in the intact network using a DC power flow
• Correct line flow violations in the intact network one-by-one by re-
dispatching generators in the most economic way
• Correct post-contingency security violations one-by-one by re-
dispatching generators in the most economic way
Fundamental principle: Use pre-calculated linear sensitivity factors to
model the effects of the changes in generation and network topology
on all line flows
Linear Sensitivity Factors
• Types of Linear Sensitivity Factor:
 Generation outage sensitivity factor (GOSF)
 Line outage sensitivity factor (LOSF)
GOSF
•Definition
 
• GOSF relates the approx. change in power flow in line ‘i-j’ due to outage of
generator at bus ‘k’.
• Case 1: The lost generation is exactly compensated by the slack bus.

GOSF of line ‘i-j’ due to generator outage at bus ‘k’


= Change in power flow through line ‘i-j’
= Change in generation at bus ‘k’
GOSF
• Simulation
  of generation outage: (Generator outage at bus k)
• New line flow (line ij) after generator outage at bus k:
  𝑛
𝑓 =𝑓 0𝑖𝑗 + 𝛥 𝑓 𝑖𝑗 =𝑓 0𝑖𝑗 +𝛼 𝑘𝑖𝑗 𝛥 𝑃𝑘 =𝑓 0𝑖𝑗 − 𝛼 𝑘𝑖𝑗 𝑃 0𝑘
𝑖𝑗
• In general, are constant.
• If fijn  fijmax, the operator should be alerted for taking appropriate actions.
GOSF
• Case 2: The lost generation is compensated by all the remaining ‘on-line’
generators.
• Here, each of the on-line generators would take part of the generation in a
particular ratio.
• Most commonly used method is to divide the lost generation in proportion
to the maximum MW capacity of the on-line generators.
• The proportion of generation picked up by on-line generator ‘g’ is:
Mmax= Total no. of generators
Pg
 gk  MPamax =, Max.
g  k rating of generator ‘a’
 a
a 1
P max
 gk  proportionality factor for generator g
k
when unit k fails
GOSF
• As the generation sensitivity factors are linear in nature, superposition
principle is applicable.
• The new line flow with several generators participating to fulfill the
lost generation:
M
fijn  f ij0   ijk Pk    ija Pa  ak
a 1
k

• Here it is assumed that none of the remaining on-line generators hits


the maximum generation limit.
Calculation of GOSF
• DC load flow is used for this calculation.
• Assumptions in DC load flow:
 The system is lossless, all lines are purely reactive.
 All bus voltages are maintained at 1.0 per unit.
 The angular differences between the end bus voltages of any line is relatively
small.
 So, cosθi ≈ cosθj , and sin(θi – θj) ≈ θi – θj .
• With these assumptions, line power flow equation becomes:
VV 1
sin(i   j )  (i   j ), xij is the reactance of line ij
i j
Pij 
xij xij

• So, the line flow becomes a linear combination of terminal bus voltage
angles.
Calculation of GOSF
• The current flow over the line ‘i-j’ is:
Vi  V j
I ij 
jxij


 V cos 
i i  jVi sin  i  V j cos  j  jV j sin  j 
jxij
Vi cos  i  V j cos  j  j (Vi sin  i  V j sin  j )

jxij
(Vi sin i  V j sin  j ) 1
  (i   j )
xij xij
• So, in DC load flow, the expressions of line power flow and line
current are identical in per unit.
Calculation of GOSF

• From Fast Decoupled load flow:  P   B      B 1  P   X   P


• How to get B’ for DC load flow?
 Read the line data and bus data of a network.
 Ignore the line resistance and shunt susceptance.
 Form the Ybus matrix (n x n). n No. of buses in the network.
 Eliminate the entries of Ybus associated with the swing/slack bus.
 Take the imaginary components of Ybus. Change the signs for all entries.
 This is the required B’ matrix.
 X    B 
1

• Calculate
• Fill up the row and column entries of the [X] matrix (corresponding to
the swing bus) with zeros.
• This is to signify that the swing bus angle will not change during any
kind of outage.
Calculation of GOSF- Example
Calculation of GOSF- Example
• Updated line (branch) data: Ignore R and Bshunt
From bus To bus R (p. u ) X (p. u) Bshunt (p.u)
1 2 0 0.2 0
1 4 0 0.2 0
1 5 0 0.3 0
2 3 0 0.25 0
2 4 0 0.1 0
2 5 0 0.3 0
2 6 0 0.2 0
3 5 0 0.26 0
3 6 0 0.1 0
4 5 0 0.4 0
5 6 0 0.3 0
Calculation of GOSF- Example
• Form Ybus Matrix using the updated line data.

-j13.333 j5.0 0 j5.0 j3.33 0


j5.0 -j27.333 j4.0 j10.0 j3.33 j5.0
0 j4.0 -j17.8462 0 j3.8462 j10.0
j5.0 j10.0 0 -j17.5 j2.5 0
j3.33 j3.33 j3.8462 j2.5 -j16.3462 j3.33
0 j5.0 j10.0 0 j3.33 -j18.33
Calculation of GOSF- Example
• Invert signs for all Ybus entries and pick the imaginary part only.
• Eliminate Swing bus Entries (rows and columns)
• That is B’ matrix. (n-1 by n-1)
27.333 -4.0 -10.0 -3.33 -5.0
-4.0 17.8462 0 -3.8462 -10.0
-10.0 0 17.5 -2.5 0
-3.33 -3.8462 -2.5 16.3462 -3.33
-5.0 -10.0 0 -3.33 18.33
Calculation of GOSF- Example
• Invert B’ to get [X] matrix. (n-1 by n-1)
0.0941 0.0805 0.0630 0.0643 0.0813
0.0805 0.1659 0.0590 0.0908 0.1290
0.0630 0.0590 0.1009 0.0542 0.0592
0.0643 0.0908 0.0542 0.1222 0.0893
0.0813 0.1290 0.0592 0.0893 0.1633

0 0 0 0 0 0
• Update [X] matrix0having0.0941
zeros in the rows and
0.0805
columns
0.0630
corresponding
0.0643 0.0813
to swing
bus entries. (n by 0n) 0.0805 0.1659 0.0590 0.0908 0.1290
0 0.0630 0.0590 0.1009 0.0542 0.0592
0 0.0643 0.0908 0.0542 0.1222 0.0893
0 0.0813 0.1290 0.0592 0.0893 0.1633
Calculation of GOSF- Example
• GOSF for the Example Six bus System Pij
 
k
ij    X  P
Line flows Bus 1 Bus 2 Bus 3 Pk
1
Pij   i   j 
l = 1 (line 1-2) 0 -0.47 -0.40
l = 2 (line 1-4) 0 -0.31 -0.29
xij
l = 3 (line 1-5) 0 -0.21 -0.30
l = 4 (line 2-3) 0 0.05 -0.34 i  X ik Pk
l = 5 (line 2-4) 0 0.31 0.22  j  X jk Pk
l = 6 (line 2-5) 0 0.10 -0.03
l = 7 (line 2-6) 0 0.06 -0.24
l = 8 (line 3-5) 0 0.06 0.29
l = 9 (line 3-6) 0 -0.01 0.37
l = 10 (line 4-5) 0 0 -0.08
l = 11 (line 5-6) 0 -0.06 -0.13
Calculation of GOSF- Example
• 
• Sample GOSF Calculation:
 k = 2  Generation outage at bus 2
 l = 1 (line 1-2)  i = 1, j =2
 xij = x12 = 0.2 (from given line data)
 Xik = X12 = 0 (from X matrix)
 Xjk = X22 = 0.0941 (from X matrix)
 So, GOSF,
Calculation of GOSF- Example
• 
• Sample GOSF Calculation:
 k = 3  Generation outage at bus 3
 l = 2 (line 1-4)  i = 1, j =4
 xij = x14 = 0.2 (from given line data)
 Xik = X13 = 0 (from X matrix)
 Xjk = X43 = 0.0590 (from X matrix)
 So, GOSF,
Line flow Calculation using GOSF
•  Sample Post-Contingency Line Flow Calculation

 Base case flow on line 1-4, = 43.6 MW (usual AC load flow)


 Base case generation on Bus 3, = 60 MW (usual AC load flow)

 GOSF, = - 0.2950 (DC load flow linear sensitivity factor)
 Post-contingency line flow, =+ = 43.6 + (-0.2950) * (-0.295)
= 61.3 MW
Line Outage Sensitivity Factors (LOSF)
f l
LOSFl ,k  0
fk
• LOSFl,k is the line outage sensitivity factor while monitoring line l
after an outage of line k.
• Δfl is the change in MW flow in line l.
0
𝑓
 
• 𝑘is the pre-outage flow in line k
fl n  fl 0  LOSFl ,k f k0
• Updated line flow of line l with line k out,
Line l Line k

Bus i Bus j Bus n Bus m


Line outage simulation
Bus n Bus m Bus n Bus m

Lines to Lines to
remainder of remainder of
the network the network

Line k Line k
Line k before Line k after
Pnm outage outage

(a) (b)

Bus n Bus m

Line k outage
Line k simulated with
injections at bus n
and bus m
Pnm
Pn
Pm
(c)
Line outage simulation
• The opening of line k can be simulated if:
 Pn  Pnm and
 Pm   Pnm
 This means whatever power injection happened at bus n is taken out of bus m
through line k.
 This also means:
 No flow through the breaker connecting bus n to the remaining system and
 No flow through the breaker connecting bus m to the remaining system
 So, both breakers will carry zero flow (as if they are open).
LOSF Expression   
 P 
• From FDLF formulation:    X  P ; P   n 
  
 

 mP

• Now,  n  X nn Pn  X nm Pm


 m  X mn Pn  X mm Pm
• Define:
 Quantities before outage:  n , m , Pnm
 Incremental changes due to outage :  n ,  m , Pnm
 Quantities after outage: n , m , Pnm
LOSF Expression
1  
• If the line reactance is xk , Pnm   n   m

xk
 
• Then,  n  ( X nn  X nm ) Pn
 m  ( X mn  X mm ) Pn
• Then, Pnm  1 n  m  1   n   n   m   m   1   n   m   n   m 
 
xk xk xk
1
• Or, Pnm  Pnm   X nn  X mm  2 X mn  Pn
xk
 
 
• Then, Pn  
1
 Pnm
1  1  X  X  2 X  
 xk nn mm mn

LOSF Expression
1
fl xl
 i   j 
• Generalized LOSF , dl ,k  0 
fk f k0
1  i  j 
   
xl  Pnm Pnm 
1
   i ,nm   j,nm 
xl
LOSF and δ’s
• If neither n nor m is the reference bus: i  X in Pn  X im Pm

• If neither i nor j is the reference bus:

( X in  X im ) xk ( X jn  X jm ) xk
 i ,nm  and  j,nm 
xk   X nn  X mm  2 X mn  xk   X nn  X mm  2 X mn 

1  ( X in  X im ) xk ( X jn  X jm ) xk 
• Then, 
dl ,k   
xl  xk   X nn  X mm  2 X mn  xk   X nn  X mm  2 X mn  
xk  X in  X jn  X im  X jm ) 
  
xl  xk   X nn  X mm  2 X mn  
LOSF and δ’s
• If either n or m is the reference bus:

• If either i or j is the reference bus:

X in xk
 i,nm  if m is the reference bus
xk  X nn
 X im xk
 i,nm  n is the reference bus
if
xk  X mm

• If bus i itself is the reference bus:  i,nm  0


Example : Six –bus system LOSF
k=1 k=2 k=3 k=4 k=5
(Line 1-2) (Line 1-4) (Line 1-5) (Line 2-3) (Line 2-4)
l=1 0.64 0.54 -0.11 -0.50
l=2 0.59 0.46 -0.03 0.61
l=3 0.41 0.36 0.15 -0.11
l=4 -0.10 -0.03 0.18 0.12
l=5 -0.59 0.76 -0.17 0.16
l=6 -0.19 -0.06 0.33 0.22 0.23
l=7 -0.12 -0.04 0.21 0.51 0.15
l=8 -0.12 -0.04 0.20 -0.38 0.14
l=9 0.01 0.005 -0.03 -0.62 -0.02
l = 10 0.01 -0.24 0.29 0.13 -0.39
l = 11 0.11 0.03 -0.18 0.12 -0.13
Example : Six –bus system LOSF
k=6 k=7 k=8 k=9 k = 10 k = 11
(Line 2-5) (Line 2-6) (Line 3-5) (Line 3-6) (Line 4-5) (Line 5-6)
l=1 -0.21 -0.12 -0.14 0.01 0.01 0.13
l=2 -0.06 -0.04 -0.04 0 -0.33 0.04
l=3 0.27 0.16 0.18 -0.02 0.32 -0.17
l=4 0.23 0.47 -0.40 -0.53 0.17 0.13
l=5 0.30 0.17 0.19 -0.02 -0.67 -0.19
l=6 0.24 0.27 -0.03 0.31 -0.26
l=7 0.27 0 -0.20 0.58 0.20 0.44
l=8 0.27 -0.17 0.47 0.19 -0.42
l=9 -0.03 0.64 0.60 -0.02 0.56
l = 10 0.24 0.14 0.15 -0.02 0.15
l = 11 -0.23 0.36 -0.40 0.42 -0.18
Example : Six –bus system LOSF Calculation

• Sample LOSF Calculation:


 K = 2  Line outage between buses 1 and 4. So, n =1, m =4.
 l = 9  Effect to be seen on line between buses 3 and 6. So, i = 3, j =6.
 xk = x14 = 0.2 (from given line data)
 xl = x36 = 0.1 (from given line data)
 Xin = X31 = 0, Xim = X34 = 0.05897, (from X matrix)
 Xjn = X61 = 0, Xjm = X64 = 0.05920, (from X matrix)
 Xnn = X11 = 0, Xnm = X14 = 0, Xmm = X44 = 0.10088, (from X matrix)
0.2  0  0  0.05897  0.05920 
 So, LOSF, dl ,k     0.0050
0.1  0.2   0  0.1088  2*0  
Example : Six –bus system LOSF Calculation

• Sample LOSF Calculation:


 K = 8  Line outage between buses 3 and 5. So, n =3, m =5.
 l = 9  Effect to be seen on line between buses 3 and 6. So, i = 3, j =6.
 xk = x35 = 0.26 (from given line data)
 xl = x36 = 0.1 (from given line data)
 Xin = X33 = 0.1659, Xim = X35 = 0.09077, (from X matrix)
 Xjn = X63 = 0.12895, Xjm = X65 = 0.08927, (from X matrix)
 Xnn = X33 = 0.1659, Xnm = X35 = 0.09077, Xmm = X55 = 0.12215, (from X
matrix)
0.26  0.1659  0.12895  0.09077  0.08927 
 So, LOSF, d l ,k     0.6005
0.1  0.26   0.1659  0.12215  2*0.09077  
Example : Six –bus system
•  Sample Post-Contingency Line Flow Calculation

 Base case flow on line 3-5, = 19.1 MW (usual AC load flow)


 Base case flow on line 3-6, = 43.8 MW (usual AC load flow)
 LOSF, dl,k= 0.6005 (DC load flow linear sensitivity factor)
 Post-contingency line flow, =+ dl,k = 43.8 + (0.6005) * (19.1)
= 55.26 MW
OPF Example: 3-bus, 4-line Network
Line data (Sbase = 100 MVA) Bus 1 Line 1 Bus 2
Line no, l bl (p u) fl max (MW) G1 G2
Line 2
1 5 100
2 5 100
200 MW
3 5 60
Line 3 Line 4
4 5 80

Generator data
Gen no. gi min (MW) gi max (MW) Ci ($/MWh) Bus 3
G3 50 MW
1 100 250 20
2 20 100 40
3 0 50 50 What is the most economic N-1 secure
dispatch assuming we approximate the
network with a DC power flow?
DC OPF- Relaxation Approach
• Solve
  an economic dispatch ignoring network constraints and contingent states
min F = 20g1 + 40g2 + 50g3
subject to:
g1 + g2 + g3 = 250 = Pdem
g1 100
-g1 -250
g2 20
-g2 -100
g3 0
-g3 -50
DC OPF- Relaxation Approach
• Solve an economic dispatch ignoring network constraints and contingent
states
55
 Since generator 1 is the cheapest
50
we maximize its output
45
 By inspection, we get
40
 g1[0] = 230 MW
35
 g2[0] = 20 MW

g 3 (MW)
 g3[0] = 0 MW 30

F[0] = 5400 $/h 25


ED solution
20
In the g2-g3 plane we only have g2 =20 MW
g3 = 0 MW
15
(g1 = 230 MW)
upper/ lower generation limits 10

0
0 10 20 30 40 50 60 70 80 90 100
g 2 (MW)
DC OPF- Relaxation Approach
• Calculate the line flows in the intact network using a DC power flow
 By inspection, B’ and C matrices are
 5 5 0 
 15 10 5  5 5 0 
B '   10 15 5 C 
 5 0 5 
 5 5 10   
 0 5 5 
a

b
5 0 0 0
0 1 1 1 0
0
1 2
5 0
diag (bl )    A   1 1 0 1  C  diag (bl ) AT
c d 0 0 5 0  
   0 0 1 1
0 0 0 5
3
DC OPF- Relaxation Approach
• Calculate the line flows in the intact network using a DC power flow

 Assigning Bus 1 as the reference bus,


the X and H matrices are

0 0.4 0.2 
0 0 0  0 0.4 0.2 
X   0 0.08 0.04  H 
0 0.2 0.6 
 0 0.04 0.12   
0 0.2 0.4 
DC OPF- Relaxation Approach
• Calculate the line flows in the intact network using a DC power flow
 The bus voltage angles corresponding to the economic dispatch are

0 0 0   2.3   0 
 0  XP 0  0 0.08 0.04   1.8   0.164 
   
0 0.04 0.12   0.5  0.132 

 The corresponding line flows are0 0.4 0.2   82 


 0 0.4 0.2   2.3   
 1.8   82
f 0 = C 0  HP 0  100  
   66 

 0 0.2 0.6 
   0.5  
 0 0.2 0.4   16 
DC OPF- Relaxation Approach
••Calculate
  the line flows in the intact network using a DC power flow
Line 3 is overloaded! 20 MW
230 MW
How can we mitigate this 82 MW
overload ?
We will redispatch the generation G1 G2
ensuring that f3 60 MW 82 MW
In other words, we will “tighten”
200 MW
the relaxation
66 MW
16 MW

0 MW G3 50 MW
DC OPF- Relaxation Approach
••Correct
  the line flow violations in the intact network by redispatching generations in the most
economic way

How do we tighten the relaxation to impose f 3 60 MW ?


o We determine the effect of redispatching on f3 through generation shift factors h3i

f3  f 3[0]  h31P1  h32 P2  h33 P3

oThat is f 3  f 3[0]  h31 (g1  g1[0] )  h32 (g 2  g[0]


2 )  h33 (g 3  g [0]
3 )

 66  0.2(g 2  20)  0.6(g 3  0)


 70  0.2 g 2  0.6g 3

70  0.2g 2  0.6g3  60
o We append the constraint to relieve flow limit violation in line 3
DC OPF- Relaxation Approach
•• Correct
  the line flow violations in the intact network by redispatching generations in the
most economic way
 The tighter version of the economic dispatch is
min F = 20g1 + 40g2 + 50g3
subject to
2g2 + 6g3 100
g1 + g2 + g3 = 250 = Pdem
g1 100
-g1 -250
g2 20
-g2 -100
g3 0
-g3 -50
DC OPF- Relaxation Approach
• Correct the line flow violations in the intact network by redispatching
generations in the most economic way 55

 In the g2-g3 plane, we see 50

45
o Upper/lower generation limits
40
o The flow limit on line 3 cutting off
35
the previous solution

g 3 (MW)
30
 Which generator should we use to 25

relieve this overload? 20

15
Generator Rate of g2 (MW) g3 (MW) Cost of line Pre-fault flow limit on line 3
10
providing line relief relief ($/h)
relief (MW/MW) 5

2 -0.2 50 0 600 0
0 10 20 30 40 50 60 70 80 90 100

3 -0.6 20 10 300 g 2 (MW)


DC OPF- Relaxation Approach
• Correct the line flow violations in the intact network by redispatching
generations in the most economic way
 We choose the cheaper option
where generator 3 provides relief 55

 The optimal dispatch is now 50

 g1[1] = 220 MW 45

 g2[1] = 20 MW 40
 g3[1] = 10 MW 35
 F[1] = 5700 $/h

g 3 (MW)
30 New ED
g2 = 20 MW
25 g3 = 10 MW
(g1 = 220 MW)
20

15

10

0
0 10 20 30 40 50 60 70 80 90 100
g 2 (MW)
DC OPF- Relaxation Approach
• Correct the line flow violations in the intact network by redispatching generations in the
most economic way

Verify that this dispatch meets all line 20 MW


220 MW
flow limits 80 MW
Indeed!
G1 G2
What is cost of imposing the line
flow limits? 80 MW
200 MW
60 MW
20 MW

10 MW G3 50 MW
DC SCOPF- Relaxation Approach
• Correct post-contingency security violations by redispatching generations in the most
economic way

We perform N-1 security analysis


Take lines out one at a time and evaluate corresponding impacts on remaining lines
Can we do so without recalculating everything ?
Yes! Using line outage distribution factors, dl|k

 Startingwith line 1, we find the immediate impact of losing this line using the factors dl|k
for K =1 and l = 2, 3, 4
DC SCOPF- Relaxation Approach
•• Correct
  post-contingency security violations by redispatching generations in the most
economic way
First we get for buses i = 1, 2, 3 and n =1 and m=2, the values of
1|12  0
 X 22
 2|12   0.1333
1  b1 X 22
 X 32
3|12   0.0667
1  b1 X 22

And then the line outage distribution factors dl|1 , l = 2, 3, 4



d 21  b2 112   212  0.6667 
d31  b 
3 112
 312   0.3333
d 41  b 
4 212
 312   0.3333
DC SCOPF- Relaxation Approach
• Correct post-contingency security violations by redispatching
generations in the most economic way
20 MW
 We then find flows associated to the 220 MW

failure of line k = 1 (or equally k =2) G1 G2


 For example, in l = 2 (or equally l =1)
133 MW
f 2|1  f [1]
2 d f
2|1 1
[1]
200 MW

 80  0.6667x 80 86.7 MW
46.7 MW

=133.3 MW > f 2max


 Similarly, for in l = 3 and 4 10 MW 50 MW
G3

f 3|1  86.7 MW > f 3max


 Post- contingency overloads on
f 4|1  46.7 MW  f max
4 lines 2 and 3
DC SCOPF- Relaxation Approach
• Correct post-contingency security violations by redispatching
generations in the most economic way 20 MW
220 MW
 Likewise, we find flows associated to the 110 MW

failure of line k = 3 G1 G2
 For example, in l = 1 (or equally l =2) 110 MW
f1|3  f1[1]  d3|1f 3[1] 200 MW

 80  0.5 x 60 40 MW

=110 MW > f 2max


 Similarly, for in l = 4 10 MW G3 50 MW

f 4|3  40 MW < f 4max


 Overloads on line 1 and 2
DC SCOPF- Relaxation Approach

•  How do we mitigate these overloads?



 We redispacth the generation ensuring that f2 100 MW and f3 60 MW
when lines 1 (or equally 2) are taken out
To mitigate the effect of line contingency k on line l, we impose, using
compensated generation shift factors
l
f̂ l  f l|k   l i | k Pi  f l
max

i 1

where f  𝑙∨ 𝑘 is the flow on line l after the failure of line k prior to redispatching
DC SCOPF- Relaxation Approach
 The outage of line k =1 (or k =2), implies two flow violations. One on l = 2
(l =1) and the other on l = 3
 Starting with k = 1 and l =2 , we have

f̂ 2  f 2|1   21|1 (g1  g )  22|1 (g 2  g )   23|1 (g3  g )  f


[1]
1
[1]
2
[1]
3
max
2

 So, we need to calculate the values of  2 i |1 for i = 1, 2, and 3


 By inspection, we get
2 1
 21|1  0  22|1    23|1  
3 3
DC SCOPF- Relaxation Approach
 Hence, with k =1 and l =2 ,we have

2 1
f̂ 2  133.33  (g 2  20)  (g 3  10)
3 3
 Limiting the line flow to 100 MW and simplifying, we obtain

2 g 2  g3  150

 What next?
 We add this constraint to the original economic dispatch along with
the constraint limiting the pre-fault flow on line 3
DC SCOPF- Relaxation Approach
That is min F = 20g1 + 40g2 + 50g3
55

subject to 50

2g2 + g3 ≥ 150 45

2g2 + 6g3 ≥ 100 40

g1 + g2 + g3 = 250 35

g 3 (MW)
g1 ≥ 100 30

25
-g1 ≥ -250
20
g2 ≥ 20
15 Post-fault limit
-g2 ≥ -100 10
on line 2 due to
failure of line 1
g3 ≥ 0 5
-g3 ≥ -50 0
0 10 20 30 40 50 60 70 80 90 100
g 2 (MW)
DC SCOPF- Relaxation Approach
• Correct post-contingency security violations by redispatching
generations in the most economic way
55

50
The “tighter” optimal dispatch is now
45
 g1[2] = 175 MW 40

 g2[2] = 75 MW 35

 g3[2] = 0 MW

g 3 (MW)
30

 F[2] = 6500 $/h 25

20
New ED
g2 = 75 MW
15
 Verify that the pre-fault limits are satisfied 10
g3 = 0 MW
(g1 = 175 MW)

 Is the l =2 post-fault flow due to the failure 5

of line k =1 within limits? 0


0 10 20 30 40 50 60 70 80 90 100
g 2 (MW)
DC SCOPF- Relaxation Approach

• Correct post-contingency security violations by redispatching


generations in the most economic way

 Now looking at the effect of the failure of line k =1 (or k =2 ), on line l = 3

f̂ 3  f 3|1   31|1 (g1  g1[1] )   32|1 (g 2  g 2[1] )   33|1 (g 3  g3[1] )  f 3max


 So, we need to calculate the values of 𝜓  3 𝑖 ∨ 1 for i = 1, 2, and 3
 By inspection, we get
1 2
 31|1  0  32|1    33|1  
3 3

 Limiting the post-fault flow to 60 MW thus requires


g 2  2 g3  120
DC SCOPF- Relaxation Approach

• Correct post-contingency security violations by redispatching


generations in the most economic way
55
min F = 20g1 + 40g2 + 50g3 50

45

subject to 40
Post-fault limit on line 3
due to failure of line 1
g2 + 2g3 ≥ 120 35

2g2 + g3 ≥ 150

g 3 (MW)
30

2g2 + 6g3 ≥ 100 25

g1 + g2 + g3 = 250 20

g1 ≥ 100 15

-g1 ≥ -250 10

g2 ≥ 20 5

-g2 ≥ -100 0
0 10 20 30 40 50 60 70 80 90 100
g3 ≥ 0 g 2 (MW)

-g3 ≥ -50
DC SCOPF- Relaxation Approach

• Correct post-contingency security violations by redispatching


generations in the most economic way
55

50

Another “tighter” optimal dispatch is now 45


New ED
g1[3] = 160 MW 40 g2 = 60 MW
g3 = 30 MW
g2[3] = 60 MW 35
(g1 = 160 MW)

g 3 (MW)
30
g3[3] = 30 MW
25
F[3] = 7100 $/h 20

 Verify that the pre-fault limits are 15

satisfied 10

 Is the l =3 post-fault flow due to the 5

0
failure of line k =1 (k =2) within limits? 0 10 20 30 40 50 60
g 2 (MW)
70 80 90 100
DC SCOPF- Relaxation Approach

• Correct post-contingency security violations by redispatching


generations in the most economic way

 Lastly, looking at the effect of the failure of line k =3 on line l = 1 (or equally l =2 )
f̂1  f1|3   11|3 (g1  g1[1] )   12|3 (g 2  g 2[1] )   13|3 (g 3  g3[1] )  f1max

 We calculate the values of𝜓


  1 𝑖 ∨ 3 for i = 1, 2, and 3
 By inspection, we get
1 1
 11|3  0  12|3    13|3  
2 2

 Limiting the post-fault flow to 100 MW thus requires

g 2  g3  50
DC SCOPF- Relaxation Approach

• Correct post-contingency security violations by redispatching


generations in the most economic way
min F = 20g1 + 40g2 + 50g3 55

50

subject to 45

g2 + g3 ≥ 50 40

g2 + 2g3 ≥ 120 35

2g2 + g3 ≥ 150

g 3 (MW)
30 Post fault limits on line
1 and 2 due to failure
2g2 + 6g3 ≥ 100 25 of line 3

g1 + g2 + g3 = 250 20

g1 ≥ 100 15

-g1 ≥ -250 10

g2 ≥ 20 5

-g2 ≥ -100 0
0 10 20 30 40 50 60 70 80 90 100
g3 ≥ 0 g 2 (MW)

-g3 ≥ -50
DC SCOPF- Relaxation Approach

• Correct post-contingency security violations by redispatching


generations in the most economic way

 What is the optimal dispatch here?


 It is the same as we found by correcting for the loss of line 1 (or equally 2)!
 Constraints on post-fault flows in line 1 and 2 after the loss of line 3 are
inactive
 That is, they are redundant
 So is the pre-fault flow limit on line 3!
Observation
We call the failure of line 1 (or equally 2) an umbrella contingency
because it covers for other, less stringent, failures
DC SCOPF- Relaxation Approach

• Correct post-contingency security violations by redispatching


generations in the most economic way
 • Pre-fault limits
 Is the final economic dispatch
60 MW
160 MW
secure with respect to 60 MW
 Pre-fault line flow limits? G1 G2
 Post-fault line flow limits?
60 MW
200 MW
40 MW
20 MW

30 MW G3 50 MW
DC SCOPF- Relaxation Approach

• Correct post-contingency security violations by redispatching


generations in the most economic way
  • Post-fault flows: Loss of line 1 or 2
 Is the final economic dispatch
60 MW
160 MW
secure with respect to
 Pre-fault line flow limits? G1 G2
 Post-fault line flow limits?
100 MW
200 MW
60 MW
40 MW

30 MW G3 50 MW
DC SCOPF- Relaxation Approach

• Correct post-contingency security violations by redispatching


generations in the most economic way
  • Post-fault flows: Loss of line 3
 Is the final economic dispatch
60 MW
secure with respect to 160 MW
80 MW
 Pre-fault line flow limits?
G1 G2
 Post-fault line flow limits?
80 MW
200 MW

20 MW

30 MW G3 50 MW
DC SCOPF- Relaxation Approach

• Correct post-contingency security violations by redispatching


generations in the most economic way
  • Post-fault flows: Loss of line 4
 Is the final economic dispatch
60 MW
160 MW
secure with respect to 70 MW
 Pre-fault line flow limits? G1 G2
 Post-fault line flow limits?
70 MW
 All limits are met! The algorithm 200 MW

terminates 20 MW

 What is the cost of security?


30 MW G3 50 MW

You might also like