Power System Analysis - 2(BEE703)

.IN
C
Module 4

N
Economic Operation of Power Systems

SY
Introduction

U
VT

Studied smart, not hard — thanks to [Link]

INTRODUCTION

.IN
EPS: ECONOMIC ASPECTS

C
• Proper assessment of load requirement of a

N
given area where power is needed.

SY
• Power to be supplied using available units:

U
thermal, hydroelectric, nuclear, etc.
VT
• Factors considered while choosing a unit: fuel
available/ cost, availability of sites for stations,
nature of load to be supplied, etc.
Studied smart, not hard — thanks to [Link]
Further,
• Load is not constant due to varying

.IN
demands at different times of the day.

C
N
• EPS

SY
- Is expected to supply Reliable and
Quality Power
U
VT
- Should ensure Continuity of Power
Supply at all times.
Studied smart, not hard — thanks to [Link]
EPS: OPERATIONAL ASPECTS

.IN
• Electric energy is generated at power

C
stations far away from the load centers.

N
SY
• Large, long lines (grid lines) wheel the
generated power to substations at load
U
centers.
VT
• Many electrical equipment are used for
proper T&D of the generated power.
Studied smart, not hard — thanks to [Link]
Economic loading of generators and
interconnected stations:

.IN
Economic efficiency is achieved when all the

C
units are loaded so that the fuel cost of power

N
generation is minimum. The units then share

SY
the load to minimize the overall generation cost.

U
This economical approach of catering to the load
VT
requirement is called as ‘Economic Dispatch’.
The main factor in economic operation of power
systems is the cost of generating the real power.
Studied smart, not hard — thanks to [Link]
IN EPS, COST HAS 2 COMPONENTS:

Fixed Costs: Capital investment, tax paid,

.IN
labour, salary, etc. – Independent of load.

C
N
Variable Costs: Losses, daily load

SY
requirements, purchase/sale of power, etc. -
depends on load.
U
VT
Factors affecting operating cost of units:
Generator efficiency, losses, fuel cost, etc.
The fuel cost is the most important factor.
Studied smart, not hard — thanks to [Link]
EPS is a mix of various types of units:

.IN
• Hydro, thermal, nuclear, hydro-thermal,
wind, tidal, etc.

C
• Each type of unit contributes its share for the

N
SY
total operating cost.

U
Since fuel cost is a predominating factor in
VT
thermal (coal fired) plants, ELD is considered
usually for a given set of thermal plants in the
foregoing discussion.
Studied smart, not hard — thanks to [Link]
 PROBLEM OF ELD:
* The problem of Economic Scheduling:

.IN
UCP: objective- to find various generators to be in
operation among available ones, satisfying

C
N
constraints, so that operating cost is minimum,

SY
solved for specified time, based on forecasted load.
ELD: objective- to find power output of each

U
presently running units to meet the specified load
VT
(including losses), so that fuel cost is minimum.
* The problem of OPF: deals with delivering real
power to load points with minimum loss.
Studied smart, not hard — thanks to [Link]
PERFORMANCE CURVES:

.IN
1. I-O Curve: plot of fuel input- Btu/Hr. as a function of
power output,

C
N
SY
U
VT

Studied smart, not hard — thanks to [Link]

[Link] Rate Curve: plot of heat rate- Btu/kWh, as a function of power
output

.IN
C
N
SY
U
VT

Department
Studied ofhard
smart, not EEE,—GSSSIETW, Mysuru
thanks to [Link]
3. Incremental Fuel Rate Curve: plot of
incremental fuel rate (IFC) in Btu/kWh as a
function of power output,

.IN
C
N
SY
U
VT

Department
Studied ofhard
smart, not EEE,—GSSSIETW, Mysuru
thanks to [Link]
4. Incremental Fuel Cost Curve: plot of incremental fuel cost (IFC) in
Rs./kWh as a function of the power output, where, IFC in Rs./kWH =
(Incremental fuel rate in Btu/kWH) (Fuel cost in Rs./Btu)

.IN
C
N
SY
U
VT

Department
Studied ofhard
smart, not EEE,—GSSSIETW, Mysuru
thanks to [Link]
IFC is a measure of how

.IN
costlier it will be to produce
an increment of power output by that unit.

C
N
The Cost Curve can be approximated by:

SY
* Quadratic Curve: Ci(Pi) = ai+biPi+ciPi2 Rs./Hr.

U
* Linear curve: d(Ci)/dPi = bi+2ciPi Rs./MWHr.
VT
Generally, the Quadratic curve is used to represent
cost curve, with IC curve given by the Linear curve.
Studied smart, not hard — thanks to [Link]
IFC is a measure of how

.IN
costlier it will be to produce
an increment of power output by that unit.

C
N
The Cost Curve can be approximated by:

SY
* Quadratic Curve: Ci(Pi) = ai+biPi+ciPi2 Rs./Hr.

U
* Linear curve: d(Ci)/dPi = bi+2ciPi Rs./MWHr.
VT
Generally, the Quadratic curve is used to represent
cost curve, with IC curve given by the Linear curve.
Studied smart, not hard — thanks to [Link]
CONSTRAINTS IN ELD:
Primary (equality) constraints:

.IN
Pi-PDi-PL=0; Qi-QDi-QL=0; i=buses

C
Secondary (inequality) constraints:

N
SY
Pimin ≤ Pi ≤ Pimax Qimin ≤ Qi ≤ Qimax

U
VT i=generating units
Spare Capacity Constraints:
ΣPGi(Gen.) ≥ ΣPL(Losses) + PSR + ΣPDi
PSR -Spinning Reserve
Studied smart, not hard — thanks to [Link]
Constraints contd.

.IN
Thermal Constraints:

C
Simin ≤ Sbi ≤ Simax i=branches, Sbi -branch MVA

N
Bus voltage and Bus angle Constraints:

SY
Vimin ≤ Vi≤ Vimax δijmin ≤ δij ≤ δijmax
i=1,2,….n; j=1,2,….m

U
VT
Other Constraints:
Timin≤Ti≤Timax Ti -% tap-setting of TC transformer
Psimin ≤ PSi ≤ PSimax PSi -phase shift of PST.

Studied smart, not hard — thanks to [Link]

SPINNING RESERVE:
- Amount of generation available from all units minus

.IN
present load plus losses supplied. PSR = Σ PGi-{ΣPDj+ΣPL}

C
- Must be made available so that loss of one/ more units

N
does not cause a large drop in system frequency.

SY
- Must be allocated to units based on Council rules.

U
- One such rule is as follows: ‘SR must be capable of
VT
making up for loss of the heavily loaded unit in the system’
- Must be spread around system to avoid problem of
‘bottling of reserves’ and to allow various parts of the
system to run as ‘islands’, when they become disconnected.
Studied smart, not hard — thanks to [Link]
 SOLUTION TO ELD:

.IN
The solution to ELD problem is obtained as per equal

C
incremental cost (EIC) criterion which states that:

N
SY
‘All the units must operate at the same

U
incremental fuel cost for economic operation’
VT
This EIC criterion can be derived as per LaGrangian
multiplier method for different cases as under.

Studied smart, not hard — thanks to [Link]

CASE (i) Solution to ELD
without inequality constraints:

.IN
Consider a system with N units supplying load PD

C
MW. Let MW limits and losses be negligible.

N
Suppose that the fuel cost of unit-i is given by:

SY
Ci(Pi) = ai+ biPi+ ciPi2 Rs./Hr. so that

U
ICi = d(Ci)/dPi = bi+ 2ciPi Rs./MWHr.
VT
Total cost, CT = Σ Ci(Pi) i=1,2,…N
ELD problem can be stated
mathematically as follows.
Studied smart, not hard — thanks to [Link]
Minimize CT = Σ Ci(Pi) i=1,2,…N

.IN
Such that Σ Pi = PD

C
This optimization problem can be solved by

N
LaGranje’s method as follows.

SY
LaGranje’s function L is: L = CT + λ (PD – ΣPi)

U
The minimum cost is obtained when:
VT
∂L/∂Pi=0 and ∂L/∂λ=0 i.e.,
∂CT/∂Pi – λ = 0 and
PD - Σ Pi = 0 (which is same as constraint)
Studied smart, not hard — thanks to [Link]
Further, since cost of a unit

.IN
depends only on its own output, we have,
∂CT/∂Pi=∂Ci/∂Pi=dCi/dPi i = 1,2,…N

C
Thus, dCi/dPi – λ = 0 i = 1,2,…N

N
SY
dC1/dP1= dC2/dP2= …=dCi/dPi=….=dCN/dPN
= λ, a common IFC, Rs./MWHr.

U
VT
This is stated in words as: ‘For optimum generation
(power output) of the generating units, all units
must operate at equal incremental cost (EIC)’
Studied smart, not hard — thanks to [Link]
Expression for
System Lambda, λ:

.IN
C
Consider ICi= d(Ci)/dPi

N
or λ = bi+2ciPi Rs./MWHr.

SY
Simplifying, Pi =[λ-bi]/2ci MW

U
so that VT Σ Pi= PD= Σ [λ-bi]/2ci
Solving, we get,
λ = {PD + Σ [bi/2ci]}/ {Σ [1/2ci]}
Studied smart, not hard — thanks to [Link]
CASE (ii) Solution to ELD
With Capacity constraints:

.IN
Consider a system with N units supplying load PD

C
MW. Let MW limits be considerable and losses

N
negligible. With fuel cost of unit-i given by:

SY
Ci(Pi) = ai+ biPi+ ciPi2 Rs./Hr. so that

U
ICi = d(Ci)/dPi = bi+ 2ciPi Rs./MWHr.
VT
Total cost, CT = Σ Ci(Pi) i=1,2,…N
ELD problem can be stated
mathematically as follows.
Studied smart, not hard — thanks to [Link]
Minimize

.IN
CT=ΣCi(Pi) i = 1,2,…N

C
Such that ΣPi=PD

N
SY
and Pimin ≤ Pi ≤ Pimax

U
Where, VT
Pimin is the min. output and
Pimax is the max. output by unit i.
Studied smart, not hard — thanks to [Link]
The necessary conditions for solution
of this optimization problem are:

.IN
λi =dCi/dPi = λ Rs./MWHr. for Pimin ≤ Pi ≤ Pimax

C
λi = dCi/dPi ≤ λ Rs./MWHr. for Pi = Pimax

N
λi = dCi/dPi ≥ λ Rs./MWHr. for Pi = Pimin

SY
Thus, if outputs of a unit, according to optimality rule, is:

U
* < min. value, then it is set to Pimin, IC > sys.λ
VT
* > max. value, then it is set to Pimax, IC < sys.λ
* Between the max. and min. values,
then IC = system λ.
Studied smart, not hard — thanks to [Link]
The sequential procedural
steps are as follows:

.IN
C
• First, find output according to EIC Criterion

N
• If output< minimum value, then set the value = Pimin,

SY
• If output> maximum value, then set the value = Pimax,
• Adjust demand for remaining units after accounting for

U
settings made for such units (units which violate the limits)
VT
• Finally, use EIC criterion, for remaining units. Here, the
system lambda is determined by only those units whose
power output values are with in specified MW limits.
Studied smart, not hard — thanks to [Link]
CASE (iii) Solution to ELD
With Transmission losses

.IN
considered - PENALTY FACTOR METHOD
Consider a system with N units supplying load PD

C
N
MW. Let MW limits and losses be considerable.

SY
With fuel cost of unit-i given by:
Ci(Pi) = ai+ biPi+ ciPi2 Rs./Hr. so that
U
VT
ICi = d(Ci)/dPi = bi+ 2ciPi Rs./MWHr.
Total cost, CT = Σ Ci(Pi) i=1,2,…N
Let PL be line losses in the system.
Studied smart, not hard — thanks to [Link]
ELD problem can be stated
mathematically as follows:

.IN
Minimize CT = Σ Ci(Pi) i = 1,2,… N

C
Such that Σ Pi = PD + PL

N
SY
This is solved by LaGranje’s method as follows:
LaGranje function L is: L = CT-λ(ΣPi-PD–PL)

U
The minimum cost value is obtained when:
VT
∂L/∂Pi=0 and ∂L/∂λ=0 i.e.,
dCT/dPi –λ(1-dPL/dPi)=0 and
ΣPi-PD–PL=0 (which is same as constraint)
Studied smart, not hard — thanks to [Link]
 Further, since cost of a unit
depends only on its own output, we have,

.IN
∂CT/∂Pi=∂Ci/∂Pi=dCi/dPi i = 1,2,…N

C
Thus, dCi/dPi–λ(1-dPL/dPi)=0 i=1,2,…N

N
or dCi/dPi = ICi = λ(1-dPL/dPi)

SY
U
So that we have for optimal operation,
VT
λ=ICi /(1-dPL/dPi) =ICi(1-dPL/dPi)-1 = PniICi Where,
Pni is penalty factor of unit i =(1-dPL/dPi)-1 = (1-ITLi)-1 ;
ITLi =dPL/dPi is Incremental Transmission Loss of unit i,
and λ is in Rs./MWHr.
Studied smart, not hard — thanks to [Link]
This is stated in words as under:

.IN
‘For optimum generation of generating units,
when transmission losses are considered, all the

C
N
units must operate such that the product of

SY
incremental fuel cost and their penalty

U
factor must be the same for all units’
VT
Note: If losses are negligible as in case(i) , then,
ITLi = dPL/dPi = 0, Pni=1.0 so that λ = ICi as before.

Studied smart, not hard — thanks to [Link]

CASE (iv) Solution to ELD
With Transmission losses

.IN
considered – B-COEFFICIENTS METHOD

C
N
Consider a system with two units supplying

SY
currents I1 and I2 to load current IL. Let Ik1

U
and Ik2 be the currents flowing through a
VT
general branch element, k of resistance Rk,
with current Ik as shown in figure below.
Studied smart, not hard — thanks to [Link]
 .IN
C
N
SY
Branch
currents

U
VT in a two
unit
system

Studied smart, not hard — thanks to [Link]

Ik = Ik1+ Ik2 = Nk1I1+ Nk2I2

.IN
Where, Nk1 and Nk2 are current
Distribution factors of units 1 and 2.

C
Here, Ik1 and IL and Ik2 and IL are assumed to

N
have the same phase angle and thus, they are

SY
added as real numbers.

U
VT
Let I1=│I1│σ1=│I1│cos σ1+j│I1│sin σ1
I2=│I2│σ2=│I2│cos σ2+j│I2│sin σ2
σ1,σ2 are phase angles of currents.
Studied smart, not hard — thanks to [Link]
Consider magnitude
of current Ik, in branch k, as

.IN
Ik = {Nk1│I1│cos σ1+jNk1│I1│sin σ1}+

C
{Nk2│I2│cos σ2+jNk2│I2│sin σ2}

N
= {Nk1│I1│cos σ1+Nk2│I2│cos σ2}+

SY
j {Nk1│I1│sin σ1+Nk2│I2│sin σ2}
Thus,

U
│Ik│2 = Nk12│I1│2 + Nk22│I2│2 + 2Nk1Nk2│I1││I2│cos(σ1-σ2)
VT
However, P1=3V1I1cos1; P2=3V2I2cos2; PL=3│Ik│2Rk
where, P1,P2 are MW outputs, V1,V2 are line voltages, 1,2
are power factor angles and PL is transmission loss.
Studied smart, not hard — thanks to [Link]
After simplification, expression for
transmission loss as a function of

.IN
plant generation can be obtained as:

C
PL = Nk12RkP12/(V12cos21)+ Nk22RkP22/(V22cos22)

N
+ 2Nk1Nk2RkP1P2 cos(σ1-σ2)/(V1V2cos1cos2)

SY
= B11P12 + B22P22 + 2B12P1P2
Where, B coefficients are called, loss coefficients.

U
VT
In general, for a system of n units we have,
PL=  Pi Bij Pj Where,
i j
Bij=NkiNkjRk cos(σi-σj)/(ViVjcosicosj)
Studied smart, not hard — thanks to [Link]
Note:
1. The B coefficients are represented in units of reciprocal

.IN
MW, (MW-1)

C
2. For a 3-unit system, PL=B11P12+B22P22+B33P32+

N
2B12P1P2+2B13P1P3+2B23P2P3

SY
3. PL= PTBP Where, PT=[P1 P2 P3], vector of unit MW

U
output values, B = [B11 B12 B13; B21 B22 B23; B31 B32
VT
B33], the loss coefficient matrix
4. The B coefficient matrix is a square, symmetric
matrix of order N.

Studied smart, not hard — thanks to [Link]

Note contd.:
5. The following are the assumptions

.IN
made during the above analysis:

C
• All load currents maintain a constant ratio to load

N
current (Nki=constant).

SY
• The voltage at any bus remains constant.

U
• The power factor of each bus source is constant
VT
(i=constant).
• The voltage phase angle at load buses is constant
(σi=constant).
Studied smart, not hard — thanks to [Link]
Note contd.:
6. The Incremental Transmission loss,

.IN
ITLi of an unit can be expressed in terms

C
of its MW output as under: Consider

N
PL =   PjBjkPk

SY
j k

U
VT=   PjBjkPk+ PjBjiPi
j k≠i j
=   PjBjkPk +  PjBjiPi + BiiPi2
j k≠i j≠i
Studied smart, not hard — thanks to [Link]
Note contd.:

.IN
Thus, PL=   PjBjkPk+  PiBikPk+  PjBjiPi+BiiPi2

C
j≠I k≠i k≠i j≠i

N
Thus,

SY
ITLi = dPL/dPi

U
= 0 +  BikPk +  PjBji + 2 BiiPi
VT
k≠i j≠i
= 2  PkBik
k
Studied smart, not hard — thanks to [Link]
Example-1:
The costs of two steam plants are (with P1 and

.IN
P2 in MW): F1 =(2.3P1+0.0062 P12 + 25)106 kCals/Hr.

C
F2 =(1.5P1+0.01 P22 + 35)106 kCals/Hr.

N
The calorific value of coal at plant 1 and plant 2 are

SY
4000 kCals/kg. and 5000 kCals/kg. The cost of coal is
Rs.55/- and Rs.65/- per Ton.

U
Find: (i) Incremental Fuel Rate-kCals/MWHr
VT
(ii) Incremental Fuel Cost-Rs./MWHr and
(iii) Incremental Production Cost -Rs./MWHr if
the cost of other items is 10% of the IFC/plant.
Studied smart, not hard — thanks to [Link]
Solution:
(i)Incremental Fuel Rate in kCals/MWHr

.IN
IFR1=dF1/dP1 = (2.3+0.0124P1)106 kCals/MWHr

C
IFR2=dF2/dP2 = (1.5+0.02P2)106 kCals/MWHr

N
SY
(ii)Incremental Fuel Cost in Rs./MWHr
IC1= [dF1/dP1 in kCals/MWHr] [cost of coal in Rs./ton]

U
[calorific value-1 in kg./kCals] 10-3
VT
= (2.3+0.0124P1)106 (55) (1/4000) (10-3)
= 31.625 + 0.1705 P1 Rs./MWHr.
Similarly, IC2= (1.5+0.02P2)106 (65) (1/5000) (10-3)
= 19.9 + 0.26 P2 Rs./MWHr.

Studied smart, not hard — thanks to [Link]

Solution contd. :
(iii)Incremental Production Cost in

.IN
Rs./MWHr if the cost of other items can be

C
taken as 10% of the incremental fuel cost/plant.

N
Effective value of IC are given by:

SY
IC1eff = 1.1 (IC1)

U
= 1.1 (31.625 + 0.1705 P1) Rs./MWHr.
VT
= 34.7878 + 0.1875 P2 Rs./MWHr.
IC2eff = 1.1 (IC2)
= 1.1 (19.9 + 0.26 P2) Rs./MWHr.
= 21.45 + 0.286 P2 Rs./MWHr.
Studied smart, not hard — thanks to [Link]
Example-2:
The incremental costs of a two

.IN
unit system are given by:

C
IC1 = (0.008 PG1 + 8.0)

N
IC2 = (0.0096 PG2 + 6.4)

SY
Find the incremental cost and the

U
distribution of loads between the two units
VT
for optimal operation for a total load of 1000
MW. What is this value if the same total load
is equally shared among the two units?
Studied smart, not hard — thanks to [Link]
Solution:
For the total load values of PT = 1000 MW,

.IN
if the load is shared equally among the two units then:

C
PG1 = 500 MW; PG2 = 500 MW

N
With λ1 = 12 Rs./MWHr and λ2 = 11.2 Rs./MWHr.

SY
(unequal lambda values)

U
Now, for optimal operation, we have as per EIC
VT
principle, the IC’s to be equal.
Thus, IC1=IC2; PT = P1+P2 = 1000 are the
equations to be solved for output power values.
Studied smart, not hard — thanks to [Link]
Solution contd. :
Thus, IC1 = 0.008 PG1 + 8.0

.IN
= IC2 = 0.0096 PG2 + 6.4

C
= 0.0096 (1000 - PG1) + 6.4

N
Solving, we get, PG1 = 454.54 MW

SY
PG2 = 545.45 MW
Further, λsystem is calculated using any one of the IC

U
equations as: λsystem = λ1 = λ2 = 11.64 Rs./MWHr.
VT
Thus, with λsystem = λ1 = λ2 = 11.64 Rs./MWHr., the total
load is optimally shared between the two units and the
operating cost would be at its minimum.
Studied smart, not hard — thanks to [Link]
Example-3:
The fuel costs in Rs./Hr. for a plant

.IN
of three units are given by:

C
C1 = (0.1P12+40P1+100)

N
C2 = (0.125P22+30P2+80)

SY
C3 = (0.15P32+20P3+150)

U
Find the IC and the distribution of loads between
VT
the units for optimal operation for a total load of
400 MW, if the max. and min. capacity limits for
all units are 150 MW and 20 MW respectively.
Studied smart, not hard — thanks to [Link]
Solution:
Consider the IC curves given by:

.IN
IC1= dCi/dPi = (2ciPi+bi) Rs./MWHr

C
= (0.20P1+40) Rs./MWHr Similarly,

N
IC2=dC2/dP2 = (0.25P2+30) Rs./MWHr and

SY
IC3=dC3/dP3 = (0.30P3+20) Rs./MWHr

U
For the total load of PT = 400 MW, for optimal
VT
operation, as per EIC principle, the IC’s are equal. i.e.,
IC1=IC2=IC3 and PT = P1+P2+P3 = 400 MW.
Also, the system lambda is given by:
λ={PD+(bi/2ci)}/ {(1/2ci)}i=1,2,3 = 63.78 Rs./MWHr.
Studied smart, not hard — thanks to [Link]
Solution contd. :
Using this common system

.IN
lambda, the MW output values of all the

C
three units are obtained from their IC curves as:

N
P1= 118.9 MW, P2= 135.12 MW and P3=91.90 MW.

SY
(All the values are found to be within their limits)

U
Thus, VT
with λsystem = λ1 = λ2 = λ3 =63.78 Rs./MWHr, the
total load is optimally shared between the 3 units
and operating cost would be the minimum.
Studied smart, not hard — thanks to [Link]
Example-4:
The ICs of a two unit system are given by:

.IN
IC1 = 0.008 PG1 + 8.0

C
IC2 = 0.0096 PG2 + 6.4

N
Find the incremental cost and the

SY
distribution of loads between the two units

U
for optimal operation for a total load of 900
VT
MW. Also determine the annual saving in
cost in optimal operation as compared to
equal sharing of the same total load.
Studied smart, not hard — thanks to [Link]
Solution:
For a load of PT = 900 MW, if load is shared

.IN
equally among the units then:

C
PG1=PG2=450 MW.

N
Now, for optimal operation, as per EIC principle, the

SY
IC’s are equal. i.e.,
IC1=IC2 and PT = P1+P2 = 900

U
are the equations to be solved for output power values.
VT
Thus, IC1=0.008PG1+8.0=IC2=0.0096(900-PG1)+6.4
Solving,
PG1=400 MW; PG2=500 MW. (λ=11.2 Rs./MWHr.)
Studied smart, not hard — thanks to [Link]
Solution contd. :
Further, Increase in cost by Unit 1

.IN
if it supplies 450 MW (equal sharing)

C
instead of 400 MW (optimal sharing) is:

N
C1= ∫ IC1 dPG1 = ∫ (0.008PG1+8) dPG1

SY
= |(0.004PG12+8PG1)|400450 = Rs.570/hr.

U
Similarly, Decrease in cost by Unit 2 if it supplies
VT
450MW (equal sharing) instead of 500 MW (optimal
sharing) is:
C2= ∫ IC2 dPG2 = ∫ (0.0096PG2+6.4) dPG2
= |(0.0096PG22+6.4PG2)|500450 = Rs. - 548/hr.
Studied smart, not hard — thanks to [Link]
Solution contd. :

.IN
Thus,
the net saving in cost in

C
optimal operation is given by:

N
Rs. 570 – 548 = Rs. 22/- per hour

SY
or
It is equivalent to an annual saving

U
VT
in cost of
(assuming continuous operation):
= Rs.(22)(24)(365)
= Rs. 1,92, 720/- PA.
Studied smart, not hard — thanks to [Link]
Example-5:
The fuel costs of a two

.IN
generator system are given by:

C
C1 =α1 + β1P1+ γ1P12; C2 =α2 + β2P2+ γ2P22

N
Where, β1=40, β2=30, γ1=0.1, γ2=0.125, and

SY
α1,α2 are constants. How will the load of 150 MW

U
be shared optimally between the two units? Also
VT
determine the saving in cost in Rs./Hr. in optimal
operation as compared to equal sharing of the
same total load.
Studied smart, not hard — thanks to [Link]
Solution:
Consider the IC curves:

.IN
IC1=dC1/dP1 = β1+ γ1P1 = 40+0.2P1 Rs./MWHr
IC2=dC2/dP2 = β2+ γ2P2 = 30+0.25P2 Rs./MWHr

C
For PT=150MW, if load is shared equally

N
SY
among the units then: PG1=PG2=75 MW
Now, for optimal operation, we have IC’s to be equal.

U
i.e., IC1=IC2 and PT=P1+P2=150
VT are the
equations to be solved for output power values.
Thus, IC1= 40+0.2P1= IC2= 30+0.25P2= 30+0.25(150-P1)
Solving, P1= 61.11 MW and P2= 88.89 MW
(λ = 52.222 Rs./MWHr.)
Studied smart, not hard — thanks to [Link]
Solution contd. :

.IN
Further, The increase in cost by Unit 1 if it
supplies 75 MW instead of 61.11 MW is:

C
C1= ∫ IC1 dPG1 = ∫ (40+0.2P1) dP1

N
= |(40P1+0.1P12)|61.1175 = Rs.737.344/hr.

SY
U
Similarly, decrease in cost by Unit 2 if it supplies
VT
75 MW instead of 88.89 MW is:
C2= ∫ IC2 dPG2 = ∫ (30+0.25P2) dPG2
= |(30P2+0.125P22)|88.8975 = Rs. - 707/hr.

Studied smart, not hard — thanks to [Link]

Solution contd. :

.IN
Thus,
the net saving in cost in

C
optimal operation is given by:

N
= Rs. 737.344 – 707

SY
= Rs. 30.344/- per hour
(or

U
VT
it is equivalent to a saving of:
= Rs. (30.3)(24)(365)
= Rs. 2,62,800/- PA)

Studied smart, not hard — thanks to [Link]

Example-6:

.IN
The fuel cost function in Rs./Hr. for
3-thermal unit plant is given by (with P’s in MW):

C
F1 = 350 + 7.20 P1 + 0.0040 P12

N
SY
F2 = 500 + 7.30 P2 + 0.0025 P22
F3 = 600 + 6.74 P3 + 0.0030 P32

U
Find the optimal schedule for a total load of 450 MW. Also
VT
compute the costs of operation for this schedule. Compare
the same when the three generators share the same total
load equally among them.

Studied smart, not hard — thanks to [Link]

Solution:

.IN
Consider the IC curves as under:
IC1=dF1/dP1 = 7.2+0.008P1 Rs./MWHr

C
N
IC2=dF2/dP2 = 7.3 + 0.005P2 Rs./MWHr and

SY
IC3=dF3/dP3 = 6.74 + 0.006P3 Rs./MWHr

U
For optimal operation, we have the common system
VT
lambda given by:
λ = {PD+(bi/2ci)}/ {(1/2ci)}  i= 1,2,3
= Rs. 8/ MWHr.
Studied smart, not hard — thanks to [Link]
Solution contd. :

.IN
Using λsystem, the MW output values of
all 3 units are obtained from their IC curves as:

C
N
P1= 100 MW, P2= 140 MW and P3=210 MW.

SY
The operating costs for this schedule are

U
found by using the cost curves as:
VT
FT(Optimal operation) = F1+F2+F3
= 1110+1571+2147.7 = 4828 Rs./Hr.

Studied smart, not hard — thanks to [Link]

Solution contd. :
Similarly, the operating costs for

.IN
equal sharing of total load are also

C
found by using cost curves as:

N
(with P1= P2= P3=150 MW).

SY
FT(Equal sharing) = F1+F2+F3

U
= 1520+1621.25+1078.5
VT
= 4849.5 Rs./Hr.
Thus, saving in cost in optimal operation is:
= 4849.5 – 4828 = Rs.21.75/- per hour.
Studied smart, not hard — thanks to [Link]
Example-7:

.IN
Given that
IC1=(40+0.2P1)

C
N
IC2 =(30+0.25P2)

SY
Calculate and tabulate the load shared by

U
each unit for optimal operation if the total
VT
load varies from 50 to 250MW, in steps of
50MW, given that [Link] is 125 and min.
MW is 20 for both the units.
Studied smart, not hard — thanks to [Link]
Solution:
Stage 1: Consider IC’s at Pmin:

.IN
IC1|P1=P1min = 20MW = 40+0.2(20) = 44 Rs./MWHr

C
IC2|P2=P2min = 20MW = 30+0.25(20) = 35 Rs./MWHr

N
Thus, IC2< IC1; i.e., EIC holds good only from the stage

SY
where, λsystem=44 Rs./MWHr.
Now find P2 corresponding to this λ:

U
P2=(44–30)/0.25=56MW so that PTotal=20+56=76 MW
VT
Thus, until PT=76 MW, EIC will not be feasible, Unit 1
will work at its min. load of 20MW and all the additional
load is shared by unit 2 alone till λ=44 Rs./MWHr.
Studied smart, not hard — thanks to [Link]
Solution contd.:
Stage 2: Consider IC’s at Pmax:

.IN
since the Unit 2 is expected to reach its max. limit

C
earlier, find: IC2|P2=P2max = 125MW = 30+0.25(125)

N
= 61.25 Rs./MWHr .

SY
And P1|λ1=61.25 = (61.25 – 40)/0.2 = 106.25 MW
so that then PTotal = 106.25+125= 231.25 MW

U
VT
Thus, after PT=231.25MW, EIC ceases to hold good;
Unit 2 will work at its maximum load sharing of 125
MW only and all the additional load variations are
shared by unit 1 alone untill P1 also reaches 125 MW.
Studied smart, not hard — thanks to [Link]
Solution contd. :
Stage 3: In summary, EIC holds good only for

.IN
PT and λsystem values which satisfy the limits:

C
76 ≤ PT ≤ 231.25 MW and

N
44 ≤ λsystem ≤ 61.25 Rs./MWHr

SY
For the total load values of PT = 100, 150 and 200 MW,

U
the equations to be solved for the output power values
VT
are: IC1=IC2; PT = P1+P2 and λsystem is calculated using
any one of the IC equations. The values so obtained for
the said range of load values are tabulated as under.
Studied smart, not hard — thanks to [Link]
Solution contd. :

.IN
PT P1 P2 λsystem Remarks
50 20 30 44 Unit 2 only shares load

C
N
76 20 56 44 Unit 1 and Unit 2

SY
100 33.33 66.67 46.67 share total load as per
150 61.11 88.99 52.22 EIC Criterion. The

U
88.88 111.11 57.78 system works with a
200
VT
common system λ
231.25 106.25 125 61.25
250 125 125 61.25 Unit 1 only shares load

Studied smart, not hard — thanks to [Link]

Example-8:

.IN
Given that

C
F1=110+30P1+0.09P12 12≤P1≤125 MW

N
SY
F2=135+12P2+0.1P22 25 ≤P2≤125 MW
Calculate and tabulate the load shared by
U
VT
each unit for optimal sharing of the total
load in the range 50-250 MW in steps of
100 MW. Also find λsystem in each case.
Studied smart, not hard — thanks to [Link]
Solution:
Consider IC’s:

.IN
IC1=dC1/dP1 = 30+0.18 P1 Rs./MWHr.
IC2= dC2/dP2 = 12+0.2 P2 Rs./MWHr.

C
N
Stage 1: Consider the IC’s at Pmin:

SY
IC1|P1=P1min = 12MW = 32.16 Rs./MWHr
IC2|P2=P2min = 25 MW = 17 Rs./MWHr

U
Thus, IC2<IC1; EIC is valid only from λ=32.16 Rs./MWHr. Now
VT
find P2|λ2=32.16 = 100.8 MW so that then PTotal = 112.8 MW
Thus, until PT=112.8 MW, EIC is infeasible, Unit 1 is at its
minimum load of 12 MW and the additional load is shared by
unit 2 alone till λ=32.16 Rs./MWHr.
Studied smart, not hard — thanks to [Link]
Solution contd.:
Stage 2: Consider the IC’s at Pmax:

.IN
Since the Unit 2 is expected to reach its max.

C
limit earlier, find:

N
IC2|P2=P2max = 125MW = 37 Rs./MWHr

SY
P1|λ1=37 = 38.88 MW so that then
PTotal = 163.88 MW

U
VT
Thus, after PT= 163.88 MW, EIC ceases to hold
good; Unit 2 will work at its max. load of 125 MW
and all the additional load is shared by unit 1
alone until P1 also reaches 125 MW.
Studied smart, not hard — thanks to [Link]
Solution contd. :
Stage 3: In summary, EIC holds good only for

.IN
PT and λsystem values which satisfy the limits:

C
112.8 ≤ PT ≤ 163.88 MW and

N
32.16 ≤ λsystem ≤ 37 Rs./MWHr

SY
For the total load value of PT = 150 MW, the equations

U
to be solved for the output power values are: IC1=IC2;
VT
PT = P1+P2 and λsystem is calculated using any one of the
IC equations. The values so obtained for the said range
of load values are tabulated as under.
Studied smart, not hard — thanks to [Link]
Solution contd. :

.IN
PT P1 P2 λsystem Remarks

C
50 12 38 32.16 Unit 2 only shares load

N
SY
112.8 12 100.8 32.16 Unit 1 & Unit 2
Share total load as

U
150 31.58 118.42 35.67
VT per EIC Criterion.
System works with a
163.88 38.88 125 37 common system λ
250 125 125 37 Unit 1 only shares load

Studied smart, not hard — thanks to [Link]

Example-9:

.IN
A system is fed by two plants with IC functions as under:
IC1= 28+0.16P1; IC2= 20+0.25P2 in Rs./MWHr

C
The max. & min. loads on the units are 100 MW & 10 MW

N
respectively. Determine the min. cost of generation for

SY
supplying load as under based on EIC criterion:

U
VT
Load 12Midnight 6 am– 12noon- 2-6 6-9 9pm–
Duration –6am 12noon 2 pm pm pm 12Midnight
Load in
MW 60 100 80 120 175 50

Studied smart, not hard — thanks to [Link]

Solution:
Stage 1: Consider the IC’s at Pmin:

.IN
IC1|P1=P1min = 10MW = 29.6 Rs./MWHr

C
IC2|P2=P2min = 10 MW = 22.5 Rs./MWHr

N
Thus, IC2<IC1; i.e., EIC holds good only from

SY
system lambda equal to 29.6 Rs./MWHr.
Now find P2 corresponding to this Lambda:

U
P2|λ2=29.6 = 38.4 MW so that then PTotal = 48.4 MW
VT
However, all the required loads to be supplied
are above this total load of 48.4 MW!
Studied smart, not hard — thanks to [Link]
Solution contd.:

.IN
Stage 2: Consider the IC’s at Pmax:
IC1|P1=P1max = 100MW = 44 Rs./MWHr

C
IC2|P2=P2max = 100MW = 45 Rs./MWHr

N
SY
P2|λ1=44 = 96 MW so that then
PTotal = 196 MW

U
VT
Again, it is observed that all the
required loads to be supplied are
below this total load of 196 MW!
Studied smart, not hard — thanks to [Link]
Solution contd. :
Stage 3: In summary, EIC holds

.IN
good for all the load values specified.

C
N
Now for the total load value of PT = P1+P2, where EIC is

SY
shown to holds good, the equations to be solved for the
output power values are: IC1=IC2; PT = P1+P2 and λsystem

U
is calculated using any one of the IC equations.
VT
The values so obtained for the said range
of load values are tabulated as under:
Studied smart, not hard — thanks to [Link]
Solution contd. :
PT P1 P2 λsystem

.IN
48.4 10 (min.) 38.40 29.60

C
50 10.97 39.03 29.75

N
60 17.07 42.13 30.73

SY
80 29.27 50.73 32.68

U
100 41.47
VT 58.53 34.63
120 53.66 66.34 38.59
175 87.20 87.80 41.95
196 100 (max.) 96.00 44.00
Studied smart, not hard — thanks to [Link]
Example-10:

.IN
The fuel input in Btu/Hr. for unit 1 & 2 :
F1={P1+0.024P12+80}106; F2={6P2+0.04P22+120)106

C
The max. & min. loads on the units are 100 MW & 10

N
MW respectively. Find the min. cost of generation for

SY
supplying a load as under

U
with fuel cost at Rs.2.0 per Mbtu:
VT
Load 12 Midnight– 6 pm– 12
6 am–6 pm
Duration 6 am Midnight
Load 50 MW 150 MW 50 MW
Studied smart, not hard — thanks to [Link]
Solution:
Consider the incremental cost curves given by:

.IN
IC1=dC1/dP1 = 2+0.096 P1 Rs./MWHr
IC2=dC2/dP2 = 12+0.16 P2 Rs./MWHr

C
Consider the IC’s at Pmin:

N
SY
IC1|P1=P1min = 10MW = 2.96 Rs./MWHr
IC2|P2=P2min = 10 MW = 13.6 Rs./MWHr

U
Thus, IC2 >IC1; i.e.,
VT
EIC holds good only from λ=13.6 Rs./MWHr.
Now find P1 corresponding to this Lambda:
P1|λ1=13.6 = 120.833 MW (>P1max)
So that then PTotal = 130.833 MW.
Studied smart, not hard — thanks to [Link]
 Solution contd.

.IN
Thus, until PT=130.833 MW, EIC
will not be feasible, Unit 2 will work

C
at its minimum load of 10 MW and the

N
additional load is shared by unit 2 alone till λ=13.6

SY
Rs./MWHr. However, this is not further feasible since
the unit 1 reaches its max. value of 100 MW within this

U
VT
range!!! Hence, EIC ceases to exist for any given range
of load. The load sharing is thus constrained by the
MW limits for the various loads specified:
50 MW and 150 MW, as shown by the table below:
Studied smart, not hard — thanks to [Link]
Solution contd. :
PT MW P1 MW P2 MW Remarks

.IN
<20 Infeasible

C
20 10 10

N
Here unit 2 is at its min. limit of 10
50 40 10

SY
MW, only unit 1 shares the load till it
100 90 10
hits its own max. limit of 100 MW

U
110 100 10 and unit 1 determines the system λ
VT
150 100 50 Here unit 1 is at its max. limit of 100
180 100 80 MW, only unit 2 shares the load and
200 100 100 determines the system λ
>200 Infeasible
Studied smart, not hard — thanks to [Link]
EXAMPLE 11

.IN
C
N
SY
U
VT

Department
Studied ofhard
smart, not EEE,—GSSSIETW, Mysuru
thanks to [Link]
 .IN
EXAMPLE 12

C
N
SY
U
VT

Department
Studied ofhard
smart, not EEE,—GSSSIETW, Mysuru
thanks to [Link]
 .IN
C
N
SY
U
VT

Department
Studied ofhard
smart, not EEE,—GSSSIETW, Mysuru
thanks to [Link]
EXAMPLE 13

.IN
C
N
SY
U
VT

Department
Studied ofhard
smart, not EEE,—GSSSIETW, Mysuru
thanks to [Link]
 .IN
C
N
SY
U
VT

Department
Studied ofhard
smart, not EEE,—GSSSIETW, Mysuru
thanks to [Link]
EXAMPLE 14

.IN
C
N
SY
U
VT

Department
Studied ofhard
smart, not EEE,—GSSSIETW, Mysuru
thanks to [Link]
 .IN
C
N
SY
U
VT

Department
Studied ofhard
smart, not EEE,—GSSSIETW, Mysuru
thanks to [Link]
In general, for a system of n units we have,
PL=  Pi Bij Pj Where,

.IN
i j
Bij=NkiNkjRk cos(σi-σj)/(ViVjcosicosj)

C
N
SY
U
VT

Department
Studied smart, of EEE,—
not hard GSSSIETW,
thanksMysuru
to [Link]
EXAMPLE 15

.IN
C
N
SY
U
VT

Department
Studied ofhard
smart, not EEE,—GSSSIETW, Mysuru
thanks to [Link]
 .IN
C
N
SY
U
VT

Department
Studied ofhard
smart, not EEE,—GSSSIETW, Mysuru
thanks to [Link]
 .IN
C
N
SY
U
VT

Department
Studied ofhard
smart, not EEE,—GSSSIETW, Mysuru
thanks to [Link]
 .IN
C
N
SY
U
VT

Department
Studied ofhard
smart, not EEE,—GSSSIETW, Mysuru
thanks to [Link]
 .IN
C
N
SY
U
VT

Department
Studied ofhard
smart, not EEE,—GSSSIETW, Mysuru
thanks to [Link]
EXAMPLE 16

.IN
C
N
SY
U
VT

Department
Studied ofhard
smart, not EEE,—GSSSIETW, Mysuru
thanks to [Link]
 .IN
C
N
SY
U
VT

Department
Studied ofhard
smart, not EEE,—GSSSIETW, Mysuru
thanks to [Link]
 .IN
C
N
SY
U
VT

Department
Studied ofhard
smart, not EEE,—GSSSIETW, Mysuru
thanks to [Link]
 .IN
C
N
SY
U
VT

Department
Studied ofhard
smart, not EEE,—GSSSIETW, Mysuru
thanks to [Link]
 .IN
C
N
SY
U
VT

Department
Studied ofhard
smart, not EEE,—GSSSIETW, Mysuru
thanks to [Link]