Module-7

Distribution System Loss

Reduction & Voltage Profile
Improvement
A.K. Mishra
IOE, Nepal

10/30/24 AKM/distplng/Distribution loss 1

reduction
Parameters effecting Distribution
Feeder Loss and Voltage drop
• Let’s consider a distribution feeder branch

Vk+1 Vk
Rj Xj
Pj+jQj

• The branch Feeder loss 2 2

(Technical) i.e. I2R Loss
Pj  Q j
2
Rj
is given by: Vk
10/30/24 AKM/distplng/Distribution loss 2
reduction
 And the Voltage drop is given by:
Vk 1 Vk  IZ
Pj  jQ j
Vk  * ( R j  jX )
Vk
Pj R j Q j X j  Pj X j Q j R j 
Vk  *  *  j *  * 
Vk Vk  kV V k 
Pj R j Q j X j
Vk  *  *
Vk Vk
• In case of transmission line, the line carry same P and
Q over a longer distance and just for a line thus is very
easy to compute.
 In a distribution line (whether a primary or secondary)
The P and Q along the feeder do not remains the
same. Hence the loss even for a single feeder is
represented as:
10/30/24 AKM/distplng/Distribution loss 3
reduction
 2 2
Pj  Q j
 j Vk
2
Rj

From the above expressions for the minimum

Loss and voltage drop:
– Pj to be minimum
– Qj to be minimum
– Rj to be minimum
– Xj to be minimum
– Vk to be maximum
• Fortunately all of these three are possible for
distribution system.
10/30/24 AKM/distplng/Distribution loss 4
reduction
Branch power to be minimum
• For minimum loss and voltage drop;
– As far as possible, the branch Real and Reactive power
flow for each branch should be minimum
– May not be possible in all cases
– At least the considerations to the branches carrying
relatively higher powers.
• This can be achieved by
– Optimum configuration (for planning decisions)
– Reinforcement (for existing system)
– Reconfiguration (operational decisions)
– Addition of Real and reactive powers sources along the
distribution feeder (Distributed generator/Reactive
compensation)

10/30/24 AKM/distplng/Distribution loss 5

reduction
Let's consider five different distribution transformers of same size and
same number of customers are desired to be connected from the
substation through a radial distribution feeder of similar conductor. Out of
practically feasible feeder layout options two are shown in Fig.(a) & (b).
2 km B C
B C
2 km 2 km
2 km 2 km
2 km A 2 km
2 km A S/S D
S/S D

2 km 2 km

E E
(a) Option-A (b) Option-B
– The load to be served in both the cases same
– Conductor length same
• But it can be shown that the losses in option – A is about 1.5
times greater than option-B
10/30/24 AKM/distplng/Distribution loss 6
reduction
Suppose the transformer size at each of the node is S kVA.
Then the power flow through the branches can be easily
computed and is shown in Fig. below.

2 km
B 3S C B C

S
2 km 4S 2 km 2S 2 km 2 km S
2 km A 2 km A 2 km
S/S D S/S D
5S 5S 3S
S
2 km S
2 km

E
(a) Option-A E
(b) Option-B

LossA 52  42  32  22  1 S 2
 2 2 1.486486
LossB 5  1  3  1  1 S
2 2 2

10/30/24 AKM/distplng/Distribution loss 7

reduction
• It is to be noted that this difference in losses are only due to
the fact that the branch power to serve the same load are
significantly different in two options.
• Alternatively it can be said that the node powers have to
travel larger distance in option-A than option-B.
Summarized in Table below:

Node Traveling distance (km) The idea is;

Option-A Option-B
A 2 2 To configure the feeder in
B 4 4 such a way that the distance
C 6 6
D 8 4
of all the load points to
E 10 6 source should be minimum.

10/30/24 AKM/distplng/Distribution loss 8

reduction
Different nodal demands and conductors
• If the load at different nodes are different then;
 Li S i Should be minimum.
• Where:
– Si Demand at ith Node
– Li distance of ith Node from S/S
• If the different conductors are in the path;
Li  l j rj
j
• Where:
– j represents the branch in ith path and
– lj length of jth branch
– and rj p.u. length resistance of jth branch
10/30/24 AKM/distplng/Distribution loss 9
reduction
 Example
• Eight distribution load centers are to be electrified by
taking a power supply from a 33/11 kV Area-sub
station using a 11 kV primary distribution feeder.
The three physically viable options for the primary
distribution feeders are given in figure next slide.
• The peak load at all the load centers are 200 kVA at
0.9 p.f.
1st case:
Assuming all the conductors are Dog in all the three
options.
2nd case
– Option-A: main line Dog and laterals Weasel
– Option-B: All rabbits &
– Option-C SS-T1, T1-T2 & T2-Tt Dog rest Rabbit.
10/30/24 AKM/distplng/Distribution loss 10
reduction
 T1 T2 T3 T4
5 km 5 km 5 km 5 km
S/S

2 km 2 km 2 km 2 km

T5 T6 T7 T8

Option-A

T1 T2 T3 T4
5 km 5 km 5 km 5 km
S/S

2 km

T6 T7 T8
T5 5 km 5 km 5 km

Option-B

T1 T2 T3 T4
5 km 5 km 2.5 km 2.5 km 5 km
S/S

T7 T8
T5 5 km T6 5 km
2.5 km 2 .5km

Option-C

10/30/24 AKM/distplng/Distribution loss 11

reduction
For all the six options in Problem compute the net present worth of the
capital investment and energy loss and hence suggest which option you will
prefer with reason. Consider the following;
 The conductor details for 11 kV

Conductor R(Ω)/km X(Ω)/km Construction

cost/km
in USD
Dog 0.3349 0.315
5450
Rabbit 0.6648 0.335
4300
Weasel 1.1122 0.345
3840
 Current Energy cost at 33/11kV S/S. 8 cent and increases by 5%
each year
 Assume Annual LLF 0.25
 discount rate 10% (expected rate of return)
 Project life 20 years
10/30/24 AKM/distplng/Distribution loss 12
reduction
 Load Flow solution
Voltage (kV) Option A Voltage (kV) Option B Voltage (kV) Option C
Bus All Dog Dog+Weasel All Dog All Rabbit All Dog Dog+Weasel
No. i ii i ii i ii
1 10.64 10.63 10.62 10.33 10.66 10.66
2 10.37 10.37 10.48 10.10 10.36 10.36
3 10.20 10.20 10.40 9.95 10.19 10.19
4 10.12 10.11 10.31 9.80 10.15 10.15
5 10.62 10.59 10.56 10.02 10.08 10.04
6 10.36 10.33 10.43 9.99 10.13 10.08
7 10.18 10.15 10.35 9.85 10.13 10.08
8 10.10 10.07 10.30 9.77 10.08 10.04

Avearage Yearly
Loss Option A Option B Option C
All Dog Dog+Weasel All Dog All Rabbit All Dog Dog+Weasel
Active Loss watt
peak 69080.12565 66816.3212 55632.4921 123791.6757 74266.91897 76552.02617
Yearly Energy
Loss 453856.4255 438983.2303 365505.473 813311.3095 487933.6576 502946.812

10/30/24 AKM/distplng/Distribution loss 13

reduction
 Distribution Line capital cost Computation

Conductor Cost in
Dog Weasel Rabbit USD
Options
Option A (i) Km 28 152600
Option A (ii) Km 20 8 139720
Option B (i) Km 37 201650
Option B (ii) Km 37 159100
Option C (i) Km 37 201650
Option C (ii) Km 14.5 22.5 175775

10/30/24 AKM/distplng/Distribution loss 14

reduction
 Total present worth cost in USD
Energy loss cost 483750 467900 389580 866880 520070 536070
(energy+Construction) 636350 607620 591230 1025980 721720 711850

• Highest Construction Cost is obtained for Option B (i)

All dog Configuration
• Lowest Construction Cost is obtained for Option A (ii)
• Regarding the Construction Cost Option A (ii) is best
• The Best Option for the Construction of the project
would be Option B (i) All dog conductor configuration

10/30/24 AKM/distplng/Distribution loss 15

reduction
Reinforcement
A Primary distribution network shown below the load at each
buses has increased significantly that from the planning and
hence the losses are high.
S/S
5 km T1 5 km T2 5 km T3 5 km T4
The load at each buses are
2 km 200 kVA , 0.9 pf lagging and
T5 5 km T6 5 km T7 5 km T8
all conductors are rabbit
• It is desired to explore the viable options to improve the voltage
profile and loss reduction
• The idea is technically viable and economically best
• Out of many options one option could be; Disconnecting the T1-
T5 line at T1 and connecting T1 to the S/S with a new Feeder.
(reducing P &Q between S/S-T1)
• Second option could be change the conductor between S/S to
T1 (reducing resistance)
10/30/24 AKM/distplng/Distribution loss 16
reduction
For both of the above option compute the following
1. Minimum voltage at the feeder
2. Power Loss reduction and energy loss reduction with
annual LLF is 0.2
3. Saving in energy loss.
4. Annual saving if energy loss current rate is 7 cent per unit
and cost of conductors:

Dog 5450 USD/km

Rabbit 4300 USD/km

Assume the suitable cost for a new feeder and the life for
reinforcement is 20 yrs.

10/30/24 AKM/distplng/Distribution loss 17

reduction
Without Reinforcement
The load flow results:
Total peak power loss in the conductor is obtained as = 8, 73,150
Watt

Energy loss in kWh (in one year) =

peak powe loss * loss load factor ( LLF ) * time in one year
kWh
1000

LLF= 0.2

Total energy loss in one year = 15, 29,800 kWh

10/30/24 AKM/distplng/Distribution loss 18

reduction
After Reinforcement
Case1
Changing the conductor between substation to T1 by Dog
conductor

Total system peak power loss = 5, 28,000 Watt

Total energy loss in one year = 9, 25,060 kWh

Case2
Disconnecting the T1-T5 line at T1in the given feeder and connecting
T1 to the substation by a rabbit conductor and with new feeder

Total system peak power = 5, 99,290 Watt

Given LLF= 0.2
Total energy loss in one year = 10, 49,900 kWh
10/30/24 AKM/distplng/Distribution loss 19
reduction
Saving in option-1
With Dog conductor between sub-station to T1, the Present worth for new
conductor placement= length in km * cost of conductor/km
PW=5450*5=27250 USD
Life of conductor (N) = 20 years
Interest rate (i) = 10%

Annual worth (AW) = PW

1  i * i
N

 1  i N  1
1
AW=27250*1.120 *0.1*
1.120  1
Annual worth = 3200.82 USD
Annual saving replacing rabbit conductor between substation to T1 by Dog
conductor = 0.07*(15, 29,800-9, 25,060)-3200.82= 39, 130.98 USD

Hence annual saving= 39, 130.98 USD

10/30/24 AKM/distplng/Distribution loss 20

reduction
Saving option-2
Cost of rabbit conductor when two feeder are separated from substation to
T1 = 5* 4300=21500 USD
Cost including cost of circuit breaker, relay, lighting arrestor, CT, PT
=12, 000 USD
Total present worth (PW) = 12, 000+21500=33, 500 USD

Interest rate (i)=10%

For life of new feeder 20 year

Annual worth (AW) = PW

1  i N * i
1  i N  1
Annual worth = 3, 934.90 USD
Annual saving
= 0.07*(15, 29,800 -10, 49,900) - 3, 934.90 = 29, 658.10 USD

Hence annual saving= 29, 658.10 USD

10/30/24 AKM/distplng/Distribution loss 21
reduction
Reconfiguration (Non Automated system)
The figure below shows the single line diagram of two
feeders. The feeders are equally loaded for 21 hours each
day but feeder A has 3 hrs peaks which occur when
feeder B is off peak. The loading under peak in normal
configuration is shown in Fig.2.
10 km

200 A Feeder B

Normally closed

5 km
Normally open

600 A Feeder A

10 km
Fig.2

10/30/24 AKM/distplng/Distribution loss 22

reduction
If an additional contact is installed at the 10 km of feeder
A, the peak loading of the feeder could be reconfigured as
shown in Fig.3. Compute the annual saving. Use following
data:
Assume feeder loading is uniformly distributed
Feeder resistance: 0.4 Ω km/phase
Cost of additional contact: USD 40,000
Discount factor : 10%
Life: 20 years
LLF for peak 3 hrs: 0.8
Cost of energy: 7 cent per unit
10 km
400 A Feeder B
Normally closed
5 km
Normally open

400 A Feeder A
10 km
Fig.3

10/30/24 AKM/distplng/Distribution loss 23

reduction
SOLUTION

In case -1,
Feeder –A loading = 600/15 A/Km
= 40 A/Km
Feeder -B loading = 200/10 A/Km
= 20 A/Km

In case -2
Feeder –A loading for 10 Km = (400/10) A/Km
= 40 A/Km
Feeder –B for 10 Km equal to Loading in case -1 ie 20 A/km,
remaining 5 km should be 40 A/km ie 200 A

10/30/24 AKM/distplng/Distribution loss 24

reduction
Loss computation
For Case-1
Resistance per phase per Km =0.4 ohm
2 2
Per phase Power Loss (PL1) = ⅓[IA rLA+ IB rLB]
=⅓[(600)2*0.4*15+(200)2*0.4*10]
= ⅓*2320 kW

Three phase Power Loss (PL1) = 3*⅓*2320

= 2320 kW

10/30/24 AKM/distplng/Distribution loss 25

reduction
For Case-2
Per phase Power Loss (PL2)
= ⅓(I2 +I' 2 +I*I' )*rL1+ ⅓(I')2*(L-L1)*r+⅓[IA2rLA]
=⅓[(400)2+(200)2+400*200]*0.4*10] + [⅓*(200)2*0.4*5+⅓*(400)2*0.4*10]
= ⅓*[1200+640] kW

Three phase Power Loss (PL2) = 3*⅓*[1200+640] kW

= 1840 kW

Saving in peak power Loss = (2320-1840) kW

= 480 kW

Energy Saving per year = LLF *480*3*365 kWh

= 420480 kWh

10/30/24 AKM/distplng/Distribution loss 26

reduction
Rate of energy = Cent 7 /kwh
Total amount saving per year(Rs.) due to loss reduction
=Rate per Kwh*Saving kwh
= USD 29,433.60

Cost of Additional contact(PV) =USD 40,000

Life of contact(n) = 20 yrs
Discount rate(γ) =10 %

Converting PV into Annual cost

PV
M= = USD 4,698.38
1 1
[  ]
 (1  ) n

Net annual Saving (Rs) = Annual Saving due to energy loss –Annual cost
= USD 24,735.22

10/30/24 AKM/distplng/Distribution loss 27

reduction
Reactive Power Flow:
For the 11 kV Primary distribution Network shown below, determine the size
(kVAR) of the capacitor to be placed at the location shown in diagram to achieve
the maximum loss reduction. For simplifying the analysis assume voltage at each
node is 1 p.u.
All impedances are in p.u. at 1000 kVA.

S 1+2j
S 1+2j 1+2j
1000 kVA
33/11kV 100 kVA 100 kVA 100 kVA
0.8 p.f. lag 0.8 p.f. lag 0.8 p.f. lag

In fact the system is so simple that no any complex calculation is required and
the answer is ………….kVAR.
10/30/24 AKM/distplng/Distribution loss 28
reduction
 We have already discussed that capacitor can change only the
reactive power flow in the Feeder and the idea is minimize the
 Li Qi for minimum Loss.
A B C
S
S 1+2j 1+2j 1+2j

1000 kVA
33/11kV 100 kVA 100 kVA 100 kVA
0.8 p.f. lag 0.8 p.f. lag 0.8 p.f. lag
• The Source of reactive power are shown by arrow and
demand of reactive power is at A, B and C.
• So for Minimum loss it is very easy to say that reactive
power for B and C should come from Capacitor and for A it
should from S/S.
• Hence the Size of the Capacitor for Minimum Loss is 120
kVAR.
10/30/24 AKM/distplng/Distribution loss 29
reduction
Optimal Capacitor Placement

• Capacitors are often installed in distribution

system for reactive power compensation to carry
out
• power and energy loss reduction,
• voltage regulation improvement
• system capacity release, and
• system security improvement .

10/30/24 AKM/distplng/Distribution loss 30

reduction
• Capacitor can reduce the loss mainly due to
reactive component of line (branch) flow.
• Ideally speaking capacitors should be placed at
each and every load center so that the losses
due to reactive component of line flow is zero
• Practically it is not possible due to higher
installation cost
• Since the demand at load centers vary with time,
requires not only the capacitor to be located but
should be controlled requiring higher
accessories cost as well

10/30/24 AKM/distplng/Distribution loss 31

reduction
Capacitors at HT or LT
• If capacitors are placed at LT
– Loss will be minimum at both HT and LT
– Requires switching operations necessitating expensive
switchgear
• The provision of LT capacitors on the individual
services usually having low power factor is
generally preferred.
• The capacitors are placed normally on HT only at
some optimal locations with optimal sizing and
control switch gear.

10/30/24 AKM/distplng/Distribution loss 32

reduction
The optimal capacitor placement is the determination
of the location and size of the capacitors to maximize
the saving i.e. maximization of equation:
ncap
S = kp * LP + Ke* LE - Σ (Kcf+Kc*Qci0) .........(1)
i=1

Where:
Kp present worth of a kW of power
Ke present worth of a kWhr of energy
LP reduction of peak power loss because of compensation
LE reduction of energy losses because of compensation
Kcf Fixed cost of Capacitor placement
Kc marginal costs of capacitors per kVAR
10/30/24 AKM/distplng/Distribution loss 33
reduction
• In fact, it is required to obtain the number of optimal
capacitor locations their locations and size of the capacitors
simultaneously to give the best result.
• We can think the size of the problem if we consider the
each possible options.
• Some advance techniques such as AI based techniques
(ANN, GA) have been applied which are capable to give the
global optimum solution.
• Several classical approach have also been used which (all
of these) try to give the optimum solution as far as possible
minimizing the computations.
• This of course make some valid assumptions to simplify the
problem
• Here we are discussing one of such classical iterative
procedure to determine the optimal location and sizes
10/30/24 AKM/distplng/Distribution loss 34
reduction
The general mathematical expression for saving can be re-written as
ncap
S= kp * (LPd + LPq) + Ke* (LEd + LEq) - Σ (Kcfi+Kci*Qci0)
i=1
Where:
LPd reduction of d-component of peak power loss because of compensation
LPq reduction of d-component peak power loss because of compensation
Leq reduction of energy losses due to d-component power loss because of compensation
Leq reduction of energy losses due to q-component power loss because of compensation

With assumption that in capacitor problem bus voltage is constant at operational

voltage The expression for saving can be simplified as
ncap
S= kp * LPq + Ke* LEq - Σ (Kcfi+Kci*Qci0)
i=1

And Reactive component of current is the indicative of q-axis current Iq

10/30/24 AKM/distplng/Distribution loss 35

reduction
• The overall procedure is further simplified first finding the
expression for power loss reduction for a set of capacitors
located optimally.
• Later an iterative procedure is employed to update the
locations one by one.
• The number of capacitor locations are then identified for
which the overall saving is maximum this is possible with
the inclusion of fixed cost in objective function
• Before going to the detail procedure first consider the
correlation of loss reduction with:
– Capacitor(s) location
– Capacitor(s) Size
• To understand this, Let’s consider a section of distribution
system

10/30/24 AKM/distplng/Distribution loss 36

reduction
 4

Ic1 r3
Iq3-Ic1
Iq1-Ic1-Ic2-Ic3 Iq2-Ic2-Ic3 Iq4-Ic2
S/S
S/S 1 r1 2 r2 3 r4 5

Ic3 Ic2

Branch Sending end Receiving end Resistance Branch

No. node node () reactive
Load currents
1 1 2 r1 Iq1
2 2 3 r2 Iq2
3 2 4 r3 Iq3
4 3 5 r4 Iq4

10/30/24 AKM/distplng/Distribution loss 37

reduction
The expression of Power loss reduction can be written as:
LPq =loss without capacitor-loss with capacitor
4
LPq = Σ I2qiri-{r1(Iq1-Ic1-Ic2-Ic3)2+r2(Iq2-Ic3-Ic2)2+r3(Iq3-Ic1)2+r4(Iq4-Ic2)2}
i=1

4
= Σ I2qi ri –{r1[I2q1-2Iq1(Ic1+Ic2+Ic3)+( Ic1+Ic2+Ic3)2]+r2[I2q2-2Iq2(Ic2+Ic3)
i=1
+ (Ic2+Ic3)2]+r3[I2q3-2Iq3*Ic1+I2c1]+r4[I2q4-2Iq4Ic2+I2c2]}

= -{r1( Ic1+Ic2+Ic3)2 +r2(Ic2+Ic3)2+r3(Ic1)2+r4(I2c2)}+2{r1Iq1(Ic1+Ic2+Ic3)

+r2Iq2(Ic2+Ic3)+r3Iq3(Ic1)+r4Iq4(Ic2)}

= -{(r1+r3)I2c1+(r1+r2+r4)I2c2+(r1+r2)I2c3+2r1(Ic1Ic2+Ic1Ic3) +2(r1+r2)Ic2Ic3}
+2{[(r1Iq1+r3Iq3)Ic1+(r1Iq1+r2Iq2+r4Iq4)Ic2+(r1Iq1+r2Iq2)Ic3]}
10/30/24 AKM/distplng/Distribution loss 38
reduction
The expression for power loss reduction In
the matrix form can be written as;

= -[Ic1 Ic2 Ic3] r1+r3 r1 r1 Ic1

r1 r1+r2+r4 r1+r2 Ic2
r1 r1+r2 r1+r2 Ic3

Ic1
+2[(r1Iq1+r3Iq3) (r1Iq1+r2Iq2+r4Iq4) (r1Iq1+r2Iq2)] Ic2

Ic3

= -IcT H Ic +2bT Ic

10/30/24 AKM/distplng/Distribution loss 39

reduction
• The square matrix H and vector b can be easily be found
from known feeder configuration and capacitor location
Hii= sum of branch resistances from source to i th capacitor location
Hij= sum of branch resistances from source to the node where
i th capacitor location and jth capacitor
locations diverges
bi= sum of the product of reactive component of current and branch
resistances from source to ith capacitor location

• The size of H matrix is ncap x ncap and the size of the b

vector is also ncap
• For maximum Loss reduction, the derivative of LPq w.r.t. Ic
should be zero. This Results the size of capacitors as:
Ic(optm) = H-1b
Thus the peak power loss reduction with this capacitor power
Loss reduction can be calculated as:
10/30/24 LPq= - Ic(optm) T
H Ic(optm) +2b
AKM/distplng/Distribution loss Ic(optm) 40
reduction
The overall procedure
• There are mainly two schemes that are repeated successively
and the process stops when the choice of locations can not
contribute further improvement.
Scheme-I:
• In this scheme, for a known set of location, capacitor sizes are
found for maximum loss reduction i.e.
Ic = H-1b
Scheme-II:
• In this scheme, relocations and sizing of one capacitor is done
assuming that the rest of the (ncap – 1) capacitors are already
placed and sized.
• If suppose kth capacitor is to be relocated and sized, the
problem reduces to placement and sizing of a single capacitor.
• For this first the load current distribution is modified to consider
the ncap-1 capacitors are negative loads
• Then for each possible locations of kth capacitor the optimal
size Ici = bi/Hi.
10/30/24 AKM/distplng/Distribution loss 41
reduction
1. Read system parameters and assume a suitable
number of capacitors and their initial set of
locations.
2. Apply scheme-I to get the sizes.
3. Set iteration number 1.
4. Perform scheme-II to relocate each capacitor for
each capacitor locations individually (i.e. ncap
times)
5. Apply scheme-I to improve the sizes for
complete set of location and compute the power
loss reduction.
6. Repeat steps 4 & 5 till the choice of locations
can not contribute further improvement.
10/30/24 AKM/distplng/Distribution loss 42
reduction
Optimal Number of capacitor locations
• The procedure described till now is for known number of
capacitor locations.
• As we already discussed the cost of capacitor also
depends on at how many places we put this.
• Usually in a primary distribution feeder even consisting
of several nodes the number of capacitor location is
very few e.g. maximum not more than 4 or 5.
• Thus the procedure described earlier could be repeated
for number staring from 2 and find the optimum location
and sizes for each number and computing the saving for
each number
• To simplify the computation of energy loss it can be
assumed that the capacitor vary continuously as per the
load (though it is impractical and costly). Thus Loss of
load factor (LLF) is applicable for energy loss reduction
calculation.
10/30/24 AKM/distplng/Distribution loss 43
reduction
 Capacitor control
• As the load in a distribution do not remains the same
throughout, it may happen that loss at OFF-peak period may
be greater with the fixed capacitor determined from peak
load than without capacitor. So optimal control is necessary
this may be achieved in several ways e.g.
Option-1 all the capacitors are variable
For this since the capacitor location is already fixed hence
H matrix remains the same only compute the b vector for the
particular loading and compute the optimal size using:
Ic(optm) = H-1b
Option-2 Capacitors are fixed only ON/OFF is possible
If we have ncap capacitors placed in the system the
possible ON/OFF combinations could be 2ncap. Check which
combination suits best for which loading condition.
Another option may be only one or two variable and rest
fixed.
10/30/24 AKM/distplng/Distribution loss 44
reduction
 Voltage constraints inclusion
• One of the purpose of capacitor placement is also to improve
voltage regulation. In this case the overall objective function for
capacitor placement can be expressed as:
ncap
S = kp * LP + Ke* LE - Σ (Kcf+Kc*Qci0)
i=1
subjected to
Vmin Vi Vmax
• It is possible to optimize the above expression using advanced
techniques.
• Because the method discussed here do not include the voltage as
constraints.
• Though the capacitor placement improves the voltage level need not
necessarily able to keep the voltage at all the nodes within the limit.
• This method assumes that if the final voltage at any node do not
comes within limit, it is made by placing a voltage regulator at
appropriate location.
10/30/24 AKM/distplng/Distribution loss 45
reduction