LOAD CHARACTERISTICS

AND
LOAD FORECASTING

DEFINITIONS
CONTD

Load

Electrical power needed in kW or kVA .

Demand

The power requirement (in kVA or kW)

at the load averaged over a specified interval (15

min or 30 min). Sometimes it is given in amperes at
a specified voltage level.

Demand Intervals

The time interval specified

for demand (Di), usually 15 min or 30 min. This is

obtained from daily demand curves or load duration
curves.

DEFINITIONS
Maximum Demand

CONTD

The maximum load (or

the greatest if a unit or group of units) that

occurred in a period of time as specified. This
can be daily, weekly, seasonal or on annual basis
(for billing purpose in India it is monthly and in
kVA).

Connected Load

The sum total of the

continuous rating of all the apparatus, equipment,

etc., Connected to the system.

DEFINITIONS
CONTD

Demand Factor The ratio of maximum demand

to the total load connected to the system

Utilization Factor

The ratio of maximum

demand to the rated capacity of the system.

Load Factor

The ratio of average load in given

interval of time to the peak during that interval.

Annual Load Factor

The ratio of total energy

supplied in an year to annual peek load multiplied by

8760.

DEFINITIONS
CONTD

Diversity Factor ( Df )

The ratio of sum of

the individual maximum demands of various subdivisions of the system to the maximum demand
of the entire or complete system.

Coincident Maximum Demand ( Dg )

Any demand that occurs simultaneously with any
other demand and also the sum of any set of
coincident demands.

DEFINITIONS
CONTD

Coincidence factor (Cf )

This is usually

referred to a group of consumers or loads. It is

defined as the ratio of coincident maximum
demand Dg to sum total of maximum demands of
individual or group of loads.
Generally, it is taken as the reciprocal of the
diversity factor.

DEFINITIONS
CONTD

Load Diversity

The difference between the

sum of peaks of two or more individual loads and

the peak of combined load.

Load diversity = Di Dg
Di = individual maximum demand
Dg = coincident maximum demand

DEFINITIONS
CONTD

Coincidence Factor

This is a factor that is

usually referred in distribution systems regarding the

importance of weighted effect of a particular load.
If C1 ,C2 , ..... Cn are the contribution factors of each
of the n individual loads and D1, D2, D3 . Dn are
their maximum demands.
Dg = coincident maximum demand is taken as

DEFINITIONS
CONTD

Loss Factors

This is the ratio of average

power loss in the system to power loss during

peak load period and referred to the variable
power losses, i.e., copper losses or power loss in
conductors or windings but not to no load losses in
transformers, etc.

LOAD & LOAD

CHARACTERISTICS

A load or power requirement (also kVA) of a consumer varies widely. But

in general the consumers can be grouped into a few categories as their
needs and demands are the same.
A broad classification of loads are
(i) Domestic and residential loads
(ii) Only lighting loads (such as for street lights etc.)
(iii) Commercial loads (shops, business establishments, hospitals)
(iv) Industrial loads
(v) Agricultural loads and other rural loads

Note :

All these loads will have peak demands at different

times and for different durations.

Industrial and commercial loads may have two peak load

periods.

Agricultural loads are seasonal and vary very differently.

Lighting loads such as street lighting etc. may have

almost zero demand during day time and constant load
during 10 p.m. to 6 a.m. and a slightly higher demand
between 6p.m. to 10 p.m.

Another classification of electrical

loads is the billing categories used by the electrical
distribution authorities or State electricity boards.
This includes categories such as residential and
domestic,
industrial, commercial, rural, HT consumers and
others.

System Power Factor

In ac systems, kVA demand is more appropriate than kW and load
power factor is of importance. Typical p.f of residential, commercial and
Industrial loads are as follows.
Fluorescent lamps

0.6

Arc lamps and neon signs

0.4 to 0.7

Fans and small motors

0.5 to 0.8

Electronic gadgets

0.6 to 0.8

Domestic appliance

0.6 to 0.7

(like washing machines ,vacuum cleaners etc.)

F.H.p. motors (1 kW or less)

Water pumps (Large size 5 h.p)

Chemical industries

0.4 to 0.75
:

0.65 to 0.8

0.70 to 0.85

Domestic and Residential Loads

The important part in the distribution system is domestic
and residential loads as they are highly variable and
erotic. These consist of lighting loads, domestic
appliances such as water heaters, washing machines ,
grinders and mixes ,TV and electronic gadgets etc. The
duration of these loads will be few minutes to few hours
in a day. The power factor of these loads in less and may
vary between 0.5 to 0.7. In residential flats and bigger
buildings, the diversity between each residence will be
less typically between 1.1 to 1.15 . The load factor for
domestic loads will be usually 0.5 to 0.6.

Industrial Loads
Industrial

loads are of greater importance in distribution

systems with demand factor 0.7 to 0.8 and load factor 0.6 to
0.7. For heavy industries demand factor may be 0.9 and load
factor 0.7 to 0.8
Typical power range for various loads
Cottage and small-scale industries

: 3 to 20 kW.

Medium industries (like rice mills, oil mills, workshops,

etc.)

: 25 to 100 kW

.Large industries connected to distribution feeders (33 kV

and below)

: 100 to 500 kW

Water supply and Agricultural

Loads
Most of the panchayats , small and medium
municipalities have protected water system which
use
pumping stations. They normally operate in off
peak time and use water pumps ranging from 10
h.p to 50 h.p or more, depending on the
population and area.

Agricultural and Irrigation Loads

Most of the rural irrigation in India depends on
ground water pumping or lifting water from tanks or
nearby canals. In most cases design and pump
selection is very poor with efficiencies of the order
of 25%. Single phase motors are used (up to 10
h.p.) for ground water level 15 m in depth or less
with discharge of about 20 l/sec while multi stage
submersible pumps with discharge of 800 to 1000
l/m may require motors of 15 to 20 h.p.

Sensitive and important Loads

With

computer

applications

in

every

area,

computer loads and computer controlled process

loads are often non-linear and sensitive. They
require close tolerance limits for voltage and
frequency (voltage limit 5% and frequency 0.5
Hz with unbalance and wave form distortion less
than 3%. This requires special attention while
providing the distribution of electric power.

Load Curves and Load Duration

Curves
The consumption of electrical power or energy by any utility
varies from time to time in a day as well as during a week ,
month , season or year . For example in summer fans , AC
units , cooler etc are used but not during winter or cold
season. Industries working during day time will consume only
lighting load during night( 10 pm to 6 am) . Hence knowledge
of variation of loads and their nature is essential for
distribution planning . The load characteristics are usually
presented as load curves and load duration curves

a. Load Curves
The load ( power requirement ) of any concern or unit
is tabulated as the amount of power required or
consumed during a certain period in a day, week or
a given season.

(b) Load-Duration Curves

This is a graph obtained from load curve showing the
load in (kw) and duration over which it occurs in
descending order of load magnitudes.

Refer the following Load data on a Typical 11 kV Line

and Answer the following questions
Time Hour/
Load kW

Street
Light

Residential

Commercia Industrial
l

Agricultur
al

0.00 6.00 AM

80

200

320

100

600

6.00 8.00 AM

700

400

100

400

8.00 9.00 AM

800

400

300

9.00 10.00 AM

600

400

400

10.00 5.00 PM

500

700

400

5.00 6.00 PM

600

900

400

6.00 7.00 PM

80

800

1200

320

7.00 8.00 PM

80

1000

1200

320

8.00 9.00 PM

80

1000

1200

220

9.00 10.00 PM

80

800

1050

170

10.00 12.00 PM

80

500

320

100

400

Q1 Draw Load Curve and Load Duration Curve .

Q2 What are the peak ( maximum) demands of individual load ,
system maximum demand.
Q3 What is the contribution factors for each of the loads ?
Q4 What is the diversity factor and coincidence factor for the
above loads ?
Q5 For the feeder load given in the table , at a peak load of
1500 kW , the power loss recorded is 75 kW . If the annual
loss factor is 0.2 , what is
a. Annual average power loss , and
b. Total energy loss per year ?

Load

Growth

And

Diversified

Demands
As the residential or commercial areas grow with
increased population and new areas area added,
it will be necessary to account for the new loads
added and also to take in to account diversity
between similar loads and non coincidence
between peaks of different types of loads added.
This will optimize the additional capacity to be
added. For the purpose variation in the peaks of
different kinds of loads is taken.

LOAD GROWTH AND FORECASTING

Population growth and energy requirement do not
grow linearly but follow non linear power law or
exponential growth. The usual function that fits is
y = ka x, where y is the new value after a growth
period 'x', k is the initial value of y i.e, when x = 0 and
'a' the rate at which y increases logarithmically.

Power growth, i.e., increase in load demand

after a period x with and annual growth rate of 'g' is
usually expressed as

Pn =Po( 1 + g ) n
Po = initial power demand
g = growth rate
n = period
This is also known as compound interest law
(here 1 + g = a of the previous equation, y = kax)
Future Demands are normally estimated knowing
the growth rate factor g

LOAD FORECASTING
Based on certain conditions and trends existing and assuming
that they continue, load forecasting is a method by which future
increase in loads are predicted. There is a great need for
accurate forecasting of loads over a given period to meet with
the power and energy requirements of the future and money to
be spent. In our country there is a lot of pressure due to limited
financial and energy resources and hence electrical load forecasting is vital.
The models adopted for load fore casting are statistical models
based on Markov process, Time Series analysis and Sampling
techniques. The method used is regression analysis.

REGRESSION ANALYSIS
The mathematical modelling is done by taking the previous
growth over a period and the future trend is extra polated .
This is done by either fitting a linear or non linear curve for
the growth to get least overall error or fitting a sequence of
discontinuous non linear curves from the pervious data extra
polating the results. The factors that are taken into account
are
i.

basic trends,

(ii) seasonal variations, and

(iii) random and cyclic variations depending on

weather conditions.

The trends are fitted into either

(i) Linear increase as

P = a + bt

(ii) Exponential or compound interest law

Pn = Po(1 + g)n

or

(iii) other power laws like

(a) P = A xb (exponential growth)

(b) P =A + Bx + Cx2 (quadratic law)
may also be used.

These trends and estimates are checked with

typical correlations from available records and
actual values.
To conclude, load forecasting and energy
forecasting for future years is difficult but
necessary process in order to plan for future
power and energy requirements.