Electric Load Forecasting - Classification, Procedure and Approach

The load forecasting is the only way to predict the power/energy needed to meet the demand and supply equilibrium.Here we discuss about the  Load forecasting Classification, approach and the way of procedure which is used to complete the forecast. you have needed to read Load Forecasting - an Introduction and their significance before continue further.

Load Forecasting Classification

Load forecasting can be classified as demand forecast and energy forecast.

1.Demand Forecast - This is used to determine the reverse capacity of the generation, transmission and distribution system, Future demand can be predicted on the basis of rate of growth of demand form past history and government policy, this will give the expected rate of growth of load.

2.Energy Forecast - This is used to determine the type of facilities required i.e future demand requirements

Load Forecasting procedure

Depending on the time period of interest a specific forecasting procedure may be adopted. This may be classified as

1. Short time technique
2. Medium-term technique (Intermediate)
3. Long-term technique

1. Short Time Forecast - Short time forecasting is done for day to day operation, to ensure generation capacity for a week, and for maintaining the required spinning reserve. Hence it is usually done 24 hours ahead of when the weather forecast for the following day because available from the meteorological office. Application of short time load forecasting STF are

• To meet the short-term demand with the most economic commitment of generation sources.


• To access the power system security based on the information available to the dispatches to prepare the necessary corrective action.

2.Medium-term Forecast - This forecasting is done for 5 to 6 years and play a major role while planning the size of a power plant and construction and installation of the equipment in power plants, Applications of medium-term load forecasting MTLF

• For the estimation of fuel (Coal, diesel, water etc)


• For the estimation of peak power & energy requirement for each month of the coming year.

3.Long-term forecast - Long-term forecast may extend over a period of 20 years or even more and in advance, in order to facilitate various plan like, the preparation of maintenance schedule of the generating units, plan for future expansion of the generating capacity enter into agreements for energy interchange with neighbouring utilities

Load Forecasting Approach

Two types of approaches are available for Load Forecasting

1. Peak Load Approach


2. Energy Approach

1.Peak Load Approach - In this approach the simplest way is to extrapolate the trend curve, which is obtained by plotting the past values of annual peak against year of operation. The following analytical function can be used to determine the trend curve

1. Straight Line                y=a+bx


2. Parabola                       y = a+bx+cx^2


3. S- Curve                        y = a+bx+cx^2+dx^3


4. Exponential                 y = ce^dx


5. Gompertz                     log y = a+ce^dx

let us drive the function for straight line by assuming that the load increase in an exponential fashion, i.e. to match exponential trend curve to the historical data. Let D be the power demand is considered X0 be the reference or basic year. Under the assumption of an exponential growth of load demand, we have

D = e^a+b(x-x0)

where a & b are constant

Let us assume

x-x0 = x

so 

D = e^a+bx

Taking natural logarithms on both sides

Ln D = a+bx

substitute 

y = Ln D

then 

y = a+bx

now y and x are related to linear & hence varies in a straight line manner. The same procedure applies for the remaining. In all these function y represent the peak load and x represent time in years.

2. Energy Approach - In this approach, the annual energy sales is forecasted to different classes of customers like residential, commercial, industrial etc. which is converted to annual peak demand using the annual load factor. A detailed estimation of factors such as the rate of building a house, sales of electrical appliances and growth in industrial and commercial activities are required in this method. Forecasting the annual load factor also contributes critically to the success of the method.

Load Behaviour

1.Domestic Load - This type of load mainly consists of domestic appliances such as lights, fans, heater, refrigerators, air conditioner, mixer, oven, heating ranges, small motors for pumping and various other small household appliances. Peak real load refer to meet maximum real power load during a particular time. Peak reactive load refer to maximum reactive load during a particular time.

The daily load curve DLC of a weekday of domestic load in terms peak real and reactive loads as shown in figure.

2.Commercial Load - Commercial Load consist of lights, fan, air conditioners, heater and other electrical appliances used in a commercial establishment such as shops, restaurants and marker places. The DLC of a weekday of this type of load in term of peak load is shown in the figure.

3.Industrial Load - This type of load may be subdivided into small, medium and heavy depending on required power range. for example, small-scale industries require load upto 25 KW, Medium scale industries require load in between 25 KW to 100 KW and heavy industries require load more than 500 KW. These loads are considered as base load that contain small weather dependent variations. The DLC of a week of this type of load in term of peak load is shown in the figure.