Power System Planning.docx

4.2 Load Characteristics Let us start from a low power appliance such as a refrigerator, turning on and off, irregularly. At the same time, there are other appliances at a home, which, somehow, and to some extent, smooth out the load fluctuation of that home. Now, what happens to the load fluctuation of a distribution substation feeding several H. Seifi and M. S. Sepasian, Electric Power System Planning, Power Systems, DOI: 10.1007/978-3-642-17989-1_4, Springer-Verlag Berlin Heidelberg 2011 45

homes. Still, the smoothing becomes more apparent. The daily load of a distribution substation may look like the one shown in Fig. 4.1. On the other hand, the distribution substations are supplied, through subtransmission and transmission networks; from transmission substations. The daily load curve of a transmission substation has a general shape similar to Fig. 4.1. The same is true for the whole network consisting of several transmission substations. Instead of focusing on the actual level,1 let us now focus on the load shape of Fig. 4.1. Suppose we are going to know the load shape of the last week. Obviously we should gather the minute-by-minute data required. To simplify the task, let us assume that the load does not vary in each hour. In that case, the load shape may be drawn as shown in Fig. 4.2. It is evident that the load shape of a working day is significantly different from that of a weekend day. Moreover, even the load shape of working days may be different due to, say, the weather conditions. Let us now, go further towards the load shape of the past year. If the time step used is still 1-hour, 365 9 24 = 8760 data are required. It is evident that the task may be accomplished for a year or even for the last 10 years or more. These load shapes may be used for the detailed calculation of energy demands. However, they are of less use in planning studies, as we will see in this book. Let us, now, focus on the future, without any available data. We are going to forecast the hour-by-hour daily (or weekly) load of our test case. This load curve is used by the system operator to decide the necessary actions. On the other hand, if for evaluating the generation deficiency, the load shape of the coming summer is to be forecasted, is it really possible to do so? In other words, it is possible to predict, the hour-by-hour load for several months from now? Basically if we have to forecast the hour-by-hour load,2 we should accept the uncertainties involved.3 However, we will see in this book that we often require less detailed, but as accurately as possible, the 86420 10 12 14 0 3 6 9 12 15 18 21 24 Time (h) Load (MW)

Fig. 4.1 The daily load of a distribution substation 1 Here we are not involved with the level. In Sect. 4.4, we will come back to the point. 2 For some types of studies such as fuel and water managements. 3 See Chap. 11 for the uncertainties involved in power system planning problem. 46 4 Load Forecasting

load shape. For instance, we may want to know the variation of daily peak loads of the coming summer. In other words, only 90 data are required. Such a seasonal load curve is shown in Fig. 4.3. We have not assumed that the load is flat in each day. Instead, we have only focused on the peak values. If for planning purposes, we are going to predict the load variations for the next several years (say 10 years or more), we do not bother the daily variations. 4 Instead we may have to predict, say, the summer and the winter peaks, of the coming years. It means that for a 10-year prediction, only 20 data are required.

4.3 Load Driving Parameters Once we have talked about the various load shapes in Sect. 4.2, in this section we focus on the parameters affecting the forecasted load of future. These driving parameters are quite a few. Some typical ones are as follows 86420 10 12 14 0 3 6 9 12 15 18 21 24 Time (h) Load (MW)

Fig. 4.2 The discretized load of the distribution substation 8 6 4 20 10 12 14 16 18 20 0 10 20 30 40 50 60 70 80 90 Days Load (103 MW)

Fig. 4.3 A typical seasonal peak load curve 4 Even if we bother, who can predict the daily variations of say, 5 years from now?