Implementation of the load survey system in Taipower

This paper is to illustrate the load survey project which has been designed and implemented in Taipower since 1993. The multiple functions of system planning and operation to be supported by the load survey system have been identified to justify the cost effectiveness of the project. The stratified sampling algorithm has been applied to determine the proper size of customers for the installation of intelligent meters so that the customer load profiles solved can represent the actual system load characteristics in a sufficiently confidential level. In the study, the test customers within each customer class are selected by the random number method based on the standard deviation of power consumption of all customers in the class. The power consumption of test customers within every 15 minutes has been collected for 4 years and the statistic analysis has been applied to derive the typical load patterns of each customer class. With the precious information provided by the load survey system, Taipower is considering to revise the load forecasting and system planning, to design more proper tariff structure according to the actual contribution of power consumption by each customer class and the corresponding electricity service cost. By investigating the load composition of each customer class, more effective load management strategies are also designed by Taipower to reduce the system load demand during summer peak period.