Statistical distributions of load profiling data

This paper present an approach to estimate the customer load confidence interval using statistical distribution model. Time series data of load profile in the utility of Malaysia was analyze and this data was fit with two statistical distribution models, namely the normal and log-normal distribution. Goodness of fit test was performed in the selection of best model. Result shows that the customer load consumption for each seven days can be model using both distributions. An estimated 90%, 95% and 99% customer load confidence interval was done via both distribution models.