Modelling Consumer Type Specific Electricity Load in Iceland

Author:  Berit Hanna Czock
Year: 2018
Supervisors:  Ewa Lazarczyk Carlson and Samuel Perkin

Abstract

Accurately modelling and forecasting electricity load is important for many aspects inmanaging and maintaining a power system. Different parties in the power sector havean interest in accurate short-term, medium-term and long-term forecasts. Next to thetotal load, different stakeholders in the power market are interested in user type specificdata and forecasts. Detailed consumer related data sets can be created using smart-meters, a technology that has not yet been implemented in Iceland. Instead user typespecific data can roughly be approximated using a Monte Carlo simulation approachwhich samples values from seasonal ARIMA models. Those are based on representativeuser type curves extracted from a data set of hourly total load observations at 44substations in the Icelandic system and their corresponding end user divisions (averageload for the year 2015). This approach has several advantages over the benchmarkmethod, user type specific curves based on the average end user divisions. It howeverfails to model lower resolution patterns so that the potential for better output qualitylies in improvements of the sampling method.