Real-Time Forcast Model Analysis of Daily Average Building Load for a Thermal Storage System Control

Abstract

Thermal storage systems were originally designed to shift the on-peak cooling production to off-peak cooling production to reduce the on-peak demand. Based on the current electricity charging structure, the reduction of both on-peak and off-peak demands is becoming an exceedingly important issue. Reduction of both on-peak and off-peak demands can also extend the life span and defer or eliminate the replacement of power transformers due to potential shortage of building power capacity with anticipated equipment load increases. The next day daily average electricity demand is a critical set point to operate chillers and associated pumps at the appropriate time. For this paper, a mathematic analysis was conducted for annual daily average cooling of a building and three real-time building load forecasting models were developed. They are first-order autogressive model, random walk model and linear regression model. Finally, the comparison of results show the random walk model provides the best forecast.