Enhancing Utility Operations during Heat Waves through Large-Scale Sensing and Data Fusion

Improving utility operations to protect vulnerable populations during heat waves

Background and Motivation

Heat waves lead to sudden increases in electricity demand, posing critical challenges to power system operation. To address this, utilities should: (a) improve the ability to forecast load and demand flexibility during heat waves and better inform the resource scheduling decision hours ahead of these events and; (b) estimate population overheating risks, particularly in vulnerable communities in order to integrate them into the decision-making process of emergency response.

This project aims at enhancing utilities operation during heat waves by developing new models, based on large-scale sensing and data fusion techniques that will utilize existing data and machine learning algorithms to estimate hours-ahead electricity demand, flexibility of aggregated building stocks and overheating risks of vulnerable communities during heat waves.

Urban scale building modeling with scope and data sources

A distinctive aspect of this approach is the use of existing data – namely the Ecobee Donate Your Data (DYD) dataset which has smart thermostat data for more than 120,000 U.S. homes – to forecast demand as well as estimate residents’ indoor thermal comfort and overheating risk as a function of power outage duration. Finally, this project will also develop innovative methodologies to integrate these techniques into real-life utility planning and operations and will demonstrate them in partnership with Portland General Electric.

Objective

This project will develop a urban building energy model and novel methods to:

Forecast building electric load during historical and future heat waves;
Estimate the “overheating risk” of building occupants as a function of power outage duration (from 1 to 8 hours); and
Identify and evaluate demand side load reduction strategies through energy efficiency retrofit, operational improvements, or behavioral actions.