Development of a Coupled Bottom-Up and Top-Down Modeling Framework to Transform Long-Term Electricity Demand Forecasting Methods and Results – UROP Spring Symposium 2022

Development of a Coupled Bottom-Up and Top-Down Modeling Framework to Transform Long-Term Electricity Demand Forecasting Methods and Results

photo of presenter

Qi Cui

Pronouns: he/him/his

Research Mentor(s): Michael Craig
Co-Presenter:
Research Mentor School/College/Department: SEAS / SNRE
Presentation Date: April 20
Presentation Type: Poster
Session: Session 3 – 1:40pm – 2:30 pm
Room: League Ballroom
Authors:
Presenter: 87

Abstract

During the recent blackouts in California and Texas in 2021, peak electricity demands exceeded the limits. Since then, it become urgent for policymakers and scholars to precisely model and forecast electricity to prevent similar events from happening. With this objective, our project aims at modeling and forecasting electricity demands with various quantitative methods. Since peak electricity usage is a vital determinant for electricity supply, our main goal is forecasting when and where does the peak arrive. We first built various residential building models to generate future residential electricity demands. Then we used historical demands and meteorology data to derive future non-residential demands. With the outputs, we discovered seasonal patterns from those data and analyzed possible peak times to ensure abundant power supplies. One crucial contributor to these patterns is the use of heat pumps. We found that most electricity is used in summer and winter when heat pumps are widely used for heating and cooling. Meanwhile, more electricity is consumed in winter mainly due to the decreased efficiency of heat pump in cold weathers. Also, the pattern is region-based too since the efficiency of heat pumps decrease significantly in regions with extreme weathers like the Midwest. Our findings should be crucial for policymakers to allocate electricity resources more reasonably.

Presentation link

Environmental Studies

lsa logoum logo