Lead Performer: Argonne National Laboratory – Argonne, IL
November 20, 2018
minute read time
Lead Performer: Argonne National Laboratory – Argonne, IL
Partners:
-- Leaptran Inc. – San Antonio, TX
-- University of Texas – San Antonio, TX
DOE Total Funding: $750,000
FY19 DOE Funding: $250,000
Project Term: October 1, 2018 – September 30, 2021
Funding Type: Lab Call
Project Objective
The aim of the proposed project is to develop new weather and energy forecasting algorithms with uncertainty by incorporating spatial-temporal data in and around the building. The project will crowd-source a set of weather underground data around the building to establish a spatial correlation between those data and the building site. In addition, data from building automation systems or connected thermostats like temperature set-points, room occupied status, and supply air flows, will be incorporated into the forecasting algorithm.
This project will develop three types of methodologies to train large data sets and forecast weather, energy and peak energy demands at different prediction horizons (e.g. 5 minutes to 24 hours ahead) for the purpose of advanced building controls. The project will also run a stochastic model to estimate the probability distribution of weather and energy forecasting, and to verify the impacts of energy savings. The algorithm will be tested on different types of building locations and load patterns before testing the weather and energy forecasting in a fully instrumented building – the San Antonio Technology Center – to validate the proposed approach.
Project Impact
This project will significantly improve the implementation of advanced building controls through at least 10% performance improvement in terms of accuracy for onsite temperature and solar irradiance, peak demand, and energy usage pattern forecasting on local buildings. With these tools, grid operators will have more accurate and finer resolution information about building loads to make smarter operational decision to improve grid economics and reliability. Additionally, building owners will be able to better quantify the values that building loads can bring to the electric grid. This will accelerate building owners to participate in grid services and realize the full spectrum of potential revenue.
Contacts
DOE Technology Manager: Erika Gupta
Lead Performer: Zhi Zhou, Argonne National Laboratory