2019 - PI World - San Francisco - Water Energy and Facilities
Forecasting and Dynamic Real-Time Optimization of a Campus District Energy System using PI (Univ of Utah)
Increasing penetrations of intermittent renewable sources of energy on the grid introduces a new element of uncertainty into the grid and complex energy systems. Energy storage is a key enabling technology to resolving some of these intermittency issues. However, the performance of the energy storage can be drastically improved if coupled with smart algorithms to predict future energy needs and optimize the system dynamically over the course of some timespan. Using the University of Utah as a living laboratory and PI for data infrastructure, researchers have built a live software application that predicts future campus cooling demand using machine learning techniques. The application also solves for optimal operation of a campus cooling system with thermal energy storage and minimizes the total cost to run the system over a 24-hour period. Some of the key issues include intelligent ways to continually "learn" the system so that optimal results remain accurate and effective.
Industry
Infrastructure, Facilities and Buildings
Company
University of Utah
Speaker
Kody Powell
Dr. Kody Powell is an Assistant Professor and automation systems and simulation researcher in the Department of Chemical Engineering at the University of Utah. His research work primarily deals with smarter operation of complex energy systems in order to improve system reliability and efficiency. The research leverages system flexibility via storage, hybridization, or scheduling in a proactive manner, often including forecasts of future energy supply and demand in order to make the system perform optimally over a time horizon. Dr. Powell is also the Director of the Intermountain Industrial Assessment Center, a U.S. Department of Energy-funded program that provides technical expertise to manufacturers. Through this program, Dr. Powell is able to incorporate real-world systems, data, designs, and technical challenges from the manufacturing sector into his research.
Company
University of Utah
Speaker
Pratt Rogers
Pratt is an Assistant Professor of Mining Engineering at The University of Utah. He is deeply interested in the use of data to improve organizations. He has extensive consulting and research experience in the mining industry related to business intelligence, mining technology, machine learning operational excellence, and sociotechnical systems. Before joining University of Utah, Pratt was VP Product Development for MISOM Technologies where he was a highly effective manager and cross-functional team builder. He and his team worked on delivering software and business intelligence solutions with an integrated information platform, which delivered analytics, and reporting used for continuous improvement of safety, costs, and productivity. Pratt holds a BS and PhD in Mining Engineering from University of Arizona.
Session Code
US19NA-D2FW05