skip to Main Content

University of Hawaii

Electrical Engineering

Stochastic Optimization of Electric Water Heater Under Dynamic Pricing

Date: 2019-08-15           Add to Google Calendar
Time: 10:00am - 1:00 pm
Location: Holmes Hall 389
Speaker: Foad Najafi, PhD Candidate


With the improvement of communication technology and internet of things (IOT), the electricity demand response (DR) made it possible to control both generation and consumption in power systems. In this work, a new control strategy for controlling Electric Water Heater (EWH) using Mixed Integer Linear Programming (WH-MILP) is introduced. To balance between the possible discomfort (cold water) that DR could bring to the user and electric cost, a dual objective cost function is introduced which minimizes user discomfort and electricity bill at the same time. To schedule an electric water heater, two factors are more critical: electricity price and hot water withdrawal pattern. Many previous works only rely on electricity price for scheduling. However, this work is among the very few works that consider consumption alongside the electricity price for scheduling. But also, by treating hot water withdrawal pattern as a random variable, WH-MILP finds the best setpoints for EWH based on the result of stochastic optimization over random behavior of hot water withdrawal pattern.

The combination of a stochastic model with join probability distribution and an objective-based approach to user preferences is a qualitative improvement on previous work which only included one of these elements. To address these effects, a controller must use both a cost-based treatment of user preferences (not simple temperature constraints) and also a stochastic treatment of withdrawals. In the main body of the proposal, we present a simple probabilistic forecasting method for the joint distribution of hot water withdrawals in all hours of the coming day. This is the first model of its type for EWH withdrawals, and should be sufficient for developing and testing the stochastic water heater optimization model.