Development of a Practical Controller for the Optimised Control of Hot-Water in Grid-Connected Microgrids and Embedded Networks

Abstract

As New Zealand’s electricity industry braces for disruptive change in the face of increasing public consciousness to the worsening impacts of climate change, evolving government policy and the impositions of electrification to transport and industrial heating, the question of managing the country’s energy infrastructure is emerging as a critical component in the development of a more flexible electrical grid.

This thesis examines the role of hot water load control and its potential to contribute to the problem of managing New Zealand’s peak energy demands. Drawing from historical events such as the localised power cuts of August 2021, this study explores the effectiveness of traditional hot water ripple control and the centralised load control of hot water cylinders. While serving New Zealand well, the effectiveness of traditional ripple control (RC) is limited in its current state, requiring an exploration of more modern, flexible approaches to demand response, especially in the face of evolving technologies such as electric vehicles (EVs) and residential heat pumps.

By developing and validating a digital model of a hot water cylinder and exploring computer algorithms for its smart control in the context of a grid connected microgrid, this research estimates the potential for hot water cylinders to satisfy end-user needs centred around comfort and cost while also benefiting the wider electricity grid by reducing peak electricity demand.

Potential approaches to hot water load control including dynamic programming, shortest path search algorithms, and reinforcement learning are investigated as strategies to automate and optimize hot water load control, offering insights into the potential contribution the improved management of the humble residential hot water cylinder could provide to New Zealand on its journey towards a more sustainable, flexible electricity system.