Slide 1.pngSlide 2.pngSlide 3.pngSlide 4.pngSlide 5.pngSlide 6.png

    

International Conference on Innovative Applied Energy    

E-Proceedings ISBN: 978-1-912532-05-6

St Cross College, University of Oxford, United Kingdom

  


 

Decentralised Energy OptimisationFor Blocks of Buildings

 


 

Sean Williams, Michael Short and Tracey Crosbie

School of Science, Engineering and Design, Teesside University, UK

 

  

Paper Abstract

Policy and new technologies are transforming the energy landscape in the UK. Centralised control of electrical generation and unidirectional distribution have a finite part in a sustainable energy system. Subsidies have encouraged an increase in distributed resources. At the same time closure of larger fossil-fuelled power plants is reducing system inertia on energy networks. In this study, a decentralised proactive approach to demand-side response exploiting building thermal inertia is presented using machine learning methods and a real-time adaptation algorithm. This paper proposes a dynamic 2-step energy consumption prediction scheme that can be configured to provide efficiency opportunities and the potential to reduce energy costs in buildings. The approach adopted optimises energy usage through existing demand-side response mechanisms utilising decentralised frequency regulation. The paper concludes with a discussion on the future direction of research. 

Paper Keywords
Decentralised, machine learning, Dijkstra, thermal inertia, frequency regulation.
Corresponding author Biography

Sean Williams is a 2nd year PhD research student part of the Doctoral Training Alliance in Energy. Sean is a graduate of Teesside University, where he received a MEng (Hons) First Class in Instrumentation and Control Engineering (2016).

Sean’s early career led to a 7-year partnership with the Design Equipment & Support trading entity and organisation within the UK Ministry of Defence and positions as project engineer. It was in this capacity that his expanding expertise in aircraft mission support systems allowed him to include several high value projects in Italy, Spain, Germany and Saudi Arabia.

Specialising in control and optimisation of energy systems, Sean’s first journal publication to Applied Thermal Engineering proposed how a proactive approach to demand response employing heat transfer dynamics can be realised by employing decentralised frequency control regulation.

More recently, his attention has focused on promoting an energy optimisation system (EOS). The EOS contribution to a facilities energy management strategy is to decompose the overall multi-objective optimisation problem into 2 main parts. The first aims to create a grid frequency prediction model, and the second is based on a modified Dijkstra’s algorithm, credited for finding the shortest path between nodes in a directed acyclic graph (DAG). Here, the objective function calculates the optimal temperature setpoint by adjusting the power consumption of thermostatically controlled loads (TCL) required to regulate indoor space heating and air conditioning.

The International Conference on Innovative Applied Energy (IAPE’18)