Coupling simulation with machine learning for the development of a proactive HVAC system in the manufacturing sector

The industrial sector consumes 55% of the world's energy consumption [1]. Following manufacturing processes, the HVAC system is the second largest energy consumer in manufacturing facilities, yet is generally uncounted for and considered an indirect cost to maintain a facility [2]. Any current efforts at reducing energy demand in the manufacturing sector have been focused towards process machines rather than on the manufacturing building as a holistic energy system. Currently, HVAC systems are reactive, responding to changes to the environment as they happen, based upon requirements for thermal comfort. Manufacturing facility environments however are subject to complex interactions between machine level resources, water, heat and compressed air.;This study questions the suitability of the reactive thermal comfort based HVAC system, and proposes a proactive manufacturing based HVAC control system, utilising predicted optimum HVAC set points. Through the use of simulation, a holistic analysis of a manufacturing facility was performed, based on building location and layout, building fabrics, weather conditions and manufacturing demand in order to determine the relationship between manufacturing demand and HVAC control. A number of predictive models were analysed for suitability for use in the manufacturing, before being trained on simulation data for the prediction of optimum HVAC set points and corresponding facility indoor conditions.;Simulation was coupled with predictive modelling in order to predict building energy and HVAC energy demand, allowing for the identification of potential future spikes in consumption, followed by subsequent HVAC and manufacturing schedule optimisation, allowing for a 15.1 % reduction in peak energy demand. Through simulation and predictive modelling, the research has demonstrated the potential energy savings achieved by adopting a proactive HVAC system in the manufacturing sector. Such a methodology achieved 14.1 % energy savings over a 12-month period for an analysed case study environment. [See thesis text for references]

Name dc:type.qualificationname

phd

Level dc:type.qualificationlevel

doctoral-pg

Grantor dc:publisher.institution

University of Strathclyde

Year dc:date.issued

2020

Author dc:creator

• Mawson, Victoria Jayne

Advisor dc:contributor.advisor

• Hughes, Ben

DOI

10.48730/rcej-rf76

Identifier

T15785

Author Identifier

201953618

OAI identifier oai:identifier

oai:strathclyde:bz60cw24r