Kaleb Pattawi

Date of Award


Document Type



Santa Clara : Santa Clara University, 2021.

Degree Name

Master of Science (MS)


Mechanical Engineering

First Advisor

Hohyun Lee


Energy consumption in buildings is projected to increase more than 50% by 2050. Renewable energy sources are expected to increase and will help alleviate some of the energy demand; however, the intermittent nature of renewables makes it tough to effectively use energy in buildings. Various pricing schemes, including time of use electricity pricing, are used to encourage users to shift loads. Buildings need to be able to proactively react to changing energy sources and electricity prices in order to reduce over-generation of electricity and save money. This work demonstrates an optimized Heating Ventilation and Air Conditioning (HVAC) control strategy that considers occupancy probability and real-time occupancy information and a simulation framework that allows for complicated building controls to be simulated using EnergyPlus™, a well-established building simulation software. Furthermore, this simulation framework is scalable to multiple buildings and can be used to model energy consumption in whole communities. In the past, end users could be modelled as simple nodes that consumed energy, but with the increase in renewables and distributed power sources, end users can participate in the energy markets more actively. Now, accurately modelling individual houses is crucial for modelling energy consumption in communities. While commercial buildings have tangible economic impact, this approach is aimed at residential buildings because at scale, residential buildings can have a significant impact on energy consumption. The results show that an optimized HVAC control strategy that considers occupancy probability and real-time occupancy can reduce weekly cost of electricity by 21.8% compared to a fixed HVAC control strategy while also providing peak shaving capabilities. However, results also suggest that users can benefit more from a simpler adaptive HVAC control strategy and save 31.6%. These findings suggest that the current pricing scheme does not provide enough incentive for users to alter energy consumption behavior and as a result, the grid will not benefit from potential load shifting. Additionally, the impact of preheating and precooling with optimization is not significant and these results encourage the advancement in energy storage devices to provide more justification for a complex HVAC control strategy. This work also explores the effect of occupancy consideration on energy consumption. This outcome demonstrates that occupancy consideration can have a significant impact when modelling individual houses of up to 10.8% change in cost and 11.3% change in energy consumption. Occupancy information should be considered in future works, especially those aimed at smart grid technology and transactive energy approaches. Finally, this work demonstrates that a scalable simulation framework has the capabilities to accurately model a transactive energy approach by showing that an optimized occupancy-driven HVAC control strategy can reduce the peak load by 7.5%.