An Agent-based Dynamic Occupancy Schedule Model for Prediction of HVAC Energy Demand in an Airport Terminal Building

Abstract

Airport terminal buildings are characterised by rapidly fluctuating occupancy levels in different zones. Occupancy is one of the major factors that influences the indoor environment and building energy consumption. The paper presents an approach to couple pedestrian flow model with energy simulation to predict the HVAC energy demands in the transitional environments. For the study, a medium sized airport at Visakhapatnam, India, located in warm and humid climate zone is considered. Occupancy dynamics of the terminal building is modelled and analysed through an agent based model (ABM). The results show a significant difference in the characteristics of the occupancy profiles between the various zones in the terminal building. A coupled energy simulation is carried out using the dynamic occupancy schedule obtained from ABM. The paper presents the impact of pedestrian density on HVAC loads at different zones. Also load profile for typical days compared for the peak load performance. Finally, the paper presents the comparison of HVAC loads predicted using the occupancy schedule based on flight operations and dynamic occupancy schedule based on agent based simulations.

Ten questions concerning agent-based modeling of occupant behavior for energy and environmental performance of buildings

“…Agent-based approach is also helpful in examining the efficacy of energy-saving policies in shared environments, such as university classrooms or office spaces, where interactions and peer effects have significant influences on OB [9], [64], [65]. For occupancy-driven analysis, such as evaluating performance of lighting sensors [66] or HVAC demand in airport terminal buildings [67], agent-based occupant modeling approaches are useful to simulate the stochasticity in occupancy patterns. Other potential application of ABM could be within net-zero energy communities, where occupants' acceptance of technological solutions is central to achieving the desired energy savings, or for cases where occupant-centric BPIs such as occupant hours and peak load per occupant hour are of interest.…”

A level-of-details framework for representing occupant behavior in agent-based models

“…Later, a multi-agent model of occupant activities and decision making using the SMACH platform was developed along with a validation approach [43], [45]. In addition to office and residential building settings, agent-based occupant models have been applied in an airport for estimating cooling energy demand [46], a hospital for estimating energy consumption [47], administrative offices within schools to account for the impact of organizational issues on energy consumption [48], and a university campus for identifying an optimal HVAC strategy [49].…”

“…For instance, Vellei et al [39] adopted stochastic occupancy models to capture occupant diversity in thermostat behavior in residential buildings. Likewise, Sinha et al [46] implemented an agent-based passenger flow model for estimating HVAC demand in a terminal building. Azar et al [49] also introduced stochasticity in their occupancy model to identify energy management strategies for a university campus.…”

“…It is important to note that the implementation of LoD O-3 must be carried out with utmost attention, since the complicated models may be difficult to initialize and lack transparency and completeness [51]. An example of a relatively complicated ABM in the BPS domain was developed by Sinha et al [46] to estimate HVAC energy demand in an airport terminal building. The model features individual-level granularity, detailed internal zoning, and stochastic decision-making models (LoD O-3) that affect passenger density and, in turn, the zones' HVAC demand.…”