Energy-optimised dynamic occupancy scheduling for offices with desk sharing

Desk-sharing concepts offer significant potential for improving space utilisation in modern office buildings, but their impact on HVAC energy consumption is often neglected in seat allocation planning. This study proposes an automatic scheduling tool for energy-aware seat allocation that determines occupancy plans in advance based on expected attendance patterns, without relying on real-time sensor data. The approach is formulated as a mixed-integer linear programme (MILP) that integrates employee attendance, thermal room dynamics, HVAC capacity limits, and organisational constraints such as departmental adjacency. The model is evaluated in a representative office scenario over a full working week using 100 randomly generated occupancy cases. The results show that the optimised schedules achieve an average HVAC energy cost reduction of 29.1 % compared to conventional fixed-seat assignments, with savings ranging from 16.0 % to 45.3 %. By directly incorporating HVAC energy optimisation into desk-sharing planning, this work extends existing scheduling approaches and provides a foundation for future integration of more detailed building energy models.