Selection of the control alternative with the minimum energy consumption for the hybrid illumination system of inner spaces

Abstract The use of daylighting reduces energy consumption in buildings, which can be achieved by the implementation of the hybrid illumination systems (HIS). In general, the solutions of HIS exercise high controllability on every luminary for minimizing the energy consumption while maintaining the desired illuminance level by users. Nevertheless, it is only considered the demanded power by luminaries despite to auxiliary components (e.g. sensors, power-packs, and controllers) operate full time and their energy consumption may be representative. Therefore, this study proposes a procedure for sizing HIS for inner spaces with the lowest annual energy consumption ( A E C H I S ) based on an exhaustive search. All of the possible control alternatives are determined by a combinatory process considering calculation using Stirling number of the second kind, the number of control groups, and the number of available control strategies. The modelling of interior illuminance is based on the daylight factor matrices, the incident solar irradiation on each opening, the illuminance contribution by luminaries, and the operation conditions of control alternative. A E C H I S is defined as a function of number and power consumed by luminaries ( P l u m ) and auxiliary components ( P a u x ) considering operating conditions of HIS for every time-interval. P l u m is defined as an MINLP problem that depends on the state (On or Off) and luminous flow percentage of every luminary. The procedure is applied to two inner spaces of the university building: a meeting room and a classroom, from the integration of MATLAB and GAMS. Results show the A E C H I S for all possible alternatives and the description of the technical characteristics for the best control alternative. Also, it can be evidenced that the energy consumption by auxiliary components influences notably on sizing process because represents for majority of cases between 10% and 50% of A E C H I S , the reason why the option with the highest controllability is not the best solution.