Abstract

This thesis investigates the development of predictive control strategies for optimally or near-optimally designed solar homes. Optimal design refers to the integration of renewable energy technologies (mainly active and passive solar) with a high-quality building envelope as well as efficiency and conservation measures to achieve substantial reductions in energy consumption and peak demand. Effective implementation of these technologies requires an integrated design approach, which considers their interactions with the building and its services. Furthermore, control strategies must be an essential part of the integrated design of a building to improve energy performance and ensure occupant comfort. In optimally designed solar homes, control strategies should incorporate the collection, storage and delivery of solar energy. Weather forecasts along with an understanding of the building’s thermal dynamics (e.g., time delays due to thermal mass) enable predicting and managing loads and solar energy availability.

Design and operation strategies of a case study, the Alstonvale House, are presented. Features of this house include passive solar design, a building-integrated photovoltaic/thermal (BIPV/T) system coupled with a solar-assisted heat pump, a thermal energy storage tank and a radiant floor heating system in a thermally massive concrete slab. Design and control approaches developed for the Alstonvale House provided the basis for generalized control strategies applicable to optimally designed solar homes.

Simplified building models, which can be derived from more detailed models or on-site measurements, can facilitate the implementation of predictive control techniques. In this investigation, model-based predictive control was applied to a radiant floor heating system and the position of roller blinds in a room with high solar gains.

Predictive control can also be applied to optimize the operation of renewable energy systems. In this study, forecasts of heating loads and solar radiation were used in a dynamic programming algorithm to select a near-optimal set-point trajectory for an energy storage tank heated with a heat pump assisted by a BIPV/T system.