Active control of split system domestic solar water heaters

Abstract

Solar water heaters have the potential to make large savings in greenhouse gas emissions in Australia. Long financial payback periods are the main reason that uptake of solar water heating is not more significant. This thesis investigates the potential improvement in performance of split-system solar water heaters by the addition of an active control system. This work builds upon "low flow" collector circulation theory and addresses the poor control available from the storage tank thermostat. Modelling suggests that the thermal efficiency of the water heater can be improved by about 25°/o, primarily through reduction of tank standing losses, if the thermostat is replaced by a smart controller. Auxiliary energy consumption is reduced proportionally. If realisable, these savings recover the capital cost of the additional controller in several years. The consumer will benefit from further savings in auxiliary energy consumption over the life of the system and so the payback will be more attractive. The active control strategy is based upon predicting and controlling the energy content of the storage tank. The control strategy is energy tariff sensitive and may be set by the householder to behave in an energy efficient or a cost effective manner. A number of technologies and design improvements regarding forecasting of the energy supply and demand were also developed in this work. The auxiliary heater was moved outside of the tank and placed in series with the solar collector via a switching valve arrangement. The collector circulation pump was also used to circulate water through the auxiliary heater effectively providing a variable volume, variable temperature thermostat. A new variable power pump controller was developed for the existing circulation pump to allow fine temperature control of water returning from both the auxiliary and solar heat sources so that disruption to thermal stratification in the tank was minimised. The predictive performance of the collector could then be decoupled from the state of the tank. This thesis explores a practical implementation of the active control strategy and provides an insight into the actual performance and areas of sensitivity of the technology. The proposed design changes require more thorough validation including field trials to evaluate the load learning algorithms. Performance of the active controller would be improved if the heating circuit intake position could be actuated vertically within the tank or if hot and cold water could be fully separated in the tank.