The traditional pattern of energy consumption for heating a greenhouse shows two peaks; one at dawn and one at sunset. When large areas of greenhouses are supplied by district heating from a combined heat and power station these peaks create difficulties for the supply system, because the pattern of energy consumption is very similar for all the greenhouses. This can affect the power station's ability to meet the simultaneous demands for its supplies of heat energy and electrical power at peak periods. Strategies for levelling out the heat energy consumption of greenhouses are therefore essential in these circumstances. This paper deals with the development and testing of computer algorithms designed to spread the energy demand of greenhouses, with consequent shaving of the two peaks. Ornamental pot plants were grown in two greenhouses, employing a commercial climate computer to control room temperature, integrated room temperature, supplementary light and inlet water temperature to the heating system. One greenhouse was maintained at equal day and night temperatures, as a reference. The other employed a low day temperature/high night temperature regime (negative DIF). Physiological background to the environmental regimes employed is outlined. The negative DIF algorithms developed were able to level out the energy consumption peaks - transferring the use of energy to night time - while producing plants of unimpaired quality and unchanged production time.
