Cooling demand is far more difficult to assess than heating demand. In the cooling demand the insolation through windows is crucial and therefore also the behaviour of the occupants of the house. By closing windows and sunshades during the day and opening windows during the night the cooling demand can be greatly reduced. The cooling demand is not very much dependent on the insulation quality of the house (since windows can be opened). So we came to the conclusion that only one cooling demand for all four houses would suffice. The cooling demand was calculated with TRNSYS for the reference house with standard occupant behaviour. The set temperature was chosen at 23 oC. The cooling demand is calculated at 1.4 GJ/year. With a set temperature of 24 oC the cooling demand is only half of that at 23 oC and with 26 oC the cooling demand is almost zero. So the cooling demand is very sensitive to the set temperature. At 23 oC set temperature the maximum cooling power is 2.1 kW.

A high solar fraction for cooling is necessary to avoid natural gas driven sorption cooling (which is for single effect sorption about half as energy efficient as compression cooling). However the cooling demand is far lower than the heating demand and so the cooling plays only a minor role in the energy consumption of the typical houses.

Simulation system

An existing Ecofys program for simulation of a solar system with electric heat pump was extended with a newly written simulation module of a sorption heat pump. The sorption module is based on the zero order model described by Herold [Herold, 1996]. This is a direct thermodynamic model of a single effect sorption process in which only those processes are modelled that contribute the largest irreversibility’s. These are especially the heat exchangers. Single effect was chosen because it can be driven with a standard solar hot water system (temperature < 100 oC). The sorption model is independent of an actual process. It describes a sorption heat pump as two Carnot cycles. By choosing this sorption model we could first dimension the sorption cycle and afterwards find the actual process (water/ammonia or water/solid) which can best fulfil the requirements.

Calculations were made for the Test Reference Year (TRY) of De Bilt in the centre of the Netherlands. The TRY gives hourly values, but the calculations were made with a time step of only two minutes, therefore the TRY values were interpolated within the hour.

The houses were modelled with a low temperature heat delivery system. The delivery temperature is raised proportional to the heating power that has to be delivered (from 22 oC at zero heat demand to 30 oC at maximum heat demand). The maximum heat demand over the TRY de Bilt depends on the type of house: Big house: 10.0 kW, Average house: 5.5 kW, Reference house: 2.5 kW and Minimum house: 1.5 kW.

There was no demand reduction at night (no lowering of the set temperature at night) and consequently no demand peak in the morning.