A weather data, which generates auxiliary energy results that are closely 365/12 times smaller than the annual one, has to be selected.

In contrast to this previous test method, the days of the core phase are chosen with different criteria. They are chosen in an iterative process where the predicted simulation results of the monthly average values of the internal storage energy and the space heating demand are compared to their corresponding annual values.

Two combisystem simulations are necessary for the development of the test sequence weather data. These are an annual system simulation and a 12-day system simulation. These simulations are carried out using TRNSYS [7]. All system parameters, depending on the annual cycle, have to be

adapted to a shorter 12-day simulation cycle. Furthermore, domestic hot water draw-offs and the space heating demand should have realistic values in comparison to the annual one.

Comparison of annual and
predicted performance

Fig. 1. Iterative process used.

Afterwards, a prediction calculation is necessary in order to be able to make a comparison between annual and test sequence simulation. This comparison should show a good agreement in the space heating demand, the domestic hot water demand, the mean internal storage energy, the solar usable energy and the fractional solar consumption. A methodology for choosing the days is done with an optimization process written in Matlab [8]. It is based on an algorithm that searches the best days out of an annual reference weather file with regard to the defined quality criteria.