It has been found out that the monitoring of the system and the analysis of the operation is necessary, even after several years of operation. Since 1993 many lessons were learned and a kind of standardisation could be reached /3/. But a detailed planning is essential for the special needs of every large-scale solar thermal system.

Two typical failures or not optimal operation will be shown in two examples, which occurred at the solar thermal system of the hospital in Baden-Baden:

After recommendations of /3/ for the sizing of heat exchangers a transmission capacity of 100 W/K/m2 collector area is needed. Following characteristic design values are advised: Temperature input primary: 75 °C Temperature input secondary: 30 °C Temperature output primary: 68 °C Temperature output secondary: 33 °C This means a logarithmic temperature difference of 5 K.

The heat exchanger of the collector loop at the Hospital in Baden-Baden was not designed in the recommended way. The logarithmic temperature difference is 8,5 K instead of 5 K.

100

90- 80-

60

50 40- 30- 20

01 02 03 04 05

The measurement showed that the heat exchanger was sized smaller by approximately 40 %. This can be seen in the diagram in Figure 4. The temperature difference occurs up to 15 K instead of 5 K. Especially at sunny days and therefore high temperatures the available capacity can’t be transferred. The small dimension was due to cost reasons. The consequence is a decreasing of the collector efficiency and thus a smaller solar yield. The installed heat exchanger was not replaced because the delivered solar energy was still in an acceptable range and the warranty could be fulfilled anyway.

After two years of operation it was found that the flow rate of the secondary side of the discharging heat exchanger was decreasing constantly and finally stopped (Fig. 5). After the cleaning of the dirt trap didn’t help the heat exchanger was dismounted. There was a strong calcination on the drinking water side. This was remarkable because there was a
special thermostatic valve which avoided temperatures above 60 °C at this part of the heat exchanger. The measurements proved that the valve worked properly.

After the cleaning the solar thermal system worked like before. One year later the same problem occurred, so now the heat exchanger has to be cleaned at least twice a year. Without the measurement the failure would have been very difficult to detect. It shows that even after 2 or 3 years of trouble-free operation a constant check is necessary.

Fig. 6: Simulated and measured solar yields of the six solar thermal systems from 2000­2003

□ Simulation □ 2000 □ 2001 □ 2002 П2003

Fig. 7: Simulated and measured system efficiency of the six solar thermal systems

The relatively low yield of the solar thermal system at the hospital in Singen is caused by an unfavourable concept in the beginning. As can be seen from the yearly increasing gain the optimisation affords were successfully accomplished. This was for example a better adoption of the control concept and the change of the piping of the hot water storages.

As heating storage tank old hot water storages were used. The place of the connections of these old storages were not optimal which means almost no temperature layers can develop inside the storage.

It is remarkable that the plant at the student residence "Vauban” has reached high solar yields. The reason for this is a low solar fraction. The real hot water consumption especially in the semester brake in summer was not as high as predicted. Therefore the utilisation of solar energy also in times of a high irradiation is guaranteed. The disadvantage is a low solar fraction of app. 17 %. In Figure 8 and 9 the effect of the hot water consumption on the solar yield and the system efficiency can be seen.

Fig. 8 and 9: Influence of the specific load on the annual solar yield of the six plants monitored by the Fachhochschule Offenburg. (Each point in this figure represents a monitoring period of one complete year of operation.)

The specific solar energy price ranges between 0,107 €/kWh and 0,160 €/kWh. The exceeding of the maximum predicted costs of 0,13 €/kWh is due to a tolerance of 10 %, different weather conditions and changes in the consumption. The prices are still twice as much as the conventional energy cost. Including further optimisation potential in costs and solar yield and also the increasing of the costs of fossil energy, the solar technology especially in large-scale system gets competitive.