The following points have been important for me

• large area of cheap collectors
• simple hydraulic system
• simple Brennwert-heating (don't know the word: a heating, which uses the energy of the exhaust gas), which is controlled by the collector controller
• good external heat exchangers
• good controller
• Use of my "pre-installation"

The large collector area makes it possible to use the sun to support the heating mainly in the transition period and even in winter. If you choose a small or a large collector area: The costs for the equipment placed in the cellar are nearly the same. If the price for a square meter of collector is not much higher than the grant by the government which can be expected, it is a good idea to choose a large area.

In my case the use of standard pumps, standard mixers and standard slidegate valves makes a sustainable operation of the system possible. In case of a defect of one of these parts I can buy a new one in the next market.

As shown in a test of "Stiftung Warentest" all heatings (Brennwertkessel) are close together and they all are good. Since the heating should be regulated by the central collector controller, there is no need for high sophisticated  burner add-ons. A simple burner which can be easily maintained is a good solution here.

I think it is a good idea to use a simple storage without inner mobile parts (in case of trouble there would be no chance to get into the storage to fix it). Within this storage you will find "dead" (oxygen-free) heating water only and therefore it should not rust through. There is no need of a sacrificial anode or electrical measures. And you will always receive fresh warm water, which can be used for further cooking (tea, rice, noodles,...), so that the time for cooking will be shortened. You can do that with a normal boiler as well, but I didn't like it any longer, when I saw the mud within the boiler once I exchanged the anode. Perhaps I'm too sensitive in this case.

The be-all and the end-all is a good device to control the system. I use an UVR1611 of the Austrian company Technische Alternative and I think it is good (but I don't know many others ;-) ).

When I fitted out the attic in 1995 I laid copper tubes an insolated them, since I didn't want to move out all that stuff in the attic at a later time. Perhaps I had made three errors from current view:

• I had used solder. Following a discussion in a forum (Haustechnikdialogforum) it might become a problem with the soldered connections in summer with high temperatures when the collector comes to a standstill. The collector which I selected has got standstill temperature of 189°C (= 372°F) and this should be within the green range. Furtheron the temperatures should exist in the tubes only on the first meters and only for a short moment.
• For insulating the copper tubes I have used good Armaflex-SH, which was recommended to me in 1995. It can be used up to 120°C (= 248°F), but better would be Armaflex-HT, which can be used up to 150°C (= 302°F). The highest temperature I've seen up to now (June 2007) was a short peak of 135°C (= 275°F).
• Perhaps the cross line of the copper tubes was too scarce. First there was a frowning, then they computed and it was okay.

So I was in doubt as to wether it could work and I started to think of new solutions and new routes for the tubes or to stop the hole project. After a discussion with the solar fitter I decided to use my old work, which should not have been in vain.

Last I selected

• a thermal collector system of BUSO with 19 m² collector area and 1000 l storage
• an UVR1611 controller of the company Technische Alternative
• a heating system ("Brennwerttherme") Modula II von Paradigma
• the company EUT Haustechnik in Niederzier, who did the planning and execution

I would have liked to use a larger area on the roof, but there is a window in the middle of the roof, which prevented this. The hight and the space in the cellar and the width of the door were limiting values too. And last but not least I had to ask my purse because I started a project with an uncertain outcome and the costs should stay within the bounds of possibility.