One disadvantage of that is that it would become intermittent - it would whoosh warmer liquid down at 18 litres/min , then because the delta-T between the Sunbox and the down flow was no longer good, it would turn off.... then a few minutes later, it would come on again, have another whoosh, and so on. I am sure that this is not the reason for the solenoid valve failing because I presume it was designed for frequent use, but it seemed inefficient.
The 'Trickle all the Time' circuit is less dramatic, because it is injecting warmer liquid at only 5 litres/minute, but this means that it does not warm up the down flow enough to stop it. I hope that over the long term, this will result in more kWh dumped - because the delta-T will be favourable for a longer time. With the higher temperature delivery from the evacuated tubes, this type of connection will be the only practical one. The existing sun box pump is now working at 45 watts, and I have yet to find out what this will work out at litres/minute. I will know after a week of averaging the flow volume and clocked hours.
|My next task is to clean up this arrangement to make the|
simplest possible connection for the tubes circuit
The summer time performance will probably be the same with either system, the diurnal may be better, but I am curious about the winter performance. In Winter, with the entire ground loop going through the panels, one can see the direct effect of a sunny winter day. I was watching the GSHP working this morning with the new circuit, and recognise that the downflow temperature isn't vastly increased anymore because it is receiving only a trickle of warmer liquid that doesn't change the visible down flow temperature much. That's less 'dramatic', but it is consistent, and there is good reason (better delta-T) to think that winter performance might improve. With the liquid moving more slowly through the panels, I hope it will have time to get more warmed. Providing the piping is well insulated, the system may be more efficient.
I am glad I have done this 22-month trial of the Trickle and Whoosh method. If I had not, I would have wasted acres of blog-space speculating on how much better it might be - I now know that it is not as good, and it not as reliable when a maintenance issue arises.