Abnormal heating graphs?

[srudin]

We have a solar hot water system and use it for about 3 years now. The main parts are a German tank, a Chinese panel and a Chinese controller. It was installed by a local company in Chiang Mai (Thailand) and worked quite reliably so far. Please note that the tank and the panel are at very close distance on the same (ground) floor, maybe 2-3m of pipe for each connection (in/out).

A few months ago we exchanged the controller to a wifi model that I can connect to my HASS system. It works pretty much the same as the previous one but allows me to remote control and automate the system as well as collect some data.

Now this collected data is somewhat confusing me. Let me show a number of graphs that I got out of our system:

---1---


Looking at graph 1 I see more or less what I would expect: After sunrise the panel (T1) heats up and triggers the pump; after every pumping cycle the tank (T2) gets a bit hotter. However, the graph flattens slightly down - meaning that the hotter the water gets, the less the heat increases in the tank per pumping cycle. Oh, and please ignore the temperature increase between 2-3pm, that is the integrated heating module.

---2---


Looking at graph 2 I see the opposite of what I would expect: After sunrise the panel heats up and triggers the pump; but after every pumping cycle the tank gets a bit COOLER (?!). Only after around 10 pumping cycles it starts to slightly heat the tank.

Please note that the main difference between the two graphs is that the tank temperature in 1 starts below 30° while in 2 it starts above 45°.

In other words it seems that:
a) water heating is only effective below 40°
b) the first couple of pumping cycles seem to be ineffective or even cool down the tank
c) if the time between two pumping cycles is too long again the pumping cycles become ineffective or even cool down the tank
(Please note that besides the 2 uploaded graphs I have a lot more underlining these statements. I believe the 2 posted here basically show the idea though.)

My question:
I can understand that the water in the pipe cools down during the night or when the time between pumping cycles increases. I can also understand that the water cools down more the hotter it is as cooling happens relative to the environment. But we have around 30° air temperature during the day and 25° during the night here so the cooling within 2-3m of isolated pipe should be minimal, no? I could accept the first pumping cycle to cool down the tank a bit as it will pump the leftover water in the pipe from the night into the tank, but subsequent pumping cycles should no longer show that behavior, no? Are there other people out there having access to this kind of data? What are your observations?

Related question:
We are mostly using green plastic pipes with some isolation material around it. However, the connection out of the panel back to the tank has been replaced by a copper pipe (to handle extensive heat). I have limited scientifical knowledge but isn't metal more sensitive to temperature than platsic and passes on temperature differences quicker? I mean with an outside temperature of 30° doesn't a metal pipe cool down the 40° water inside quicker than a plastic pipe would? If yes, could this be the reason for the observations above?

 

[GXMnow]

From what I see in the graphs and what you wrote about the system operation, it sounds like the cold water left in the pipes is just cooling off the solar panel as soon as the pump turns on and then it can't heat the water in the tank.

The panel is getting hot, but it cools fast when the cold water in the pipe is pumped into the panel. The panel temp quickly drops well below the tank water temp, and cold water is going into the tank.

The only ideas I have are wrapping the pipes with thermal insulation to try and keep them warm, and maybe add a bypass valve near the tank to circulate water through only the pipes and the solar panel until the pipes warm up before switching the valve to pump the water to the tank. This is certainly adding more complexity, but it may improve total efficiency. The idea is that when the panels get warm in the morning, start pumping water in a loop of just the pipes until it all heats up hotter than the water in the tank. This will delay the water heating a little, but it will keep the pump from dumping a slug of cold water into the tank from the cold pipes.

 

[srudin]

The water in the pipe from the panel to the tank of course is an issue - but I would assume that the water in the pipe from the tank to the panel should not be a problem? I mean once it gets into the panel it should heat up no mater what the temperature before was, no?

The controller is set to start pumping once the panel temperature is 10° higher than the tank temperature and stop once the difference is less than 5°. However, according to graph 2 the panel temperature gets BELOW the tank temperature during the first couple of pumping cycles. Maybe some pumping settings are wrong? Or could it be a problem that we have only 2 temperature sensors (panel, top of the tank) and not 3 (panel, top and bottom of the tank)?

I would say that the pipes are pretty well isolated already. Cycling the water first would be an option but as you said it will make everything way more complicated. My feeling is that 2-3m of pipe can contain only a limited amount of water and I would expect this water to be flushed into the tank after the first cycle. So my confusion is not so much the temperature drop after the first cycle but why the temperature still drops on cycles 2-10.