Control Problem on $1000 Solar Water Heating System

The plot below shows the temperatures of the supply and return water for the PEX $1000 Solar Water Heating System collector.


Note that around 3:30 pm the sun intensity is such that the water going through the collector is no longer being heated at all.  The controller should switch the pump off before this happens, but instead, the pump continues to run for nearly an hour, and during this time it is actually cooling the water in the storage tank because the water is coming back from the collector is cooler than the water going out to the collector!


The way the system is supposed to work is that there is a sensor in the storage tank to sense tank temperature, and a 2nd sensor in the collector to sense collector temperature.  When the collector sensor is hotter than the tank sensor by about 15F, the controller turns the pump on.  When the collector temperature drops to within about 4F of the tank temperature, the controller turns the pump off.  Clearly the 2nd part is not happening in this case.


This is what what was causing the problem:

- The tank started out fairly cool in the morning (about 95F).


- As the day progressed, the hotter water returning to the top of the tank from the collector progressively warmed the tank from the top down.  The area near the top of the tank eventually warmed to about 115F, but the area near the bottom of the tank stayed about 95F.


- The pump intake is about 7 inches off the bottom of the tank, but the tank temperature sensor was right near the bottom of the tank.


- By late in the day, the pump intake has moved into the warmer water in the tank, but the tank temperature sensor is still in the 95F water at the bottom of the tank.


- The controller thinks that the tank temperature is 95F, but the pump is actually drawing from an area of the tank that is more like 115F.  The 115F temperature water being pumped out to the collector is actually keeping the collector sensor up toward 115F (even though the sun is off the collector, and the collector would be down at 50F if not for the hot water being pumped into it).  So, the controller thinks the tank is at 95F, and the collector is at 115F, and is happy to keep pumping.


- At about 4:30pm, I realize what's going on, and when I raise the tank sensor up to the level of the pump inlet, the controller immediately shuts down the pump.

This really surprised me because I did not expect much temperature stratification in the tank.  I felt that the fact that the pump takes water  directly from the tank rather than using a heat exchanger would keep things mixed up.  But, in this case there was a 20F temperature differential between the top area of the tank and the bottom area.   Measurements taken at various levels indicate that there was a very sharp temperature transition near the bottom of the tank, with the temperature fairly constant above an below that transition area.


Now that the problem is corrected, the temperature stratification should allow the system to operate a little more efficiently.  The water going to the collector is a bit cooler, which results in a lower average collector temperature and lower losses, while the tank top water is a bit warmer to provide hotter water to the house sooner.


Moral: Mount your sensors carefully, and then BE SURE to check that they (and everything else) are actually working the way you think they should be working.

I wonder how many systems are out there running day in and day out with these sorts of problems that are not detected because no one checks?





Gary November 19, 2008