Solar Shed -- Plumbing and Pumping


The plumbing system is the drain back style.   When the collectors are in the sun, the circulation pump is turned on, and  pumps water from the bottom of the storage tank into the lower collector manifold.   After circulating through the collectors, the water returns to the top of the storage tank via a pipe from the manifold along the top of the collectors.  All of the plumbing (including the collectors themselves) are sloped to make sure that the collectors and plumbing completely drain back into the tank when the pump stops.  For more information on how the drain back plumbing works, and other plumbing alternatives, see the Home Power Basics articles.



Collector supply plumbing arrangement.  This is all 3 ft of it!  The pump is a
Taco 008 with a bronze case -- there are ball valves in the pump flanges to
allow pump removal without draining the tank.
An insulated chamber was added around the pump when this end of the tank was insulated.


The pump must be located below the water line in the tank, so that it retains it prime when the water drains back into the tank from the collectors.  I used a bulkhead fitting (see picture) to make a penetration low in the tank for the pump inlet.  


An alternative would have been to use a double U-tube arrangement.   In this arrangement, the inlet pipe starts near the bottom of the tank, goes up to the top edge of the tank, and uses a U shaped fitting to go over the top edge of the tank, it then descends down (outside the tank) to near the bottom of the tank, and goes through a 2nd U shaped fitting to bring it up to the pump inlet, which is positioned to keep the pump below the tank water line.  It is explained in the Home Power Basics article on drain back systems.


Below the waterline connections to EPDM lined tanks are not considered a good thing to do, but the double U tube does not sound so swell either.  Finding a well designed bulkhead fitting decided the issue for me in favor of the below water line connection, but, I'm not certain its the best choice. 

The bulkhead fitting I used is pictured below.  I got it at my local AG store -- there is no manufacturers name on it.  When I installed it, I put a bit of silicone caulk between the EPDM liner and plywood to seal and cushion it better(?).



The EPDM liner gets
compressed between the thick
rubber gasket and the plywood
tank wall.

The center hole is threaded from  each side
for regular pipe threads.



I used 1 inch copper for all the plumbing lines between the collector and tank.  I also maintained steep slopes to insure complete draining of the collector.  The plumbing on my setup is very compact.  The pump inlet pipe is only inches from the tank, and the pump outlet to collector manifold is only about 3 ft.  The return pipe from the top of collector manifold is less than 20 ft long.  If my lines were longer, and given the price of copper these days, I would have considered using PEX for the part of the plumbing outside the collectors(especially for the supply run.)  But, my runs are so short it did not seem worth the bother.  The one exception I made to 1 inch copper lines was to make the last 3 ft of the return line 3/4 inch copper.  I did this because it seemed that the 1 inch return line was not running full, and I think that for more consistent pump performance, the collector should run full, so that the pumping head requirements are reduced by the fluid in the return line(?)  The short length of 3/4 seems to make the system perform more consistently and with less "gurgle".  The return pipe must end in the air pocket above the water line in the tank.  This allows air to enter the return line during the drain down process.  As long as the return pipe ends above the water line, and pipes are sloped correctly, no air venting valves, vacuum breakers or the like are needed (or desirable, since they are just something else to fail).



The pump I am using for the collector circuit a bronze case Taco 008. 

I used bronze because its basically an open system that allows new oxygen to enter.  The oxygen will corrode iron pumps over time. 

The Taco 008 provides 15 feet of head at zero flow.  The vertical distance from the pump to the top collector manifold on my system is about 12 ft, so the pump has about 3 ft of margin for the startup head requirement of 12 feet.

I need at least 1 gpm per 40 sqft collector, and the Taco is providing 10 gpm after steady flow is established.


The pump for this system is located in an unheated shed, so protecting it from freezing is an issue.  I basically mounted nearly in contact with the bare storage tank wall, and then insulated a around the outside of it with R10 insulation.  I believe that the tank will conduct enough heat into the pump area to keep it above freezing. 

There is also the issue of the pump running too warm inside this insulated cavity when the tank water is hot, the air temperature is hot, and there is no free room air circulation around the pump.  I believe that when the pump is operating, the pump temperature will be mainly influenced by the water temperature, since it has a high conductivity path to the water flowing through it.  But, I also ran some number on heat loss out of the pump cavity with a 50 watt internal power dissipation, and the resulting temperature seems acceptable.  I plan to mount a temperature sensor on the pump just to keep track of it for a while.



A conventional differential controller will be used to turn the pump off and on.  The controller turns the pump on when the collectors are about 8F hotter than the tank, and turns the pump off when the collectors are about 4F hotter than the tank.  The high-limit feature of the controller will be used to stop circulation when the tank temperature reaches 170F.


Gary 10/25/06