Solar Drainback

Housedoc

Greetings,

I have a perplexing situation with a drain back system. The system uses a 15 gallon stainless drain back tank on top of two 80 gallon non pressurized storage tanks. The attached drawing shows the piping configuration of the solar loop. I have omitted the demand side piping for clarity.



The problem - the solar pump is somehow getting air in the inlet. It happens randomly sometimes a week apart sometimes a day apart. The pump just sits there and stops pumping. You can unplug the controller, wait a couple minutes and plug it back in. The pump takes off with no problems. I cannot figure out how air can make its way from above the water line all the way down to the inlet of this pump.



Any ideas ????

Mark Eatherton

With that many ups and downs...

it may not have ever been properly and completely purged.



I'd drain it, then back fill under high pressure, then adjust water level to the site gage.



Wouldn't be the first system that never got properly filled and purged :-)



ME

Housedoc

Thanks for the input Mark

The system must have air above the water line to drain back properly. So I don't need to purge the air from the circuit. I just need to figure out how its getting into the pump inlet. If I purged all the air from the system, it would be left in a vacuum and the water would not drain from the collectors.



I have an engineer friend who has told me something I did not know. The dip tube in water heaters has a hole drilled in the side at the top of the tank to keep air at the top of the heater. He is 90% convinced that the water from the drain back tank spills into the hot port of tank A and brings entrained air with it. The air rises to the top of the "unpressurized" storage tank and is sucked into the dip tube via the hole. From there, it moves to the top of the 2nd tank where it migrates again to the top. Then the pump energizes and pulls the air from the hole in the dip tube of the 2nd tank. Then it moves to the inlet of the pump where it creates the problem.



There is also a concern that the flare at the top of the dip tube lays loosely in the nipple. The fear is it will move up and air will fill the nipple just like the hole in the side of the dip tube. I can see the physics. But it would appear to require a lot of planet alignment to work as described. Then again, I cannot come up with a better answer.



What do you think?

Mark Eatherton

The air has to be coming from somewhere...

And IF the DB tank is properly sized and piped, the solar loop pump shouldn't be sucking air. I suppose it is possible that your pump is creating enough of a vortex that it could exert impellance and cavitation on the impeller, but I'd make absolutely certain that circuit is properly purged before I head down that path.



Who's tanks, and how are they plumbed? How is the system configured in general?



My assumption is that you are using tanks with immersed coils. These coils are piped in series. Are you taking the return fluid from the array into the bottom of the first tank, then out of the upper HX tapping, and into the lower HX tapping of the next tank, then out of that HX tapping and into the pump?



Also, 42" of static positive suction head on most pumps is right on the naked edge, and when the pump starts up, the NET positive suction head will drop due to water level drop in the tank, and the pressure drop of the fluid resistance through the circuit. I completely missed that in the first pass... As the solar fluid gets hotter, the potential for steam flash and cavitation at the impeller increases.



The holes in the dip tube are there to act as an anti siphon feature, just in case the water supply fails and tries to drain the potable side of the system out. But again, I need clarification of how this system is pumped/piped.



In order to do a proper purge/fill and flush, you'd have to add another drain cock to the loop so you could force purge all fluid out of the circuit, then power purge the loop to guarantee no residual air is in the circuit. If the tank is properly sized, and the circuit completely purged, the eye of the impeller should be seeing a good constant suction pressure of around 60" WC, and should not cavitate.



ME

Housedoc

Diagram

Mark,

Take a look at the attached drawing in the initial post. It shows the piping schematic et al.

Mark Eatherton

I saw that...

I didn't see any output to DHW or to space heating. How are you accomplishing that?



Whose tanks?



Do they have heat exchangers inside of them? More than one HXer per tank?

Housedoc

System set up

The tanks are American Water Heater 80 gallon electric heaters. The elements are not connected. There are no heat exchangers internal. The hot solar storage water comes off the left tank from a port on the top. It moves through a plate exchanger at the same time as cold water is pulled from the bottom of a third water heater which is potable. Solar heat is exchanged in the counter flow plates and provides hot water for the house. The solar water goes back to the cold tank to the right. The potable water heater and heat exchanger work perfectly. I left them out of the drawing for clarity.

Mark Eatherton

Got it...

It's now clear in my minds eye.



It all goes back to my previous statements about not being properly purged and not being able to maintain the required NPSH to the main solar circulator.



In order to guarantee that their pumps will perform to spec, most all circulator manufacturers require a minimum of 5 ' of net positive suction head. In your case, as you begin drawing water out of the reservoir tank, the NPSH drops, and just as the fluid is almost to the top of the system, and establishing a siphon on the downcomer, the pumps poops out due to lack of NPSH, and begins cavitation, hence air binding.



In general, the O2 that is in H20 doesn't just jump out of suspension. It needs heat and negative pressures to give it incentive to jump.



In order to guarantee a total and complete purge of the solar loop through the storage tanks, I would add a ball valve and a purge cock between the DB tank and the first solar tank in the loop. This way, you can use street water pressure (and associated velocity) to force purge that loop and guarantee it is bubble free. I would pump the water in at the DB tank connection and purge it out near the circulator.



You also need to calculate the NPSH and make sure that as you draw fluid out of the tank to fill the loop and array that you maintain the manufacturers minimum required NPSH. Worst case, you may have to install a larger DB tank to achieve this.



ME

Mark Eatherton

PS.....

If the holes in the sides of the dip tube really bother you, you can eliminate them by either making your own dip tube out of copper, or use an alternate tapping, like the drain tapping to get a low connection to the tank, eliminating any possibility of trapped air getting in your face.



There is ALWAYS a way :-)



ME

hot_rod

a couple thoughts

could you remove the drain cock, install a brass tee and locate the pump down low instead of pumping up through the dip tube?



Also pressurize the entire system to 30 psi or so if possible. That assures your NPSH and helps preventing flashing when the water hits a hot absorber.



hr

Housedoc

Thanks guys

I am moving the pump inlet down to the drain cock, taking the dip tube out of the picture. The pump is supposed to be fine with 18" of inlet pressure. Typically that's all they have with a stand alone drain back tank.



Then I have to add means of purging air from both tanks. My idea was to install a stand pipe off the top of both tanks with a boiler drain at the top, above the water line. This way I can be assured the tanks hold nothing but water. Then I wanted to add a boiler drain above the water line or in the top of the drain back tank to make sure we have an air charge to facilitate drain back.



Hot Rod, you lost me on the pressure idea. I think it has something to do with raising the boiling point. Right?

Fortunat

Sucking air

Your engineer friend's descriptions sounds plausible.



What's the water level in the drainback tank when the system is operating and how is the water returned to that tank. If the water level is low or the return water is entraining air in the tank I could see how it could make its way around to the pump.



~Fortunat

hot_rod

The higher the fluid temperature

the more NPSH you need.  The pump manufacturers should have that info in their tech sheets.  They usually indicate the pressure required at a few different temperatures.

  Yes the idea is to increase the boiling temperature, you can look up those relationships by Googling boiling temperature of water under pressure.  Ideally you want to see 250- 300F, as the collector will stagnate at those temperatures, maybe higher for evac tubes.



hr

Housedoc

Thanks everyone

Thanks for the great feedback. I rearranged some piping today following the suggestions. Solar pump is quiet as a mouse. No air. I added boiler drain above the water line to infuse air and slightly pressurize. Drain back is also quicker. Thais again guys!