Greetings... new member here. I hope someone can help with this.

MY SETUP:

I have a hydronic radiant floor heating system. 3 zones; basement, 1st floor, 2nd floor. 1 pump per zone. Heat source is 50 gal Polaris water heater. Pex tubing is mounted under subfloor with aluminum plates, reflective material, and insulation. The system is "open" in that the heater is also my domestic water heater. A mixing valve diverts hot water to my Manabloc water distribution system. Hot water from sinks is replenished into the system from the cold supply and runs through the floor before going into the water heater. It's actually a very efficient system when it works as intended.

The hot water supply comes out the top of the tank, goes down 16" and to my three pumps. The return tubing is configured to basically run downhill all the way to the heater. It has no uphill path at all. It returns at the bottom of the tank.

THE PROBLEM:

I've only fired up the system for the first time this winter. Under normal operating conditions, the system works great! The system cycles on every 3 hours or so when outside temps are below freezing. The rooms are VERY comfortable. However, if the outside temp increases and the system is not called for heat for an extended period, let's say 8 hours, the cold water in the tubing wants to move down (naturally) and go into the water heater. This causes the heater to start cycling every 10 minutes to recover the set temp on the heater. This is a tremendous waste of expensive propane. AND if it's doing this in the Winter, what will it do in the Summer when no heat is needed at all?!

WHAT I'VE TRIED ALREADY:

I've installed a spring check valve above the water heater on the hot side, but this is not enough to stop the natural water flow.

WHAT DO YOU THINK??

Before I go about reconfiguring my system, I wanted to see if any of the experts out there could offer some insight. If I created an uphill path to the water heater for the cold return lines, would that be enough to stop the cold water from flowing downward or would the shear volume of cold water at the higher elevation easily overcome the force necessary to go back uphill? What about more spring check valves?

Any help would be greatly appreciate! See the attachment for a picture of the system.

Thanks,
Doug

Install a zone control valve on the radient loops.

Install a zone control valve on the radient loops.

That is a good idea. But because of the nature of the system being "open" to the domestic water, all three zones must be allowed to freely circulate water when hot water is used at sinks and tubs. Otherwise, I would risk stagnant water in the system that might be unpleasant when it eventually comes out.

However, you did give me an idea. If I replaced the swing check valve sitting atop the mixing valve with a single zone control valve, I think your idea would work. I'd need for my I-link control unit to open this valve whenever each pump comes on. I'm not aware of this devices capabilities so I'll have to research that.

I'd still like to hear if there's a possible solution that uses the force of gravity instead of electronics and valves.

Thanks!
Doug

Let me think about it Might have some questions

I've installed a spring check valve above the water heater on the hot side, but this is not enough to stop the natural water flow.

The spring check should have worked since it should not let any water back flow into the unit. I believe a thermal check is also available. You might also try moving the spring check away from the hot water source a bit since I have heard of hot water creating a problem with some of them.

http://www.heliotropethermal.com/solar_check_valve/solar_check_valve.htm

http://gothermal.thomasnet.com/viewitems/hydronic-controls/ball-cone-check-valves?&forward=1

"The spring check should have worked since it should not let any water back flow into the unit."

I should have been more clear on where the water is flowing and the purpose of the check valve. The spring check valve's purpose is to provide a small amount of resistance in hope of stopping the natural convection of hot water when the system is idle. It's not there to prevent backflow. I'm using the valve in an unconventional way.

How about a check valve at the discharge of each pump?

When all else fails install a snap selonoid valve. That is a positive shut off and nothing will flow back.

I thought it was "If all else fails, read the directions".

Your solution makes a lot of sense. Normally closed valve connected to pump power. If it can't get through the pump, nothing will flow.

Directions? We don't need no stinking directions.

That's an incredible picture which you've posted before.

The only thing I can see not quite right is that the floor shoul have an epoxy coating. In some buildings, you can eat off the mechanical room floor. In general, you can't.

It is sealed with a clear epoxy but they forgot to add the fine grit mix and had to do it over. You will really bust your A$$ on it the way it was.

And this is different from a zone control valve, how??

And this is different from a zone control valve, how ???

Yes different.

Depending upon the brand of zone valve it might not be a positive shut off. A snap selonoid is a positive shut off device.

Basically, if you prevent the displacement of a fluid, the fluid can't move. The pump when off is leaky. A zone valve may be leaky.

Positive shutoffs at both ends would be troublesome because of possible freezing issues. If the shutoff exists at one end then there is still a place to expand too.

<snip> If I replaced the swing check valve sitting atop the mixing valve with a single zone control valve, I think your idea would work. I'd need for my I-link control unit to open this valve whenever each pump comes on. I'm not aware of this devices capabilities so I'll have to research that.

I'd still like to hear if there's a possible solution that uses the force of gravity instead of electronics and valves. <snip>

If that circulator control box has the end switch contacts available, it could be done with the snap solenoid valve hvac1000 recommended. I don't think it could be easily done without control valves. Thermosiphoning is hard to beat when it has a head as high as the second floor. If you ran your cold water returns down to the floor and then back up into the top of the water heater it would provide more of a heat trap, but probably not enough to overcome 12 to 15 feet of head. There are figures out there to calculate what you need, I can't remember what they are or where to find them any more.

Really, what you need to do is keep that return line as warm as the feed line. If there is no temperature differential across the zone supply and return lines, with a bit of heat trap in the basement on both lines you can stop the thermosiphon. Or you could work it so that your supply lines (from the circulators) get cold BEFORE the return line. Then the thermosiphon would want to run backwards, toward the check valve. Of course it would meet heat somewhere near or at the check valve, you would need a good thermal break in that line. (It just occurred to me, hopefully there is no path through the mixing valve around that solenoid valve.) Bundle the supply and return together inside the same insulation. Note added: {or take the insulation off the supply and superinsulate the return line} scratch the section in brackets. Occasionally in the non-heating season you still will have problems with that because of the cold feed and warming of the uninsulated line.

Another note, just remember, any valves and whatnot you use must be for potable water. Heating system parts will not work long term, they are not designed to work in constantly changing water. You did use bronze circulators didn't you?

EPM

<snip>Positive shutoffs at both ends would be troublesome <snip>

Can't have positive shutoffs at both ends. He feeds his cold supply water into the system just after the first shutoff. It has to be able to flow through everything in order to prevent stagnation. That back check above the mixing valve (or something that does the same thing) MUST be there too because of the restrictions in the cold supply to the water heater.

EPM