Pumping vertical in An open system

I think installing a flat plate heat exchanger after the boiler in this situation would be advisable, less trouble to go up and over with a pressurized system. A check valve in the pump would prevent the water from running back into the boiler when the pump is off but it's not as reliable imo.

Not a big deal. If you can mount the pump feed and the return line below the water level in the Garn there are no issues other than properly sizing the pump.
I would suggest a circulator with a higher head rating than something like a Taco 007. The actual size depends on the heat load you are
moving heat to and the piping size.

You mean to say a "circulator" not a "pump".

As long as you maintain a flooded suction condition, the circulator is properly sized
to work against the head pressure to be encountered in your heating loop.

There is no reason you cannot use a "circulator" with a check valve cartridge to
do this. You just have to be sure to have flooded suction with the "circulator"
at ground level that is all you need to do besides choosing what size "circulator" you wish to use.

You also need to improve your drawing if you want more help. A circulator can push water
for miles; it all depends on the pipe size, desired flow and pressure and whether it is a
flooded suction system with a check valve.

You can access The bell and Gossett home page and determine what size circulator
you need for the heating loop "after" you decide what size tubing you are using.

Do not forget that your water circulation distance is more than double the actual total length
when factoring in heat exchangers and in floor heating loops and Modine heaters.


You have much more work to do with this before you can make any determinations about any
equipment controls and the like.

Make sure you put some air vents in at the high points, to let air out when needed. Those are great places for airlocks. Something manual I think - not something that scrubs air all the time. Those have potential to let air in also.

No way you can go under the windows? Might simplify if so.

I will suggest keeping the check valves out of that circulator, and have the circulator pushing upwards. this makes any air in the lines move thru quickly without air locking the circulator. properly sizing the loop with the circulator and piping will help. several things: keep open flow paths as simple as reasonable, allow for manual air elimination at downward corners where air will collect, size the circulator to push 3-4 FPS thru the loop piping, which will allow the fluid flow to entrain the air and drag it out of the loop.
generally, open loops are fine below the fluid level, but if above the fluid level, should be a single fluid path. these aren't hard rules, but will make your life easier.

k

WIth both lines submerged, air vents and check valves should not be needed. Air vents can tend to leak and all air is pushed out of the system. If the circulator is adequate, it will always push through a simple system like this.
Check valves are only going to add pressure drop. I would only consider one if the system was thermosyphoning. This appears unlikely in this installation.

A circulator with a check valve will prevent air
bubbles with a flooded suction piping set up.

You have to be concerned with the total head pressure created by the system against the circulator
which is the entire length of the circuit and any appliances(valves, water to air heat exchangers,
flat plate heat exchangers, in floor heat systems connected to it.

Why not just invest in some square hydronic pipe so you can run it along the floor or under the windows and you can paint it any color you wont have the mess your contemplating now nor will there be a chance it could be damaged as you could install some 2 by 4's to protect it if you want to cover it on the two exposed sides being the top and front.

If your investing in a Garn or have invested in one a few lengths of square hydronic pipe and elbows will not matter in the scheme of things and a covered pipe run along the floor with thick lumber can be made to look nice too with paint.

The hydronic tubing is available on this side of the pond so no worries there either and its soldered like copper pipe.

Depending on where you are you may have to go to a very large plumbing supplier to order it or ask if they have it in inventory. Don't bother with the big box stores unless they sell flat panel radiators from Europe to buy the square tubing.

They have used square hydronic tubing for many years in Europe to avoid issues of freezing pipes in wall cavities in exterior walls and they go up and over doors reducing leaks to a bare minimum with it rather than under the floor like we foolishly do here.

Don't guess. Calculate. Here is a head loss calculator where you enter required GPM, length (just vertical), inside diameter and a coef of friction.
http://www.engineeringtoolbox.com/hazen-williams-water-d_797.html

We did this to go over a doorway and our tanks are vertical so the supply/return lines are about 8-9' above the level of the demand pump. Since this is the highest point in our system we have several air vents on our pressurized system. But, unless I'm missing something, the only additional head is due to the typical friction and going thru fittings and extra length of piping. I like to think of this problem as loop of chain hanging on a sprocket. The weight of chain going up is equal to weight going down, or no "head" other than friction losses. Unless I misunderstood, it seems the OP was concerned about requiring a bigger pump to go over the window. Other than line and fitting losses, vented or pressurized, whatever water goes up an equal amount is going down (no head loss) in a closed loop system.

I can't see any difference between pressurized or non-pressurized systems regarding pumping a vertical loop of water. I'll let the pros correct me if I'm missing something. Thanks in advance.

The total height you move the water is the biggest factor in head loss calculations. If you go down 5 feet, up 6, then down 5, you really only went up 1 foot. Sure, there is more friction, but for a small loop system, it will not matter.

In a close loop system, your static head is basically nothing.

I just read this as a refresher, because I forgot a great deal of fluid dynamics. It's been 20 years since I took that!
http://www.michiganair.com/newsletters/2008-2/section4.pdf

This system as described is very much like a closed loop when both ends are submerged as in the Garn.
The catch is if and when the system gets air in it. Then the circulator pump has to overcome the vertical head and frictional head of the installation.
I have done these systems a lot since we make unpressurized tanks.
They have their quirks but once set up properly, they work well.
A couple caveats: if the water is not treated (it is treated in the Garn) then only use non-ferrous pumps. Do not trust air vents in the situations we are talking about. They can suck air into the system if it goes under a vacuum. Those high spots are areas that are under vacuum when the pump is not operating.

That's why I humbly asked, being unfamiliar with open systems. In my mind I visualized the open tank at the top of the loop in this discussion. But the line that runs over the window ends up above the level of the open tank.... which means water can't drain towards the closed, demand side of the circuit, but sure can back to the open tank. Thx Tom. I assume in a Garn system, the boiler is the only vessel vented? I can see how an open system could get tricky. I'll return to spectator mode.

There is no need to vent this. The system is basically functioning as a closed loop as long as both the feed and return are submerged or taking and returning water below the water line in the Garn boiler. In this case a vent it not needed.
It is important to have a pump big enough to match the vertical head and head loss of the rest of this loop.
You can easily determine the frictional head loss with online calculators. The vertical head is about 8 feet which is added to the frictional loss of the heat distribution system. This is all the piping to and from the Garn and other hardware you are pumping through.
It will be bigger than a taco 007!

tom in maine said:

There is no need to vent this. The system is basically functioning as a closed loop as long as both the feed and return are submerged or taking and returning water below the water line in the Garn boiler. In this case a vent it not needed.
It is important to have a pump big enough to match the vertical head and head loss of the rest of this loop.
You can easily determine the frictional head loss with online calculators. The vertical head is about 8 feet which is added to the frictional loss of the heat distribution system. This is all the piping to and from the Garn and other hardware you are pumping through.
It will be bigger than a taco 007!

Click to expand...

Wouldn't some manual venting be required at the high spots to get rid of airlocks - at least on start up?

IF everything is right, it will self purge, but a couple of ball valves with a hose bib between them on the inlet side of the circulator. you can then use a garden hose to push water back thru the pipe into the GARN, and then forward thru the circulator and piping to flush air out. that will get most of the air out, and then the circulator will have a good shot at getting the rest of the air out.

I have a system that will flush out the air in a system 20 feet above the top of the open tank. it takes 2 taco 007's in series, but it does it. it runs thru 6 4x8 foot solar hot water panels. it's been sitting there running every day with no attention for 5 years.

there is no need to worry about reading more fluid dynamics, Invest in 2 paperbacks, "Pumping Away" and "Classic Hydronics" authored by Dan Holohan. He explains hot water heating for the expert and layperson alike in no nonsense writing with plenty of examples that he has encountered as a trouble shooter and sales representative for B+G.

As I said earlier the square tubing can follow along the wall on the floor and can be covered by lumber and you do not have to do anything fussy.

All you need at the end of this run or the very end is a boiler drain to power purge the air from one end to the other and do the same for each loop one at a time with a boiler drain on the return leg of each loop, all your air will be gone then. The flow of cold water will push all the air out and then your done. Any air left in the system will be pushed back to your "Garn Boiler" and then to the air space above the boiler.

I can say from this past winter, my first year, that my pump had to push water up my garage wall to my in ceiling HX and back. I purged this line free of air before my first fire and eventually being the highest point air naturally settled back up there. I shut down the boiler a couple times in warmer weather days and the air collected in my HX and line like crazy.

I believe it had been said already to put a T fitting at the 90 back down and put a auto vent or a small valve to release any possible air throughout the season. I would suggest that as well, it's easier to plumb it in now and never use it then to drain the loop and buy more fittings to add it later.