thermosyphon

Mike T., Swampeast MO

Absolutely nothing wrong with the home run piping with constant circulation.

It's only the bottom supply/top return on a panel radiator that I suggested might be less-than-ideal.

TRVs are in use on the rads? Did the reversed piping on that one panel also result in reversed flow through the TRV?

TRVs are unidirectional with regards to flow. While water will flow through them in the opposite direction when they are open, they will not work properly. Usually you hear loud, repeated banging as the TRV begins to throttle down. Even if you didn't hear the banging, I suspect that reversed flow would significantly alter the delta-p of a TRV body.

I'll try to find the diagram showing the effect on output of different connection methods to the same panel radiator.

Once a circulator is running in a system it's no longer a system of natural convection--it's forced convection. The forces generated by the circulator--even a small one--are overwhelming compared to those of natural convection.

When the water goes into a huge wide spot in the road (like a radiator) velocity will slow sufficiently that natural convection will affect how the water is distributed <I>inside the radiator itself</I>. It will not however in any way affect the quantity of water moving through the radiator as such is under the control of <I>forced</I> convection.

In a forced system with fairly low velocity (as you will have here with a directly connected Vitodens), if you put the supply at the bottom of a panel radiator with the return at the top, opposite side, natural convection inside the radiator will tend to allow the supply to rise to the top of the radiator--primarily in the sections nearest the supply connection. Since water is "lazy" and will find the path of least resistance to satisfy the forced convection the hottest water will then tend to travel across the top of the radiator to the return. The wider the panel, the greater the effect and the panel will not heat very evenly. In this case, the natural convection inside the panel is your enemy.

If you put the supply on top and the return on the bottom opposite however the hot supply water is <I>already</I> where it wants to be via natural convection. The hot water tends to spread out across the top of the radiator and be drawn down through the entire radiator towards the return and the panel heats very evenly. The forced convection continues to overwhelm natural convection inside the panel. That's why I called it "counter-flow"--because such an arrangement is counter to natural convection inside the panel.

scott markle_2

Thermosyphon effect in panel radiators

I have been following the smart pump thread recently and got to thinking about the issues of conversion of electrical energy to moving water etc. Intuitively 5 watts seems like hardly enough energy to move a meaningfull amount of water even if this conversion was 100% efficient.

My thoughts then wandered to a panel radiator job that did recently. This was a multiple radiator manifold piped system direct piped to a vitodens in an unheated basement. My calculations indicated that the built in circulator would provide adequate residual head at the flow rates that I desired.

Inadvertently I reverse piped one of the panel rads. Interestingly this switching of supply and return almost completely prevented flow to this particular radiator. Once this piping error was corrected flow to this radiator (as indicated by it's output) was completely consistent with the rest of the radiators on the system.

What this indicated to me was the extent to which the thermosyphon effect was having on this type of radiator. It's my guess that a system piped in this way to radiators of this type may be experiencing significant assistance to it's pumping requirements by this effect.

Once a continuous circulation system like this gets up and running and assuming a reasonable delta across the surface of the radiator it may be that these systems require much less "pump power" than a conventional flow rate head loss calculation would indicate.

Any thoughts on this?

Steamhead (in transit)

How exactly

was it mispiped? Supply on bottom/return on top?

To Learn More About This Professional, Click Here to Visit Their Ad in "Find A Professional"

scott markle_2

Yes

The piping on these units is left-right= supply-return, I'm fairly certain supply is on top return at the bottom. Colder water having higher mass and falling.

While it's not hard to understand how this works it was surprising to what extent this physical property could override the forced circulation.

This observation leads me to believe that thermosyphoning may be assisting pump induced circulation more than one may think in certain circumstances .

zeke

The reverse piping problem you mention should not have been caused by the thermal siphoning but perhaps a check valve oriented in the "wrong" direction. The thermal "pumping" should have easily been overcome by the circ pump pumping in the "wrong" direction.

By the way, unless you are talking branch piping of the order of 1-1/2 to 2 inches in diameter as the "old" natural convection systems were designed, this socalled thermal siphoning is minimal. That is precisely why modern piping has pumps.

scott markle_2

thermosyphon

Zeke, Thanks for the reply I was afraid this thread would die without any relevant input. Troubleshooting this problem was especially frustrating because I made the same assumptions as you. There are no check valves on the individual circuits or in the radiator valves (as far as I know).As I stated this was a home run piped arrangement off a radiant balancing manifold.

The output from the misspiped rad. was dramatically effected by this. Correcting the misspiping fixed the problem completely.

One Consideration is that rads. required fairly low flow rates and thus a low head as well.

Because the outputs from the various rads. was sufficient I made no adjustments to the balancing valves,all wide open.

I assume that the thermosyphon effect was enough to cause system flow to largely bypass the miss piped radiator in favor of a path of lower resistance. It must be that the design of these steel panel rads. creates a fairly strong thermosyphon.

Mike T., Swampeast MO

I presume the "incorrect" piping found the supply on the bottom and the return on the top at the opposite side. Correct?

What make a model of panel radiator?

I quite highly doubt that your experience is related to thermosyphoning (e.g. gravity flow). Gravity flow works throughout the entire system from the lowest to the highest elements. With a directly connected Vitodens I can nearly guarantee that the head loss in the heat exchanger (not to mention its configuration and the fact that the HX is the HIGH point of the boiler) will prevent any amount of gravity flow. There is truly no way for gravity to make water flow UP into the "Slinky" and then back down to supply the system.

Much more likely that you found the consequences of less-than-ideal piping of a panel radiator.

I'll try to dig, but I remember finding an output chart of some brand of panel radiator with various supply/return connection combinations. Am 99.9% positive that the best method found the supply on the top and the return at the bottom of the opposite end--at least for the particular panel radiator observed.

Supply at the top and return at the bottom on the opposite side will set up a natural counter-flow inside the panel. The hottest water naturally wants to stay at the top but the since high pressure (supply) ALWAYS flows to low pressure (return) typical internal design of a panel will force the hot supply water to distribute itself fairly evenly throughout the panel. Other piping configurations will involve a greater degree of "short circuiting" where the more and more of the hot supply water is drawn directly into the return.

It may be possible that panel rad manufacturers intentionally use this counter-flow to ensure maximum distribution of the supply throughout the panel with a controlled delta-p. Pipe incorrectly and both distribution of the supply and delta-p could be profoundly affected.

Mike T., Swampeast MO

Direct-Connected Vitodens

You have my best respect Scott for going through the numbers, designing carefully and using the Vitodens in the way it's really made to work...

scott markle_2

counter flow

mike, I'm not suggesting that this effect is creating a flow through the boiler HX. Obviously the circulator is required to overcome the head loss here and maintain required flow. As you point out the physical layout of the HX would have no reason to induce a thermal convection.

I'm not sure why you would feel that home run piping would be a "less-than-ideal" arrangement for a trv'd radiator constant circ. system.

The bottom line is that when I corrected this error my radiator performed flawlessly, Totally consistent with the rest of the rads.When pipe backwards this radiators output was pathetic,clearly something was wrong, Initially I believed I had a blockage.

Again it should be understood that I was feeding a 8 outlet manifold from a direct piped vitodens, 3 new burnham cast iron rads. a small loop of ultra fin, 3 trv'd radison panel rads.

I think you may have the counter flow logic backwards.I believe It's not that hot water rises but that cold water sinks. This is how old engines cooled before water pumps. The radiator induced circulation. Hot water enters the top of the radiator, it begins to cool, thus becoming heavier, it begins to sink, creating a flow, this force is strong enough to "pump" water through the waterways of the engine.

If hot water is piped to the top of a panel rad. natural convection will assist it's movement to the outlet at the bottom. I don't see a counterflow mixing effect here.

Bottom piped cast iron is a different story, here the thermal convection is separate from the circulator induced flow through the bottom of the radiator.

scott markle_2

Thermosyphon effect in panel radiators

Mike, this makes more sense to me, but still my intuition say natural convection will create top to bottom flow, heavier cooling water falls right?

This misspipe did not merely reduce output the effect was dramatic loss of output. Perhaps as you have described, flow is essentially bypassing majority of rad. surface. And this is not a flow restriction from opposing thermosyphon as I originally hypothesized.

Mike T., Swampeast MO

Scott: Did the mis-pipe result in water flowing the wrong way through the TRV? If so, I suspect that is a major culprit here.

With the supply on top and return at bottom opposite, natural convection might aid in the even distribution of heat throughout the panel, but it will not in any way affect the quantity of water flowing through the panel. I really suspect however that it's the forced convection (and internal design of the panel) that really matters.

Standing iron radiators certainly work with bottom-bottom connections whether under true gravity or forced circulation. Why the difference? The rads themselves are different. Their passageways are numerous and HUGE. Even the dead men considered them having zero head loss in their systems. While it may not be much, panel rads do have head loss.

Standing iron radiators naturally establish their own natural convection systems--even when you're using forced circulation. The only real way to screw this up (assuming bottom-bottom connections) is to move an insane amount of water through the rad such that velocity from supply to return is so high than the natural convection inside the rad never has a chance to get going.

If you use the Ferris wheel analogy for gravity hot water circulation, bottom-bottom connected rads become their own little natural convective system--almost like a double Ferris wheel. The rads are the sub-wheels with natural convection keeping water flowing up and down through them quite evenly while the overall wheel is providing the power to heat the rad. The bottom-bottom connections actually aid in this due to the lack of any head loss in the rad itself.

Neglecting some systems using special fittings that allowed a single supply riser to serve rads on multiple floors, about the only time you'll see the top connection of a standing iron rad used with gravity is for overhead systems. In these systems the hot water from the boiler zoomed to the attic via huge "express" risers and then fell down the system, through the rads and into the boiler as it cooled. In this case, you'll find the supply on top and the return on the bottom. Special fittings allowed a single drop supply to serve rads on multiple floors and special devices even allowed both connections to be on the same side of the rad. The rads heated evenly because natural convection allowed the water to fall throughout the entire rad but in this case the rad itself isn't acting like the sub-wheels of a double Ferris wheel. Water is only falling through the rad--it's not rising and falling inside of it.