Willis solarsyphon

[Billy]

I know nothing, but what if....

the larger column of colder, denser water (relativity) in the tank is pushing down (gravity) into the HE or W syphon with a much smaller pipe size.

Could this be anything to do with hydraulic jack principal and mechanical advantage and not exclusively the hot water rising  You see what I'm getting at or am I perhaps firmly in lah lah land.   

What think the brains out there.

Be brave, shoot me down, my shoulders are broad. 

Just a thought.

Billy the Barge

 

[KLD]

Billy,
the diameter of the vessels should have nothing to do with the relative buoyancies.
Try this as a more intuitive experiment: take one small diameter tube and one barrel, connect the two with a length of tube on their bottom ends, and fill with water. The water level will be the same in both pipes. (Well, except for the case when the small diameter is so small that capillary forces distort the picture. But enough of that:-)
The only parameter determining the outcome of this experiment is the density of the water.
If you were to lower the density of the water in one compartment, say by heating it (were did that sun go?), then its level would rise.

Now, if you install an overflow from one vessel to the next, the rising level would yield a flow into the colder vessel, thereby heating it. Eventually, both vessels are back at the equal, but higher temps (and slighty expanded volume; where did the F&E tank connection go?  ;-)
In other words, the average density in each vessel determines, whether at all and which way round the water flows. (As Wookey has pointed out earlier.)

Klaus

[Billy]

Klaus,

I've just crashed and burned, when you put it like that,  but thanks.  I think I am getting confused with atmospheric pressure and sealed tubes.  Not that I know anything about that either!  I just happen to have a bucket with a valve and hose tail in the base, will experiment to understand more fully.

The Willis caught my I as I have just ordered some of Ivan's tubes.  I have a twin coil tank but have them plumbed to other stuff.  The syphon seemed to be the answer to my little problem but I have not got 500mm to spare.   I too am confused as to why the Willis has to be tee'd in at 500mm and how heated warm water finds its way over the hot tank water, but not half as clever as the blokes who used to use a single horizontal pipe for flow and return for gravity heating in the old New York blocks.  I reckon that is extreme stratification and why doesn't the boundary mix, I guess it does to some extent.

Wish I had your lots brains,

Billy the Barge

 

[lightfoot]

Hi Klaus,

Sorry, I may be misreading your post some how (it's getting late ) - But......

Billy,
the diameter of the vessels should have nothing to do with the relative buoyancies.
Try this as a more intuitive experiment: take one small diameter tube and one barrel, connect the two with a length of tube on their bottom ends, and fill with water. The water level will be the same in both pipes. (Well, except for the case when the small diameter is so small that capillary forces distort the picture. But enough of that:-)

Correct and also, the static pressure (which in a open system is relative to the hight (head) of the water level above a given datum) will be the same in each too.

If you were to lower the density of the water in one compartment, say by heating it (were did that sun go?), then its level would rise.

Now, if you install an overflow from one vessel to the next, the rising level would yield a flow into the colder vessel, thereby heating it. Eventually, both vessels are back at the equal, but higher temps (and slighty expanded volume; where did the F&E tank connection go?  ;-)
In other words, the average density in each vessel determines, whether at all and which way round the water flows. (As Wookey has pointed out earlier.)

Klaus

If one of the compartments / vessels was heated, natural convection will cause the hotter, more buoyant water to rise / float on top of the cooler water within the compartment / vessel.  However, discounting any displacement due to the thermal expansion of the water, the actual water level will not rise (if that's what you meant ?). 

If the two compartments / vessels are connected at the bottom (or anywhere below the water level) then in a open system, under steady state conditions and the same static pressure, the water levels will always be the same in both compartments / vessels and/or connecting pipework - regardless of any temperature differences.  As water in a system will always try to maintain the same level. 

(Note: for simplicity, the above discounts any dynamic pump pressure / velocity effects that may occur within a system)

If the two compartments / vessels are also connected at the top as well as the bottom and one of them is heated, then natural convection will again cause the hotter, more buoyant water to rise and induce gravity circulation (thermosyphon) between the two compartments / vessels.  The water velocity / turbulence needs to be kept to a minimum, to avoid mixing / dilution - to promote good stratification.



Billy, As far as hydraulics/jacks go, I think you may have been getting confused with the potential pressure rise caused by displacement of a fluid etc, ie if you try and force 2 ltr of water into a 1 ltr container - then something has got to give (sorry it's maybe not the best analogy - but hopefully you get me drift  )



Lightfoot.

[KLD]

Lightfoot.

my thinking was as follows:

firstly, I agree that in corresponding tubes the liquid will rise to equal levels. But only as long as the liquid has a homogeneous density, i.e. the density is constant throughout the system. Lets consider the simplest system, in which you have just one tube bent into a u shape. At any given height, the pressure in both arms is identical. This pressure is a result of the mass of liquid above the chosen level, and the section area of the tube. (And the atmospheric pressure, but let's assume that is constant and equal for both arms.)

Now, if you were to replace the liquid in one arm of the U with a different liquid of, say, lower density, then you'd need a larger volume of this new liquid to yield the same mass, and therefore the same pressure at our chosen level.

The lower density could also be achieved by heating the one side of the U. The "thermal expansion" is exactly the increase in volume I'm after As long as it's just a U, i.e the tubes are only connected at their bottom ends, no displacements occur.

Whether the heating of the one arm is even throughout its whole length, or whether there is any part warmer, it's the average density that determines how high the level will rise.

 
Klaus

[ericw]

I think some clue to the answer to the question of whether thermocycling can take place if the temperature in the side arm is lower than the temperature at the top tank connection can be obtained by looking at vastly simplified model.

Consider that the cylinder and HE heated side arm to be a simple 0 shaped loop.

One side representing the side arm is full of warm water.
The other representing the tank has a layer of (perfectly stratified) hot water on top of cold water.

To work out when thermosyphoning can take place you must look at the relative magnitude of the forces at the bottom of the 0
The force from the side arm will be its height multiplied by the density of the warm water
The force from the cylinder side will be the height of cold water multiplied by its density plus the same for the hot water layer.

Interestingly this shows that it is driven not only by the relative densities (temperatures) as expected, but also by the position of the hot-cold interface in the cylinder. When there is only a thin layer of low density hot water then its perfectly possible for the system to thermocycle even though the side arm temperature is lower than the hot water temperature, as the average density in the cylinder can be higher than that in the side arm.

So initially it will thermosyphon at relatively low temperatures but the thermosyphon temperature will increase as the hot-cold interface is pushed further down the cylinder.
Of course in real life there are no nice sharp boundaries and it will work on the average density in the two sides but I believe the principle still holds.

This could also explain why Willis connect their device higher up on the expansion/vent pipe.
Because its in a relatively thin pipe, it only requires a small volume of water to fill it and it will effectively create a much deeper hot layer than would happen in the cylinder alone.
If the theory above is right, this will suppress lower temperature thermosyphoning and so give hotter water at the top of the tank.

[lightfoot]

Hey Klaus,

Having thought about it, you are indeed correct with regards to the water density/level difference when two tubes/vessels are only connected at the bottom.  Have a cigar......I'll go stand in the corner.....where's the Donkey Ears smiley 



Cheers


Lightfoot.

[wookey]

I sympathise, lightfoot. I started a post arguing with him then realised half-way through that he was right actually. It's a subtle one because you hardly ever see it in practice.

[djh]

this shows that it is driven ... by the position of the hot-cold interface in the cylinder.

This could also explain why Willis connect their device higher up on the expansion/vent pipe.

Brilliant!         

Now I think I understand it.

Have an applaud.

[Ivan]

That explains why Wookey's system works well, even though his temperature sensor is connected to the cylinder rather than the side arm heatexchanger, doesn't it?

Does that also mean that you shouldn't connect the cylinder reference temperature sensor to the cylinder, when you are plumbing high into the vent/expansion pipe as per Willis's plumbing recommendations.....which means their temperature sensor recommendations are wrong?

[Brandon]

If the theory above is right, this will suppress lower temperature thermosyphoning and so give hotter water at the top of the tank.

That all makes sense, but on a cloudy day when the top of the cylinder is hot >50c from either the previous days solar or an evening WBS burn for example, then the panels may not reach that temperature, and therefore no energy is had from the panel during the day.

Surely then a slightly raised full body of water is better as it captures energy?

[Ivan]

Not really an issue with vacuum tubes - the energy they can gain is limited by the amount of sunshine, but the delivery temperature is almost independant. So you can still get a small amount of hot water on overcast days, rather than a greater amount of lukewarm water. Obviously, this doesn't apply to flat plate systems.

[chrispierce]

We are fitting our first Willis Solar syphon and require clarification. Having read through as many posts as poss on the forum, using TDC3 controller, I am confused where the best place to fit the S2 sensor should go, ie in the insulated jacket of the solarsyphon or in the lower part of the jacket. Secondly does  the 6/4 temp difference still apply. Are there any other settings which would be perculier to using a Willis SS.

[Northbridge]

I worry about reports of "the customer says they have more hot water".  Compared to what??? the solar hot water system they took out last week?

Cynical I know.... I have had a meeting with our other technical staff and a CHAP FROM OUR CYLINDER MANUFACTURER and our conclusion - do the job right and change the bloody cylinder.  second rate is second rate.  And how much solar do we do?  loads.

It's obvious to a nugget what the shortcoming of these units are, and where they fall over.  If someone can't pay the extra 300 quid over the willis unit price) for a twin coil newark cylinder then send them elsewhere for a second rate job - don't hang YOUR reputation on it.

And happy new year!

[martin]

now the bloke from the cylinder manufacturer would, of course, be totally unbiased...........there are many companies out there who do LOADS of solar installations, every single one of them cr*p........Me? - I'd sooner wait for the independent test results from the setup at Navitron HQ before leaping to possibly uninformed and plain wrong conclusions