Hot/Cold mix 1 tank

David Woycio

A previous thread has brought up a question that I have for you engineer guys. In a one tank DHW heater tank how long does it take for the layer of hot water heated by a coil or element take to cool from the lower layer of cold water in the tank? Assume that have a 120 gallon tank and have heated the top half to 120*F and then pull the plug. There is no draw on the tank so no cold water entering bottom. The starting temp is 60*F so the bottom 60 gallons is cold. There will be standby heat losses but I would rather focus on what happens to the (2) bodies of water.

Is there a finite line between the two or do we have a gradual increase/decrease in temp? Or if left alone is there a lineal increase from bottom to top?

Thanks in advance!

Brad White

Thermocline

Assuming you could even get the state you describe (sort of like those stunt bartender drinks with layers at different densities), this is what I see happening:

The boundary of a common liquid at two temperatures hence densities is called a thermocline.

Swimming in a lake in summer illustrates this phenominon where you dive down a few feet from what feels at first like bathwater then your schmecke does the shy turtle thing.

The higher the differential in temperature or density, the greater the initial resistance to change. Unlike delta-T (conduction) where you would think the propensity would be greater, here is where higher temperature differences trump that. It creates greater buoyancy and holds the hot water high and cold water low. (Low-Loss Headers work on this principle even in a dynamic state of forced flow so you know it works. Also thermosiphon traps to limit ghost flow works on the same principle.)


At that edge, the driving agent of change is conduction, a gradual square-dance of heat between the layers. With cooling at the edges of the tank being greater, one can imagine the curls of dropping cooled water going down the sides, displacing the colder water to a point that the cold water wells up in the center. A reverse convection if you will. Think Lava Lamp here. (Scott Milne still has his, I bet, to go with his Rectorseal.

Add to this the Coriolis effect, the rotation of the earth, and you really have something...especially in the far north and far south but in different directions!


At first the process would be slow, then accellerate by the momentum of falling water at the edges. As things come to equilibrium, the process would again slow because the motive force is less than you started with.


Then God invented the Dip Tube.

David Woycio

lunar effect

Brad - you forgot to mention the gravitational pull during a full moon! Thanks I get what you are saying but I'm looking for a factor of time. Several reasons for this one being the use of a single tank with (2) coils. We have always opted for separation with tanks because of the reaction you mentioned. If a single 120 gal storage tank with (2) coils from boiler and solar are used to what degree are we heating the lower portion of that single tank with our boiler zone? If the solar is hot .... not a big deal. But if we haven't seen the sun in a few days, solar depleted, then comes Mr. sun then we have a situation where we are not heating 60*F water but something higher. The higher temp is what I am interested in. Systems are at they're most efficient when heating the coldest temps. Me thinks
I will need to set up a tank in our shop and observe.
Maybe I can observe the worlds largest lava lamp!