Stratified storage tank

[Mark Holden]

I'm really enjoying reading the accumulated experience on this forum; thanks to all for sharing.
This is something I wrote earlier, mainly for solar heat systems.
However, it is also relevant for biomass and other heat sources; mainly, it would be good to use the statification tube for the return end of your heating loops, to keep the tank as stratified as possible. The point of that is so you're more likely to have very hot water at the top of your storage tank for your DHW. Also, as the tank cools, you can still extract some useful energy from the top even when most of the tank has become too cool.


The stratification tube is what introduces heat into the tank. Here's the clever part; the incoming hot water should be directed into the tank at the level where the temperature is the same as the incoming water. Cold water is removed from the bottom to be heated, and the warm layer slowly moves down. Hot water will be at the top of the tank, and be used mostly for Domestic Hot Water [dhw]. The middle of the tank, at 40°-60° will be used for home heating.
When you get strong sun, 80°-90° water will be produced, and sent to the top of the tank.

Let's say your solar collector output is at 60C. It will rise past the lower flaps, since the higher density cooler water on the other side holds them closed. It will not rise into the hotter water above, since the entering 60° water is denser and heavier than it. It will push open the flap at the point in the tank where there is the least resistance, which is at the point of equal density and temperature.
The water in the tube above the open valve will heat up through the tube walls.
The flaps should open with not more than a couple of grams of pressure, and fall closed by gravity. They don't need to seal perfectly, as at the ultra low pressure difference against them, little water will leak through. The important thing is that they don't stick, and move freely, even after years in the water. It will be hard to tell if something isn't working right, and even harder to fix.
The tank should have an access door for maintenance.
The artwork is mine, but the principle of operation is well known and proven.


This system works best with a glycol mixture, as glycol changes density with temperature more than plain water does; but for a big tank like this, glycol is too expensive. Therefore, make the tank as tall as possible, preferably 3 meters [10 feet]. The extra height equals extra stratification.

The stratification tube;

After a lot of searching, and even consulting with university professors of solar energy, I found this design. It seems to be the only one known to work consistently, but it's not possible to buy it as an individual part. It's patented by Solvis and they only sell it as part of their storage tanks. Therefore, this information is for educational purposes only!

The concept; very sensitive and large format check valves will simply let water into the heat storage tank at more or less the level where water of the same temperature and density is.

Colder, denser water in the bottom of the storage tank will hold the lower valve closed. warmer, lighter water will float above.

To test the principle of this, I built a simple test rig from a steel oil barrel and some plumbing parts. I let water out of the tank from the bottom while introducing new water in through my trial stratification tube.
This is to simulate what would be happening if it were connected to a solar collector.
This is my experimental version, not meant for actual use; the construction of this one isn't robust enough to work for 10 or 20 years in hot water. The pipe is 90mm [4"] PVC drain tube and the flaps are made from "trespa", a material similar to Formica but 6mm [1/4"] thick. It's waterproof and heavier than water.

The first try, in the first photo, didn't work well. Although hot water did exit the top while sucking the bottom closed, the valve only opened an inch or so. the water squeezing through the gap entered the tank at too high a velocity. This is known to cause turbulence, mixing, and loss of stratification.
In the second photo I've made a slight change; the pivot points are drilled into the center of the valve body, instead of using rings behind it. This made for a better balance.

The flap was weighted to have only 2 grams more weight below the pivot than above [when under water]. This worked perfectly.

In this photo, I'm letting 55C water into the 90mm stratifier tube at about 40 liters per minute. the top flap is fully open, water velocity is low. The lower flap, which was hanging slightly open before the flow was started, is now pulled closed, presumably by the venturi effect.

In this photo, I'm letting 18C water into the tank. the layer of hot water floats above as the cold water enters at the bottom, right where we want it.

I used an infrared thermometer to monitor the water temperature through the thin steel barrel. The temperature difference was clearly measurable, although there was a "fade" zone instead of the desired sharp thermocline. I suspect this was a combination of imprecise measuring technique [the steel barrel would conduct some heat from the hot to the cold area] and mixing due to the small size of my test barrel in relation to the large stratifier tube size.

The density of plain water changes only slightly as the temperature rises, so to use the effect the valves have to be very sensitive. A glycol solution would have about twice the effect, but I intend to build a tank in the region of 5 tons, and I don't want to buy 2,500 liters of glycol. And if I did, I would need another 5 ton tank to pump it into in case I needed to modify or work on the heat tank.
With water I can just dump it and pump in new.
It's important that the fluid velocity be low enough to prevent turbulence and mixing, which will cause a loss of stratification.
According to published research, outlet velocity should not be more than 0.15 meters per second. In other words, a very gentle but steady flow.
For 40mm tubing [38mm ID]; 0.192 lps [liters per second] or 11.52 lpm or 691 lph

60mm tubing [56mm ID]; 0.414 lps or 24.8 lpm or 1490 lph

80mm tubing [74mm ID]; 0.547 lps or 32.82 lpm or 1969 lph

The choice of tube diameter will depend on your fluid flow, which is determined by the size of the circulation pump and your solar collector array.
But as a rule of thumb, bigger is better.

 

[hobbyheater]

I


.

Looks interesting!

 

[Bob Rohr]

Looks interesting!

I saw a tank like that and it had some sort of fabric stratification tube, looked almost like a nylon stocking. Another manufactured enhanced the stratification with a device on the outside of the tank.

 

[Mark Holden]

I talked to a professor of "solar engineering" at a US university by email; I asked about those fabric socks, where I might get such a fabric, and what would last 20 odd years in such an environment without breaking up or clogging up.

She replied that they only test those things to prove theory and computer modeling [for the PhD students]. As soon as the testing is done, they take the rigs apart again. She couldn't tell me if the fabrics would last, but she doubted they would. so do I.

I was working on an electrically controlled system with lots of thermo-switches and electric valves, when I came across this system. It seems too simple to work, but they sell the tanks [with some gizmos on it] for 6,000 Euros [about $8,000 US]. They're the top of the line thermal solar supplier in Europe. anyway, I looked up their patents and checked it with my own experiment.

I think it will be in the public domain in about 10 years... Then they'll be available off the shelf.

 

[Mark Holden]

This circuit is to control which section of the stratified tank hot water is drawn from.
In this setup, each switch is set to the same temperature. Heat is drawn from the lowest point at which the temperature is high enough. So if your heating loop needs water at 60C, the hotter water above is held in reserve.
If a given section is too cool, the power is shunted to the switch above.
In this diagram, even the top valve closes if it's too cool; depending on your design, you might want to leave out the top switch and have the top valve open whenever the section below the top is too cool, even if the top of tank is also below your target temperature.
If you have big radiators or underfloor heating loops, probably you'll be able to run it at 50 or maybe even 40C. The lower you're able to set this temperature, the higher your efficiency will be.
There will need to be a second identical circuit for the domestic hot water [obviously that needs to run through a heat exchanger]. DHW will have to be 60 or 70C.
24VAC electric valves for irrigation systems sure are cheap; but I don't know if they'll hold up to the heat. Depending on the height of your tank, you could need 6 or more for heating and 6 for DHW.
The proper ones for heating systems cost 4 times more.
Return water from heating or DHW must enter the tank through a stratifier tube.

Design, text, and art by Mark Holden

 

[DaveBP]

I wonder, if the multi-outlet pipe is big enough to keep the flow rate (and therefore turbulence) to a minimum, do you need the flapper/valve gizmos?

Will the water 'find' its own level with an acceptable amount of mixing to make no difference.

I'm thinking about simpler is better if you don't lose enough to make it worth the risk of the "valves" sticking shut.

 

[woodsmaster]

Interesting. Looks like some fun experimenting !

 

[Mark Holden]

In my experimental rig, when I ran hot water into the tube, the lower flap visably pulled shut; if there was no flap, it would suck cold water from the bottom of the tank into the tube, mixing it with the entering hot water.

 

[Mark Holden]

I was just looking for the patent; apparently they have a Europe patent granted in 1992; 20 years ago, I assume it's expired. here's a link to their tech brochure in English, it's good reading; (broken link removed to http://www.solvis-partner.de/userfiles/2430_P12-EN_Grossanlagen_Technische_Information.pdf)
I find their prices astronomical, and they don't do big tanks like we all seem to like. But they did figure out a few things.

 

[ewdudley]

Another technique I've seen would be to use one or more cascading thermostatic diverting valves external to the tank. Might be more practicable than incorporating a custom built thermo-separator into an existing pressure vessel.

 

[Bob Rohr]

I was just looking for the patent; apparently they have a Europe patent granted in 1992; 20 years ago, I assume it's expired. here's a link to their tech brochure in English, it's good reading; (broken link removed to http://www.solvis-partner.de/userfiles/2430_P12-EN_Grossanlagen_Technische_Information.pdf)
I find their prices astronomical, and they don't do big tanks like we all seem to like. But they did figure out a few things.

There are some wild tank ideas every year at the Intersolar show in Munich. Here are a few. Also SWE Sun Wind Energy magazine has good articles on tanks from time to time. Jenni in Switzerland make some huge stratification tanks, photo shown by permission SWE

 

[Hansson]

http://translate.google.se/translat...ro.se/produkter/dalatanken-med-skiktroer.aspx

 

[Mark Holden]

Now that's a BIG tank!
A friend of mine was employed on a project with a tank like that; I never saw it but he described the setup to me.
it was a government project to solar heat an apartment block in the Netherlands. a cylindrical buffer tank stood nearly the height of the building. There's a football field of collectors on the roof.
They spent literally millions, I don't remember how many.
And it doesn't work! He couldn't tell me exactly why, but I gather that there was too much demand, not enough sun [they don't get sun in winter], and the tank wouldn't stratify.

 

[Bob Rohr]

Now that's a BIG tank!
A friend of mine was employed on a project with a tank like that; I never saw it but he described the setup to me.
it was a government project to solar heat an apartment block in the Netherlands. a cylindrical buffer tank stood nearly the height of the building. There's a football field of collectors on the roof.
They spent literally millions, I don't remember how many.
And it doesn't work! He couldn't tell me exactly why, but I gather that there was too much demand, not enough sun [they don't get sun in winter], and the tank wouldn't stratify.

A successfull 8 unit apt in Switzerland with 100% heat hot water, and excess piped to neighbors. A short You Tube at www.jenni.ch I think there is a data logger in realtime on the site also.

 

[Mark Holden]

The first diagram I started the thread with is based on a patented, commercially sold, proven design.
It seems the patent has lapsed by the way.
In the commercially sold version, the tank is about 200 gallons and filled with a glycol mix. That isn't practical for big tanks, but my solar components are supposed to be run with glycol, so the system would need a heat exchanger for the solar heat loop.
In this diagram, the heat exchanger is inside the stratifier tube, and relies on convection.
My solar collectors will be about 18 m2 [200 square feet], and in theory should make about 18kw [61,000 btu/h] at noon on a sunny January day in southern Portugal. Of course at other times of day, or when overcast, much less or zero. In summer, overheating will be an issue.

I'm interested in opinions on whether convection will do the trick in the heat exchanger. A problem is the lack of control; if there's too much flow, I'll get a large amount of water inadequately heated.
In winter, the solar collectors should attain 80-90C [170-190F] quite easily [I did some testing last winter to check this]. I really need hot water to build up at the top of the tank to have any hope of getting solar heat and dhw.
An external pumped heat exchanger would allow volume / temperature control of the secondary water; but would be less efficient and more complicated.