DIY thermal store / heat bank

[esquilax]

Pardon the length of this post but I've read plenty of extremely long and extremely educational threads on here so hopefully this will be another.

I've been reading about thermal stores / heat banks extensively on this site and others as I am thinking of making, rather than buying one. Cost is certainly one potential reason but there is also the fact that although all the literature about thermal stores says "and because they're storing primary water and are not pressurised, they can be made out of cheaper materials like mild steel and need not even be cylindrical". Great, until you actually start looking to purchase such a thing in the correct shape and size and find you might as well be looking to buy one made out of rocking horse poo (it does have excellent strength and insulating properties btw)

What I have in mind is a tank with its four sides made from a single sheet (8ft x 4ft or 2400mmx1200mm) of 3.5mm steel, with a resulting volume of 400L or so. This would leave one welded seam along the long side and two more on the top and bottom (cut from a separate sheet). The many fittings to the tank (a separate thread) would need to be welded on but I would imagine using 1" steel nipples of whatever length needed to bring the fitting out through the insulation. Not sure if immersion bosses are available in steel.

As the sides are flat, I could lag it with expanded polyurethane sheet (kingspan/celotex) which could be retained tightly to the exterior using bolts. Where I am proposing to put it, there would be 200mm free on all sides for insulation.

I am actually leaning towards a heat bank design because I've been impressed by what I've read about PHEs and I can buy one ready made. The idea of a home-made heating coil doesn't appeal to me enormously (although well done to the person who suggested in another thread filling the pipe up with sandblasting sand - I wouldn't have thought of that) but I do like the simplicity of an internal coil for DHW, so I haven't ruled it out completely.

Comments on any or all of the following would be greatly appreciated. Assuming there are no fatal flaws in my idea, I'll put together a diagram for further approval and hopefully actually make something. In which case I will keep you updated and hopefully it'll be a story on par with the guy who made the concrete tank in his garden and put a load of radiators into it.

1. Stratification.

Can anyone with a heat bank say to what extent stratification is preserved after running off hot water? At the moment, with our vented cylinder, we can turn on the immersion heater for half an hour (on "sink" setting) and get enough water for a shower. The top of the cylinder gets extremely hot, the lower part doesn't get hot at all. The hot water is drawn off without any mixing. I would be very unpopular if I installed something that works less well in this regard.

As I understand it the circulating pump driving the primary water through the PHE generally runs at a set speed, and this set speed has to cope with peak demand for DHW meaning that if there is less than peak flow through the secondary side, the primary water returned to the bottom of the heat bank will still be relatively hot. I read an idea of putting a TMV on the primary side but I don't get how this alters the flow of primary water in proportion to secondary. Maybe it doesn't. Can anyone enlighten me?

2. Reliance on thermosiphoning for stove

We have a wood-burning stove, an oil-fired boiler (used when there's no wood) and some radiators. Three downstairs, four upstairs. Two of the downstairs radiators just never really get hot, the stove bangs like hell when any sort of a large fire is lit (and you can feel adjacent pockets of boiling hot/stone cold water in the pipe while this is happening) and, best of all, the copper cylinder springs pinhole leaks on the seam literally every six months. I have now braised it maybe eight times. I didn't install any of this Oh, how I wish it was better BTW.

Anyway, this is why the "neutral point" appeals to me so much - either heat source could be used for DHW or central heating or both, without any banging or one heat source being turned into a radiator by the other. (I find it rather worrying that some people here have gone for DHW only thermal stores, makes me wonder if I'm missing something.)

This project would mean I could get rid of the current DHW cylinder and most of the other plumbing out of the kitchen and locate this thing under the stairs. Not only is this a less useful bit of space, but it is also where the stove pipework emerges. The aforementioned banging is caused by the pipework to the coil having too many twists and tees and elbows and not enough rise (under a metre overall). Based on what I've read about gravity circulation, you would not expect this thing to work at all but it does and is currently the main way of heating water in our house.

The issue with putting the new system under the stairs is that although the pipework would be enormously simplified (down to two completely straight runs of 28mm pipe, each less than 1m long) and we would go from indirect to direct heating, it would still result in LESS rise for the thermosiphoning to work with than at present. Maybe 200mm difference between the stove flow pipe and the top of the cylinder. I believe it would work but I'm not sure it's the best way to go.

I would consider getting a Laddomat / Termovar if it would contribute towards solving this problem. Having read the installation manual for a Laddomat 21/60 I'm none the wiser as to whether it would. I'm sure if I spent €400 on a lump of cast iron with a circulating pump sticking out of it I'd convince myself it was brilliant whether it was or not.

3. Heat leak radiators

We currently have no heat leak radiators per se (some of the upstairs ones do get hot on their own but it's obvious from how they were plumbed in that there was not a deliberate safety feature), although nothing bad has ever happened as a result - yet. I would still like to sort this out. Problem is that I see from other designs the heat leak radiator is often a dedicated radiator just for this purpose. This definitely wouldn't fly in my case so I'd like to use the normal upstairs radiators as the heat leak circuit - but of course I was planning on connecting these through the thermal store, not directly to the loop between stove and store. It would seem like a gross overcomplication to plumb them onto both with N.O. valves on the stove connection. Is there some other way?

4. Gravity-fed DHW and plate heat exchanger

All but the cold tap in the kitchen are currently fed from a tank in the attic, which is captured rainwater. With a heat bank, I would be hoping to switch over to a direct feed from the pump and lose the tank in the attic (not the small metal F+E tank which I'd obviously be keeping). However I don't expect to be able to do this right away as I believe part of the run from the rainwater tanks to the house was done in 15mm (because that's what my father had lying around, grrr) and therefore while pressure might be better than from the attic tank, flow rate is likely to be worse. While this is being worked out, I would therefore like to use gravity feed at first. Could I expect a flow sensor and PHE to behave correctly with gravity-fed water only? (between 0.2 and 0.6 bar)? Have those of you with flow switches found them reliable in general?

5. Galvanic corrosion

The system pipework is mostly copper (no PEX, small bit of gunbarrel that I'll probably get rid of as part of the re-routing of pipes). Should I have an anode of some kind? Should I make the tank out of galvanised steel? It's not much more expensive but I figured my radiators aren't galvanised and they've lasted for years. And the welds wouldn't be galvanised, of course.

6. Solar

I don't have any solar panels at the moment and the job is on a the long finger because it would involve cutting a channel into a wallpapered wall that extends the entire height of the house that would then require redecorating (annoying how one job can trigger loads of dependent jobs but there you go). When I do eventually do it, could I feed my solar heat into the tank with a small PHE tapped off the lower part of the tank, or would it be a big mistake not to install some sort of coil now? There won't be any access hatches on this tank so it'll be an angle grinder job to make any changes in the future.

7. Baffles

I notice some tanks mention having baffles designed to preserve a part of the tank for DHW use, an idea I like because it seems a more reliable way of achieving this objective than thermostats alone. I imagine they consist of a "shelf" around the edges of the tank with an opening in the middle. Anyone know? Has anyone got one in their tank and does it work?

Thank you all for your help!

[KLD]

"I read an idea of putting a TMV on the primary side but I don't get how this alters the flow of primary water in proportion to secondary. Maybe it doesn't. Can anyone enlighten me?"

Maybe like this? When the temperature after the PHE is too high, it is diverted back to the pump. Whether or not a TMV's actuator is fast enough is a different question.

Klaus

[Brandon]

oh, where to begin?

alas it will have o be later, as I am about to leave the house for a night dedicated to the happiness of my wife! (yes, she is coming too.)

[EccentricAnomaly]

An alternative to steel for the tank might be wood + EPDM:

http://www.builditsolar.com/Experimental/PEXColDHW/TankConstruction.htm

[esquilax]

An alternative to steel for the tank might be wood + EPDM:

http://www.builditsolar.com/Experimental/PEXColDHW/TankConstruction.htm

If I was building a really big one I might well do it that way but tank made of wood and plastic would have low Parental Acceptance Factor and PAF is a major consideration here. My father is already sceptical of the idea that it won't (shouldn't) rust...

Oh and I like welding stuff.

[esquilax]

"I read an idea of putting a TMV on the primary side but I don't get how this alters the flow of primary water in proportion to secondary. Maybe it doesn't. Can anyone enlighten me?"

Maybe like this? When the temperature after the PHE is too high, it is diverted back to the pump. Whether or not a TMV's actuator is fast enough is a different question.

Klaus

That is the arrangement I'm talking about, and that is the suspicion I have about it 

If heat banks get popular, someone should develop a circulation pump with a temperature probe built in that can be placed on the opposite side of the PHE to the pump allowing it to alter its speed according to the temperature differential. If the speed of a circulation pump could be varied using a TRIAC I might even be up to making such a thing myself but I don't think it can. (Just looked up how the Grundfos Alpha works and it uses a permanent magnet motor.)

[Ivan]

I think you need to be sure of how you are going to construct the tank before you worry too much about the heat exchanger side of things.

If you want relatively small heatbanks, assuming you have the access, try an IBC (1000litres) which is usually good to 80C, or even a bank of IBCs. Use 10mm copper pipe fed in from the top of the tank for the heat exchanger. We used this successfully for the 12 000 litre tanks on the Navitron mobile solar showers. You can buy off-the-shelf silos up to 45 000 litres in plastic, and some of them are resistant to temperatures up to about 80C.

If you want to construct your own square/oblong tank, don't underestimate the forces of containing tons of floppy water - the sides of the tank will quite likely bulge and possibly break - which is why most water tanks are round rather than square in profile. One method is to dig a pit, use 6" hollow blocks remembering to fill them vertically (and maybe add some steel rod reinforcing at the same time) - use the ground itself as outer reinforcing after the insulation, of course. If your tank is located in a dry environment, you can use straw bales as extremely effective and cheap insulation.

Incidently the heat-tolerant pond liner stuff is pretty readily available and quite cheap (there's a good supplier in the Forest of Dean, but can't remember the name of them).

[esquilax]

I think you need to be sure of how you are going to construct the tank before you worry too much about the heat exchanger side of things.

If you want relatively small heatbanks, assuming you have the access, try an IBC (1000litres) which is usually good to 80C, or even a bank of IBCs. Use 10mm copper pipe fed in from the top of the tank for the heat exchanger. We used this successfully for the 12 000 litre tanks on the Navitron mobile solar showers. You can buy off-the-shelf silos up to 45 000 litres in plastic, and some of them are resistant to temperatures up to about 80C.

Thanks Ivan. I am using ten IBCs for the rainwater and they've worked very well (apart from a few issues with blocked-up bungs). There's no room for one in or around the house though! The two options are thermal store under the stairs, or if that turns out not to be a good idea, another boring old off-the-shelf indirect cylinder in an upstairs cupboard.

[Baz]

Terminology. From a site like HeatWEB I get the impression a tank that is under pressure, even if vented in the attic, is a Thermal Store while a tank that is virtually open at the top is a Heat Bank. I'm not sure if this is a recognised difference.
Since you will need rust inhibitor in your radiator circuit it's not much different putting itin the tank Galvanisation won't last long. A big anode as used for boats is an option. Either way I suggest a filter and even a blowdown connection at the bottom.
As you are making it it won't be difficult to weld in some stays for the flat sides, and if a heat bank style no problem with 1/8 sheet anyway. My 200gall water cistern was paper thin and braced ok by its angle iron corners.

[wookey]

More later (this is one of my favourite subjects :-), but just to clarify the heat bank/thermal store thing. The way this is usually used is that if it has a coil inside for DHW it's a thermal store. If it has a PHE for DHW then it's a heat bank. So far as I know, only DPS really use this term 'heat bank' commerically. Everybody else calls both things a thermal store.

[dhaslam]

If you are making the tank  couldn't you make it to fit the shape of the space under the stairs?   

[MR GUS]

oh, where to begin?

alas it will have o be later, as I am about to leave the house for a night dedicated to the happiness of my wife! (yes, she is coming too.)

   +1

[Brandon]

oh, where to begin?

alas it will have o be later, as I am about to leave the house for a night dedicated to the happiness of my wife! (yes, she is coming too.)

   +1

GUS, how do you know she's pregnant again?

[esquilax]

If you are making the tank  couldn't you make it to fit the shape of the space under the stairs?   

I could, certainly. In fact I originally thought this tank was going to be stairs-shaped until I realised that a six-sided tank that fits comfortably under just the landing of the stairs, not the stairs themselves, would be approximately 400 litres in size and I thought that was enough.

It couldn't occupy the *entire* space under the stairs because I need to maintain access for the pipework. Even so, I'm sure I could find room to double the volume or more if that would be advantageous.

The more certain I am this is going to work, the bigger and more expensive I'd be happy to make it 

[dhaslam]

There isn't anything  in particular that isn't likely to work  but  there are  a  probably two  things that are potential problems.   

The stove will heat from the top down if you have a suitable loading valve but it won't heat the  bottom of the cylinder at all  until the  heat gradually works its way down.   

The other potential problem  is the  opposite  effect.  If the heat exchanger  returns water to the bottom of the tank  then  it will  mix the temperatures in the tank  quite quickly so it probably  will need to return  near the top to avoid this happening.       You may then find there isn't enough heat at the bottom of the store to transfer to radiators unless the oil boiler feeds that area or the store is big enough to carry over heat from a period when the heat demand was lower than the stove output.    This is why the tank size is impo

My store, 860 litres   (1 kWh per degree differential)  transfers heat to a separate DHW cylinder.   The supply to the DHW coil is  from  the top but it returns to the bottom.     When water is drawn off for one shower in the morning  transfer to the DHW cylinder quite quickly takes 7C  from the top of the store temperature and adds  1-2C to the bottom.     The next shower  is just about able to make the store all one temperature.   The DHW  cylinder,  250 litres,  does  however retain  some hot water at the top.