Home made Heat Bank

MOV = Metal Oxide Varistor - basically a spike suppressor. It has a high resistance at 240V, but the resistance then falls as the applied voltage rises. Its effect is to short out spikes of high voltage that you may typically get when switching inductive loads (which can generate big back voltages (or reverse EMF if you prefer).

It is: TMV = Thermostatic Mixing Valve. See

Mechanical - wax capsule etc

Yup, see

Is your boiler a condensing one?

With a condensing boiler, lower return temps give more efficiency. On a conventional boiler that is not true. (in fact lower return tempts might help promote condensing which is not a good thing on a conventional boiler where higher temps will prolong the life of the HE.

I'm in a similar position to the OP, trying to get my head around the details of a heatbank design.

Mine will differ slightly in that it will be built from scratch and I really need a mains pressure CH circuit (difficult pipe routings), whilst sticking with the unpressurised boiler/tank circuit.

To do this, is it simply a case of duplicating the DHW/PHE circuit? (With an expansion vessel on the CH side) How should the pipework be split off for this duplicate circuit?

What would the control system look like? Would the boiler simply be controlled by a tank thermostat, so it's function in life is simply to maintain tank temp? Would the CH circuit have TRV's on rads every location except one main rad, and then a room thermostat to cut the pump(s) on that circuit?

Finally, this diagram

Cold water (potable and under pressure) is presented at / into one side of a heat exchanger - When a 'hot' tap is opened water flows through the pipe and is observed by a flow-detector; this detection enables a pump which pumps water into the 'other' side of the heat exchanger (think, a twin-tube car radiator). (Now) Hot water at the tank's temperature is blended with cold water via a wax-cartridge thermostatic mixer valve to produce 'Hot' water at your pre-set temperature. Effectively. there's no 'Hot' water until you open a tap. The 'hot' water is off the 'cold' mains- only the energy required to bring incoming mains at (say) ten degrees is extracted from the tank to heat it up to (say) fifty-five degrees. The 'rest' of the energy remains available for use in the tank. [There's owt for nowt and you'll need to have your boiler replace the energy sometime.]

Thanks for all the specific info John.

Boiler is a Worcester Greenstar 24Ri, about a year old - so yes it's a good condensing boiler

Mike

The easiest way to explain this is that if you had an expansion vessel (assuming you're just meanign a normal vented F&E tank) your house would flood horribly :-( The DHW is directly connected to the mains via the PHE. Open a hot tap or any other point in the circuit and you'll get mains pressure and flow.

HTH

Mike

Now I think on a bit more, my mind ventures onto controls.

I seem to recall reading something about having 2 tank-stats on the tank in this kind of setup... is that correct. I'm thinking that would be ideal - one stat set at 80degC, one stat set at something much lower eg 45degC, to provide max and min temps for the store. The theory being that the boiler would fire up when the store got below

I'm thinking a simpler alternative might be to time how long it takes for the boiler to heat the tank to 80deg on an average day (ie average hot water useage and therefore initial tank temperature). Once I have this info I could just tune the timer clock for the DHW programmer (which I already have)

Mike

I can only attempt to answer you by comparison with 'my' Pandora unit.

The heatbank accepts heated 'fluid' from the boiler which flows in at the top. The working fluid is pumped back down to the boiler controlled by a wax-cylinder mixing valve. This working fluid temperature is selected by the 'thermostat' mixing valve; the duration of the circulating pump is set by the 'timer'. The thermo mixing valve accepts a 'Hot' input from the top of the heatbank and a 'Colder' flow from the bottom of the heatbank. In other words, the fluid is circulated all the time the 'timer' is 'ON' and is maintained at the setting on the wax-cylinder mixing valve. This fluid is available for circulating through the CH system -where it will relinquish heat - when the fluid flows back into the heatbank then back to the boiler to regain heat energy. The boiler should 'switch off' when the 'demand heat' line is dis-abled .

The controls seem premised on marinating the HeatBanks working fluid at a constant temperature, and the boiler supplying heat into the working fluid only when necessary.

In practise, I find that switching the boiler on for thirty odd minutes is sufficient to provide all day hot water for two/three baths/showers and all the kitchen sink requirements. Worst case - no heat in the tank - the hot fluid , for DWW and CH usage, is taken off the top of the heat tank and will apply the energy from the boiler directly to the taps and radiators. It's only then that the 'heat' sinks down through the heatbank to the bottom - they call it stratification.

I wouldn't have thought that'd be a significant consideration in the big picture, since it'd only be running for some tens of minutes a day and the power that it uses would be dwarfed by the energy used to heat the water it's pumping. However if you want to go the DIY heat bank route ..

Oh and I think it's Stuart Turner, not Taylor.

I'd be inclined not to bother with the Essex flanges and just connect the boiler flow (after the 3-port valve) and return to the top and bottom connections of the HW cylinder, Teed with the connections to the heat exchanger and pump.

Mine was 50kW which runs cold at full flow into the bath. I haven't done calculations or measurements to check but I think this is because the heat exchanger can't cope at that flow rate, so 100kW would be better. I think that's what DPS use and they ought to know!

I just use a standard Grundfos (clone) 15-50.

If you want to avoid all electronics you can get a flow switch capable of driving the pump directly as a part from DPS. (You can buy a PHE from them too.)

Use a 100kW plate heat ex. Use a flow switch not electronics as in the Stumble Wiki sketch. Use the top and bottom ports of the cylinder - maybe using a Surrey flange. If buying a new cylinder then use a direct cylinder with a Maganclean filter on the return to the boiler from the rads - this eliminates sludge build up in the cylinder. For a totally sludge free system use a direct cylinder using a 100kW plate heat ex to heat the cylinder - these work out cheaper than a coil and far, far more efficient in re-heat than a coil and return a low temp for condensing. The new Glow Worm stored water combis use a plate to heat the cylinder rather than an internal coil as they re-heat far quicker and give lower return temperatures for condensing. There will also be a higher cylinder water capacity without the coil.

All you need is a cheap £5 relay from Maplin. Only electrical. Look at DPS's web site they give the wiring diagram. The larger the cylinder the less storage temperature. You may get away with 70 to 75C. Set the bottom stat to say 80C and top to say 50 to 60C. The bigger the boiler, the faster the re-heat, and the lower the storage temperature.

You may have to live with it and set to suit. Some have winter and summer settings - lower storage temps in the summer.

You mean pressurised CH circuit?

Use a plate Heat Exchanger and pump to heat the low pressure direct cylinder. That is a DHW plate and a plate to heat the cylinder.

Have the CH off the heat bank, have TRVs on all rads and use a Grunfoss Alpha pump or the Wilo equiv in S/fix.

The boiler then only maintains the cylinder temperature - a simple task for the boiler.

Do you want to heat the CH from the heat bank cylinder too? Best to and use a Magnaclean filter on the CH rads return to the cylinder.

Sorry to be thick, but I don't follow your answer.

Yes, I wan't pressurised CH. So boiler and cylinder are directly connected to each other (and nothing else) and run at low pressure, but CH is seperated from the cylinder by a heat exchanger and is pressurised by a filling-loop?

Is the basic setup as follows?:

Direct connection between boiler and cylinder (unpressurised) with circulation pump. Cylinder to heat-exchanger number 1 and heat-exchanger to CH (pressurised on CH side) (pumps on both sides of PHE). Cylinder to heat-exchanger number 2 and heat-exchanger to DHW (pump on cylinder side of PHE)

Or did you mean a heat exchanger between boiler and cylinder?

As far as the control system goes, I follow the explaination of high and low tank thermostats.

Does the rest of the system go as follows?:

The time clock controls the boiler and pump between boiler and cylinder. (Does this pump run continuously when timer says "on", or only when boiler is firing plus a bit of overun?). The switch on the time clock for heating&hotwater/hotwater only controls whether pump(s) in CH circuit are also active. (Do these run continuously when timer says "on"?) TRV's on every rad, no room thermostat, and presumably a bypass on the circuit somewhere? No motorised valves anywhere.

Apologies if my questions ask again, what has already been explained.

Bzzz. No. They cannot be "directly" connect as one high and one low pressure. CH at 1 bar, cylidner at 0 1 bar.

By a heat exchnager? So not "directly" connected. Use a plate heat exchanger as the exchanger and pump to heat the cylinder.

I'll explain below.

Yep.

1. A pressurised CH system using a 3-way "diverter" valve, not a "mid-position" valve.
2. Use a direct cylinder, with an F&E tank top up.
3. To heat the cylinder have a plate heat exchanger (pref 100kW) and a normal CH pump, pumping water from the cylinder into the plate. When calling for DHW the normal boiler pump is on, as is the plate pump. The heat transfer is exceptionally high (the best you can get without being direct), giving low return temps for condensing.
4. Have a 100kW plate for the DHW and normal CH pump too, as per normal heat banks with a non-return valve after the pump and activated by a flow switch.
5. Have two cylinder stats to prevent boiler cycling.

Then a low pressure cylinder and a high pressure CH circuit with low return temps. This is heating the rads directly from the boiler. You can have the CH off the cylinder too. Best to do this but a larger cylinder is needed. Best a direct one again and heat via a plate. Got it?

Farnell will supply the flow switch

Flow Switch, makers site: The FS06

Well there's a bit more to it. I'll be running 2 showers, so will need a multi-installation pump and I also have a 1800x800 bath which takes years to fill on the std DHW and I even get bored waiting for the kitchen sink to fill sometimes - seems there would be many benefits to having a higher pressure DHW system. I can't afford a Pandora or similar BUT I can build a DIY Heat Bank system for prob about the same money as a pump, which will only benefit the showers in the house.

Yep, you're right. I shall have to fire my secretary!! ;-)

I read in another thread somewhere that if you set it up like this you'd have one of 2 situations depending how it was set up exactly.

Which is the best scenario?? Neither seems ideal. What of the dynamics inside the tank in the latter?

How much power does this pump use? Isn't it overkill to have a CH pump driving such a small circuit?

Thanks