Alternative to electronic weather compensation on a thermal store

[mistadave]

Now I thought of this as I was designing my thermal store setup. I was considering weather compensation controlled mixer valve coming off the tank to supply the central heating at the lowest possible temperature for the conditions. This cuts down on the amount of heat drawn off the tank, and also the amount of unnecessary mixing of the water in the store.

The disadvantage is that it costs a lot, and its something else using electricity all the time or to go wrong.

Instead I thought I could use a normal thermostatic mixer to supply the heating, and then a thermostatic diverting valve (or a mixer valve in reverse) on the return. My theory goes like this:

Assuming you have a properly sized and balanced central heating system, in my case I am aiming for a flow of 40 and return of 30. This is obviously set up for a worst case -5 outside.

If it were warmer outside, a weather compensation system would lower the flow temperature to say, 35 degrees and you might get a return of about 28 due to the lower delta T with the room temperature. A thermostatic mixer valve on the other hand could be manually adjusted to change the flow temp for different seasons, but assume you left it fixed, and then used a diverting valve on the return set to 30 degrees. Now on a cold day or if the room is being heated from cold, the return temp would be 30 or below as the rads and UFH are loosing a lot of heat, the valve then diverts this flow back to the bottom of the store.

But if the room temperature is higher the return temperature will be higher as well and so if the return is then over 30 degrees it will be diverted back round, so reducing the flow temperature and also reducing the amount of mixing going on in the thermal store.

Obviously it doesn't have the range of automatic adjustment that proper weather compensation has, but you can always manually adjust the settings for both valves if needed, and it has the advantage of adjusting the flow temperature as the room temperature increases - giving the most heat to start with and then decreasing the flow temp as the room reaches the correct temperature which would help prevent overshooting.

So, that is my theory, is it practical in the real world?

[wookey]

This is conceptually the same as UFH manifolds with adustable return temps, I think. So yes it does work. You can get them with manual control or variable electric control. The latter is unhelpfully expensive.

Doing this basically sets a max return temp to the bottom of the tank, which is a good thing. I'm not at all sure it would work as a weather compensator as such- as it'd always do the same thing - return water to tank at set temp, and circirculate it until said temp is reached. Water would go round fewer times in colder weather as it'll be coolling faster due to higher rate of heat-loses in rooms, and thus from emitters.

[SimonHobson]

When I put the store in the flat, I put a TMV in the feed to the CH so it blends the return with fresh warm water from the store. I also fitted a modulating pump and TRVs all round.
So the hot water drawn from the store should be only enough to maintain the flow temperature - and if the return temp goes up then you'll need less hot to keep the flow up to temp. If demand is low, the TRVs will close down and the overall flow will drop accordingly.

My observations were that under light heating demands, the return from the rads was little above room temperature - with the rads hot from the inlet, up one side and part way along the top, and noticeably cold at the exit end. This is due to the low flow rates allowing the rads to cool the water down to room temp by the time it leaves the rad. Obviously this won't be the case if your rads are better sized for the load - I think the flat has lower heat losses than were designed for*
Contrary to what you might expect, return temperatures go up with heating demand Under high heating demands, the flow rate is higher and so the water can reach the outlet before it's been cooled to room temperature.

I've had no complaints from the current tenants.

* After fitting the store I ran the system on the immersion heater for a while so as to measure input accurately. I reckon the heat demand even in the cold weather of Dec 10 was only about 2kW - vs being originally supplied by a (combi) boiler rated to 27kW max and I think about 9.something minimum output

I did "muse" a bit about a different setup instead of the fixed TMV. I wondered if a flow sensor could be used to adjust the flow temperature setpoint - thus avoiding running the rads with such a low flow rate and giving them a more even temperature. Ie, having the whole radiator "lukewarm" instead of a bit of it hot and the rest cold. I was thinking along the lines of setting a minimum flow rate for the system, determining a desired flow temperature (possibly applying some outside temperature compensation), and then using a motor or thermo-hydraulic actuator on the mixing valve to control the flow to that temperature.
Once you've gone electronic control, the possibilities are "somewhat expanded".

[derekmt]

there were sophiscated control system before electronics. its quite feasable to make it all mechanical. find some old control books