Boosting central heating output in one room?

[SimonH2]

I think the ideal control setup is individual electronic control for each zone - where zone is a single room in a conventional house, or maybe part of a room in an open plan layout. If the individual controls are networked back to the boiler - you can have a system where the boiler fires when any zone needs heat.

Yes, the electro-thermal-hydraulic actuators are nominally on-off. However, they have a slow opening and closing rate - and if you switch them on and off rapidly then you will get a very crude proportional control. If you have a sensitive stat with a very minimal hysteresis then it can turn the valve head on/off with quite small changes in temperature, and you may just get something close to what a basic TRV does.
IIRC we used Honeywell valves and actuators, and I checked with them on this - they said there's no restriction on duty cycle.
The ones we used had an on-delay of up to 2 minutes - ie if the wax was completely cold then it could take up to 2 minutes to open the valve. Similarly, there's a closing delay which could be up to 5 minutes if the wax was "fully hot". But switch the power off when it's only partially open and it'll start to close again quite quickly - and similarly, if you turn the power on while the valve is still closing then it'll start to open quite quickly.
So combine them with a very sensitive stat and I think they'd work - but you'd need to try it to find out. After a quick look, it does seem like they've been updated a bit since I was working with them - example here. You need to match the stroke with the valve - for the fan coil units we were using 4 port valves and 8mm stroke actuators, for radiator valves, they tend to be shorter stroke so you'd want the 4mm version.

 

[EFLImpudence]

Surely it wouldn't be the end of the world cost-wise to ballance the radiators so that living room and bedroom achieved the desired temperature all the time even when one of them is not in use.

 

[ericmark]

Surely it wouldn't be the end of the world cost-wise to ballance the radiators so that living room and bedroom achieved the desired temperature all the time even when one of them is not in use.

That was method one, with my house it worked, of sorts, but mother's house has bay windows, so if there is any sun, even in the winter, the temperature of that room can sore, not unknown before electronic heads were fitted to find the room at 28°C as the sun warmed the room as well as the central heating, the old heads marked * 1 2 3 4 5 6 were just too slow closing, remember even when closed radiator is still hot for a time.

As I said found the problem, discharged cells in the electronic head. Now back to working near enough for my needs, but I will when I get a chance try the fan again out of interest.

The idea of the wifi connection to the electronic heads is you can use ones phone to automatically change the temperature when your within a set distance from home. However since it takes over two hours to go from 16°C to 19°C in real terms not a practical option. Not many people work that far away from home, OK you could manually alter the temperature, or you could just suffer for an hour or so with a cold house, but manually one would forget, and whole idea is not to come home to cold house.

Now the gas fire can heat the room easy in half an hour at 4.5 kW and the boiler is normally around 20 kW so the boiler should be able to heat one room in half an hour, and to be fair in my own house with a Myson fan assisted radiator it does, so the stumbling block is the radiator.

Thanks to "SimonH2" I can now look at making the Myson controlled by a hub, however not sure I want to, now retired so no real saving turning it on/off with wifi, a simple programmable thermostat is all that is required, even the cheap and nasty Hive would work in my house, boiler is not modulating, however I would fit Nest just in case the boiler is changed, may as well have an OpenTherm thermostat to future proof.

But to test the idea so I know if it works would be good. What one thinks will happen, is not always what does happen. I was looking at keeping my beer at the right temperature, I thought using an old fridge/freezer not plugged in, but just as an insulated box would work well, wrong, brewing beer produces heat, so at start the temperature would rocket, seeing the over shoot with an 8W bulb as heater, I expected if I used the freezers motor I would get a see saw effect with freezer motor and bulb alternating again wrong, the air temperature dropped to 8°C but the actually fermenter held at 19.5°C as set, OK with cooling since a frost free freezer there is a fan running and the cooling coils are behind a cover, but on cooling not even 0.1°C over shoot. So having seen with beer brewing how wrong I was, would like to test before saying it will or will not work.

I was caught by all the hipe on wifi controls, I actually thought you could save money turning heating down when not required, however when the central heating failed one day, (Mother turned the thermostat down) it was 5 hours before I realised there was something wrong, the house simply does not cool down, or heat up quickly. So really no point in cheap thermostats like Hive, as said OK with Nest with OpenTherm but Hive is to my mind pointless.

 

[SimonH2]

However since it takes over two hours to go from 16°C to 19°C

That suggests the radiators are somewhat undersized. But in our house, try "doesn't reach 19˚ at all" even after 6 hours - such is the scale of the screw up make by previous owners in fitting completely inadequate heating.

but mother's house has bay windows, so if there is any sun, even in the winter, the temperature of that room can sore

Our longe has a sizeable south facing window, a huge west facing windows & french doors, and two velux roof windows. it was a sunny day on one of our viewings and even with both veluxes open it was "rather warm" - summer could be interesting. I fully anticipate days when teh lounge is hot but the rest of the house is cold. The rest of the house is solid brick so a lot of thermal mass, not sure about the extension but I think it's medium density concrete so a bit less thermal mass.

 

[ericmark]

Today got some good quality batteries, don't want them to leak and damage the electronic TRV. This has pointed out two things, one was voltage now showing 3.19 volt on both valves, it has never shown so high, instructions say normal is 3.1 volt it does not say at what voltage to change I was at 2.86 volts I think, but it does say "red flashes twice every 5 seconds = low battery needs replacing" and "red flashes once every 5 seconds = dead battery needs replacing" neither had happened so had assumed battery was OK. It also states if the battery becomes dead the plunger will fully retract, so in theory dead battery should cause room to over heat, and the room was getting too cool. It also says do not use rechargeable. I don't know what voltage is considered as discharged, I will guess more than 1.2 volts per cell, which is why it says do not use rechargeable. I will guess the cells were old, still within the use by date, but for Duracell that it seems is a very long time. It was less than 3 volt from new.

The other thing, when changing batteries the plunger fully retracts as it goes through the set-up procedure, one has to wait for the orange LED then press the button for 4 seconds while it is doing this the valve is fully opened, and the radiator got stinking hot, and the room temperature rocketed, it would seem radiators are plenty big enough, it is the anti-hysteresis software which delays the opening and in doing so also slows down how quick the radiator will heat the room.

The electronic TRV head it seems has two sensors, one for water and one for air, so it can compensate for the heat from the radiator, since they can be mounted is different positions it seems likely there is some learning involved. So on further thought I think better not to try fan on radiator as it may mess up this learning process, and now cells have been changed so it has a new battery, and it works, seems prudent to leave well alone.

 

[ban-all-sheds]

Today got some good quality batteries, don't want them to leak and damage the electronic TRV.

Lithium?

 

[ericmark]

Lithium?

Not a clue, not zinc carbon that's all I can say. Duracell Alkaline not a clue as to what metals are used inside them.

Ups it seems they are zinc carbon looking at this which means the advert with bunny is saying zinc carbon is better than zinc carbon?

However the steel shell does mean unlikely to leak which is why I used them

 

[ban-all-sheds]

http://www.energizer.com/batteries/energizer-ultimate-lithium-batteries

 

[JohnD]

IIRC lithium cells are 1.5v and alkaline are 1.2v, in some cases the extra voltage really does make a difference. I use lithium in a wireless outdoor PIR and they keep working much longer than alkaline.

 

[ericmark]

The Lithium battery it seems does have a longer shelf life and a larger mA/h capacity, however the Duracell I used was standard one, which is alkaline, there are more details but unless you down load the data sheet linked to, the manufacturers seem to hide the mA/h rating with primary cells, with secondary cells you can read the packet and see how much power is in them but with primary cells it seems to be a lottery, at the shop there is nothing to tell you if 1000 mA/h or 3000 mA/h.

In real terms one can't remember if battery changed 1, 2, 3 - 20 years ago, and since you don't want it to fail easy way is to change once a year, even if it will last 3 years, so as long as battery will last a year, any extra is wasted.

 

[SimonH2]

... the valve is fully opened, and the radiator got stinking hot, and the room temperature rocketed, it would seem radiators are plenty big enough, it is the anti-hysteresis software which delays the opening and in doing so also slows down how quick the radiator will heat the room.

It's not anti-hysteresis, it's anti-overshoot.
Basically, if the valve opens too quickly, then as you've seen, the radiator will heat up quickly and they typically store quite a bit of heat. So the room will overheat, the valve will have to shut down fully, and when the room eventually cools down again then the cycle will repeat. The same can happen with the basic TRVs, and is typically an indication that the lockshield is open too far.
It may well be that the electronic TRV does in fact do some learning - tuning a PID controller loop to try and get the fastest operation without instability.

 

[ericmark]

It's not anti-hysteresis, it's anti-overshoot.

I thought hysteresis was the sine wave caused when the system over shoots? i.e. they are the same thing?

With modulating methods to stop the over shoot it is not too bad, but where a mark/space method is used, then having two systems together can result in an increase in the over shoot.

However more of a problem is it increases the time taken, in theory a learning thermostat can work out when heating from 16˚C to 20˚C when the temperature reaches 19.2˚C it can turn down the flow to what it has already worked out will keep the temperature static and turns down to that rate. However I would not expect the so called smart thermostats to really be that smart.

In order to keep my beer at the right temperature the heater needs to be as light as possible, even then often the heat pump needs to switch on to remove the excess heat due to the over shoot, oddly it does not over shoot on the cooling cycle.

So with the rooms, I have to manually set the temperature to 24˚C for 1.5 hours then down to 20˚C so it heats up quicker without over shooting. However that only works if the room has actually cooled.

 

[SimonH2]

I thought hysteresis was the sine wave caused when the system over shoots? i.e. they are the same thing?

Not really. While the overshoot is dependent on prior conditions, it's still a linear and modelable system. Hysteresis is normally added to thermostats to stop them flapping between stats too quickly - but that means the stat of the stat at it's setpoint depends on which direction the setpoint was approached from. Ie, if the room is too hot and cools down, you can have the stat off with the room at the desired temp; but if the room was cooler and is being warmed up, you can have the room at the very same temperature but with the stat on.
https://en.wikipedia.org/wiki/Hysteresis

Overshoot is not really hysteresis in the classical sense, it's the instability caused by a control system that isn't matched to the system it's controlling. Regardless of whether the room was hot or cool when you started, if you hold the control (valve) in a set position, then the temperature will always settle at the same temperature (for a given set of conditions).

 

[ericmark]

Showing hysteresis and

showing how using a modulating controller rather than on/off how hysteresis is removed, so over shoot refers to half of the hysteresis cycle, under shoot I suppose refers to other half, I found when brewing beer yes there was simply over shoot, really never worked out why, likely due to heat from the fermenting, but when heating it would easy go over, but when cooling it never seemed to, at least not over 0.1°C which was what the controller shows.

I have found in my mothers hall, the cure was to use both a TRV and wall thermostat, the TRV is set so it starts to reduce flow before the set point of the wall thermostat, so that last degree or so takes a lot longer to reach, however the problem is doors, with doors closed it works A1, but if a door is chocked open by carers then the higher temperature of the living rooms enters the hall switching off the whole heating, so I have a second wall thermostat in the kitchen, on the side of the oven, also a silly place for a thermostat, but it means there is heating over night when the hall one is switched off with programmer, and if doors left open it is set higher than hall so keeps heating running.

I would say all in all doors are the big problem, some times open, some times closed and this alters how the heating works, you can fit all the controls you want, but variables like doors open or closed mean there is a limit to the control, except with hot air central heating, the massive fan which sucked air from living room and pumped it around the house with vents in the bottom of every door to allow air to return resulted in whole house at a single temperature, the controls in each room in theory allowed you to set bedrooms cooler, in practice it did not work, plus cost a fortune to run, in those days no double glazing and moving hot air around the room resulted in windows cooling the air.

Anyway it would seem some thing was wrong with the TRV heads, once batteries were changed, the rooms are doing exactly what is wanted, the warning lights for discharged battery were not on, but changing them cured all problems, may be because it re-sets its self on battery change? But what ever it was, it's now working OK without the need to boost any output. So since working, I am leaving it alone.