TRVs and thermostats

[adriang]

I'm in the process of getting my boiler (30 year old Ideal with HW tank, thermostat in hall, TRVs on all rads) replaced. The proposal is for a Vaillant eco+ 837 combi and to renew all the TRVs, most of which are jammed open.

I'm keen to make the control system work properly. At present, hot water is not a problem but the CH temperature is controlled more by the timer than the thermostat - it's set for a couple of hours morning and evening, if we put it on during the day the house is usually too hot in the evening. This is because the hall is cold so the thermostat doesn't cut out until the bedrooms are hot., and the TRVs don't work very well (and in any case are outwitted by rising warm air).

It seems to me that using the return flow to control the boiler will be wasteful : it can't measure usage unless its heating it, and I like a very low level of heating so I don't want it constantly warm.

The weather compensating systems might do the job in that they might be able to predict what the boiler level should be in order to get the job done (and then cut off if they find many rooms are set cold). But my wife likes some internal control of temperature : she tends to use the room stat as a colder/hotter switch. I'm not sure how that fits with a weather compensator system.

I can see the point of the distributed systems with wireless valves on rads, but I think that's going to get too expensive, I don't like lots of replaceable batteries, and some valves would be a pain to wire.

I'm thinking of putting 3 zone valves (upstairs, downstairs, bathroom) partly because the old pipework makes that very easy. I'm then looking for a thermostat solution that will make sense.

Any ideas ?

 

[D_Hailsham]

This is because the hall is cold so the thermostat doesn't cut out until the bedrooms are hot., and the TRVs don't work very well (and in any case are outwitted by rising warm air).

1. TRVs react to the air temperature, not the water temperature. So the rising war air will make the TRV shut off earlier. If you want to stop this, you must shut all the bedroom doors.
2. If the room stat shut off earlier than the TRVs, the rooms controlled by TRV would never reach temperature.

It seems to me that using the return flow to control the boiler will be wasteful : it can't measure usage unless its heating it, and I like a very low level of heating so I don't want it constantly warm.

The difference between flow and return temperatures is a measure of the heat lost.

The weather compensating systems might do the job in that they might be able to predict what the boiler level should be in order to get the job done (and then cut off if they find many rooms are set cold).

All weather compensation systems need tuning as each house has different heat loss characteristics.

But my wife likes some internal control of temperature : she tends to use the room stat as a colder/hotter switch. I'm not sure how that fits with a weather compensator system.

Sounds as if your existing room stat has a very wide hysteresis (difference between on and off temperatures) of may be 5 or 6 degrees. Most stats keep the boiler running until the set temperature is reached and then turn the boiler off. The radiators, however, continue giving off heat, so the room temperature continue to rise. Result your wife feels too hot and turns the stat down until it clicks off. Unfortunately, due to the hysteresis of the stat the off point may be 5 or 6 degrees lower. So the temperature drops and your wife start to feel cold and turn the thermostat up and the merry-go-round continues.

There are now thermostats available which take account of the fact that radiators continue to give out heat and have a hysteresis of ±0.5C, for example the Honeywell CM900 series.

I'm thinking of putting 3 zone valves (upstairs, downstairs, bathroom) partly because the old pipework makes that very easy. I'm then looking for a thermostat solution that will make sense.

Why the bathroom?

How big is your house - number of floors and area of each floor?

Presumably you would still have TRVs on the rads not directly controlled by the thermostat?

You could control the zone valves by their local room stat and use the internal switch on each zone valve to control the boiler.

 

[adriang]

1. TRVs react to the air temperature, not the water temperature. So the rising war air will make the TRV shut off earlier. If you want to stop this, you must shut all the bedroom doors.

The warm air from the hall (and other ground floor rooms) continues to heat upstairs even after all the TRVs have shut off. Agreed, shutting doors might help. Easily forgotten, though.

2. If the room stat shut off earlier than the TRVs, the rooms controlled by TRV would never reach temperature.

Yes, the hall is a reasonable place to ensure that there's always heat available for any TRV that wants it. But the effect is that it's hard to keep the temperature LOW enough.

The difference between flow and return temperatures is a measure of the heat lost.

Of course, but it's therefore necessary to send warm water around the house in order to measure the heat loss. This warms the house (perhaps made worse by ineffective TRVs)

All weather compensation systems need tuning as each house has different heat loss characteristics.

Fair enough. How difficult is that to do ? Do they attempt to tune themselves ? If perfectly tuned, how good are they ?

Sounds as if your existing room stat has a very wide hysteresis (difference between on and off temperatures) of may be 5 or 6 degrees.
...

Could be, though I'm thinking on a longer scale than immediate comfort : either the system doesn't actually control to a reasonable temperature or the desirable temperature varies according to other factors. What I'm thinking is that if weather compensation and TRVs between them keep the temperature constant, what opportunity is there to adjust actual level ? It would be nice if adjusting a TRV would do it, but the vague markings and long lag time mean that I suspect they'll be used as on/off valves if they work at all..

There are now thermostats available which take account of the fact that radiators continue to give out heat and have a hysteresis of ±0.5C, for example the Honeywell CM900 series.

That sounds interesting. Vaillant do a wireless programmable stat, which appears to communicate with the boiler more than a simple contact pair would. Is that a good option ?

Why the bathroom?

Yes, it's overkill. But : I have a stock of decent valves, the pipework already exists (former tank pipework) and there's some sense in heating the towel rail even when other circuits are off. If I didn't have the valves, that would be a good reason to have the bathroom as the only uncontrolled circuit.

How big is your house - number of floors and area of each floor?

10m x 10m, 2 floors, 4 beds (3 beds rather small). detached. Double glazing, foam filled walls, insulated loft.

Presumably you would still have TRVs on the rads not directly controlled by the thermostat?

Currently I'm expecting to put TRVs everywhere, if only because it's easier to fit them now than later and I haven't decided where the thermostat should be.

You could control the zone valves by their local room stat and use the internal switch on each zone valve to control the boiler.

I don't really want a zone valve for every room (mostly because of difficulty of fitting them) so I'm looking for a way to control them usefully. Currently my favourite choice is to measure temperature in each room and turn on valves when any room in that zone is cold, turn on boiler when either zone is cold. This is technically OK but I don't know if it will be good enough. It still doesn't really handle the problem of the upstairs getting hot even when it's all off, but I might be willing to limit the downstairs temperature when the upstairs was overheating.

 

[swidders]

Start off by simply replacing the TRVs and assessing the effectiveness before you start getting complicated, keeping one radiator with ordinary valves in the same room as the stat.

 

[D_Hailsham]

The warm air from the hall (and other ground floor rooms) continues to heat upstairs even after all the TRVs have shut off. Agreed, shutting doors might help. Easily forgotten, though.

I was referring to your to your statement about the heat rising "outwitting" the TRVs, which I took to mean that you thought the TRVs reacted to the water temperature - a not uncommon fallacy. The rising air will mean that the room gets warm quicker, so the TRVs shut off earlier. I have found that, if you leave all the doors open, the whole house will stabilize at the same temperature.

Yes, the hall is a reasonable place to ensure that there's always heat available for any TRV that wants it. But the effect is that it's hard to keep the temperature LOW enough.

Two conditions have to be met:

The rads have to be sized correctly
The system has to be properly balanced

If that is correct, all rooms will heat up at approximately the same rate, so the TRVs will kick in at the same time and the wall stat will operate. An ideal situation, I know, but there are ways round it, particularly with oversized radiators.

Of course, but it's therefore necessary to send warm water around the house in order to measure the heat loss. This warms the house (perhaps made worse by ineffective TRVs)

That is stating the obvious, so you have lost me! I'm not sure of the point you are trying to make.

All weather compensation systems need tuning as each house has different heat loss characteristics.

Fair enough. How difficult is that to do ? Do they attempt to tune themselves ? If perfectly tuned, how good are they?

Tuning has to be done manually - should be done by the installer. Some people swear by them; some swear at them!

I'm thinking on a longer scale than immediate comfort : either the system doesn't actually control to a reasonable temperature or the desirable temperature varies according to other factors. What I'm thinking is that if weather compensation and TRVs between them keep the temperature constant, what opportunity is there to adjust actual level ? It would be nice if adjusting a TRV would do it, but the vague markings and long lag time mean that I suspect they'll be used as on/off valves if they work at all

Comfort is a psychological phenomena. Someone will feel warm at 23C in the summer but complain that they are cold in the winter. Changes in temperature, which promote draughts, also make you feel uncomfortable.
All weather compensation does is adjust the flow temperature of the water so the radiators are producing the right amount of heat; you still need some way of measuring the internal temperature so the boiler is turned off when not needed. Good TRVs have a time lag of about 20 minutes, they are not designed to be turned up and down. TRVs are not on/off switches, they open and close like a car throttle.

Vaillant do a wireless programmable stat, which appears to communicate with the boiler more than a simple contact pair would. Is that a good option?

Vaillant kit is good, but make sure they will work correctly with a multi-zoned installation. One person on this site had problems with the Vaillant weather compensator because it would only work in one zone.

there's some sense in heating the towel rail even when other circuits are off.

Good point

10m x 10m, 2 floors, 4 beds (3 beds rather small). detached. Double glazing, foam filled walls, insulated loft.

Your heat loss will be between 11kW and 12kW. As your floor area is over 150m², zoning is recommended - it would be obligatory if it was a new installation. I suggest you find out the total heat output of your rads - the Stelrad Elite Catalogue is a good reference. It does not matter if the rads are larger than required.

Currently I'm expecting to put TRVs everywhere, if only because it's easier to fit them now than later and I haven't decided where the thermostat should be.

A TRV should not be in the area where the thermostat is.

I don't really want a zone valve for every room (mostly because of difficulty of fitting them)

I was not suggesting that! A thermostat on the landing, controlling the upstairs zone valve with TRVs in all rooms, and similar downstairs should be sufficient.

Currently my favourite choice is to measure temperature in each room and turn on valves when any room in that zone is cold,

The TRV will do that; it maintains the room at the set temperature with the caveat about rising heat.

turn on boiler when either zone is cold

The thermostat does that.

It still doesn't really handle the problem of the upstairs getting hot even when it's all off, but I might be willing to limit the downstairs temperature when the upstairs was overheating.

You can't change the laws of physics, though you can ameliorate their effect by keeping doors shut. I don't think your family would be too happy if you kept the downstairs cold just so the upstairs did not get too hot. Provided the rads downstairs are correctly sized, there should not be too much excess heat travelling upstairs. You should, in any case, have the upstairs TRVs set to a lower temperature than those downstairs.

Swidders suggestion is very good. Keep it simple

 

[JoeMar]

Start off by simply replacing the TRVs and assessing the effectiveness before you start getting complicated, keeping one radiator with ordinary valves in the same room as the stat.

to add to what swidders and hailsham have already said, do go for a Vaillant Controller that supports weather compensation. It takes the guess work out of setting the flow temperature and will reduce your gas consumption.

 

[adriang]

I was referring to your to your statement about the heat rising "outwitting" the TRVs, which I took to mean that you thought the TRVs reacted to the water temperature - a not uncommon fallacy. The rising air will mean that the room gets warm quicker, so the TRVs shut off earlier. I have found that, if you leave all the doors open, the whole house will stabilize at the same temperature.

Yes, I meant that the Trv upstairs would shut off even though it was the one downstairs that was sourcing the heat. (of course, the upstairs one should be off too ..)

Of course, but it's therefore necessary to send warm water around the house in order to measure the heat loss. This warms the house (perhaps made worse by ineffective TRVs)

That is stating the obvious, so you have lost me! I'm not sure of the point you are trying to make.

That, with TRVs and return flow measurement only, the boiler has to circulate warm water in order to measure whether warm water is required. That heats the house up, even if all the TRVs are off, and wastes gas in cycling the boiler. I'm looking for the best way to avoid it.

Vaillant kit is good, but make sure they will work correctly with a multi-zoned installation. One person on this site had problems with the Vaillant weather compensator because it would only work in one zone.

OK, thanks, I'll see if I can find the posting. Do these compensators look at the indoor temperature as well as outdoor, then ? And if they use the indoor temperature, do they have a sensor that's away from the boiler ? My boiler will be in an unheated garage.

Swidders suggestion is very good. Keep it simple

OK, I'll try that . It certainly seems worth putting the minimum in at first, but the simplest and most convenient approach would be a programmable wireless thermostat. If I add a weather compensator (which I think also has a programmer) and additional temperature sensing that might be redundant. Perhaps the simplest possible start is to use the existing thermostat and timer.

Finally, the software that Vaillant sell for tweaking and measuring the boiler sounds interesting .. is it worth getting ? I'm more interested in measuring than adjusting.