Tricky heating control problem

[IMM]

BMS is clearly commercial, although there are some domestic home automation products that will monitor and switch on CH, not do the controls.

They are unsuitable for this application as better and cheaper solutions are around.

Landis & Staefa do v good one, I have an old one still in its box in the garage. Forget the model number. All the large controls people will do one.

I am.

Simulated proportional feedback controls are "excellent" and do a wonderful job. BMS system have come a lot way and that they now can control, whereas the originals were monitoring and simple digital (on-off in BMS terminology) switching systems. I have experience with Landis & Gyr Visonic and Honeywell Delta systems. There is no reason why domestic BMS system cannot be cheaply available. All is needed is an I/O module where the valves, sensors, relays, etc are connected, and the module to a P/C and the appropriate software. Simple and easy. Not rocket science. The skill is setting it up, which the avearge DIYer will have a hard time doing. Getting a plumber to set one up will be near impossible.

The UFH valve will open full allowing full heat to the UFH. It will probably not be enough. If so then the rads come in to boost. When set point is near it starts to back off and the rads immediately are modulated down and then to off, with the UFH still full on.

The UFH is will be undersized, that is the point. On mild days only UFH will be required.

That is UFH for you, but this will be an undersized UFH system. The instances when the rads come in, the UFH will not be coping.

I'm not sure he has. He needs to see if a UFH system will cope first. If not abandon the UFH and stick to rads modulated by a good load compensated condensing boiler.

Compensation is feed forward control, it reduces the cold and hot spots and prevents temp under and overshoot.

That is why weather compensation is "vital" with UFH.

You can have room, wind and solar influence with a compensator. Extra modules and sensors.

That is using a two stage controller.

Two different solutions. The two stage controller will be expensive as it is commercial. Cheaper domestic compensators are available.

All clear now?

[Aidan]

Also, Try Tekmar.

know of them, but have no experience of them. Their controllers seem to be widely used for UFH. The link is the US site, they have a UK representative who should be able to advise you. I think you may find that you need different control values (PID constants ) for the UFH and the rad systems & this may be beyond the scope of such heating controllers.

[Aidan]

Translation; 'Such controllers are not currently available.'

I was hoping you could tell me where I'd find these controllers. It seems not.

It isn't cheaply available though.

Not rocket science. The skill is setting it up, which the avearge DIYer will have a hard time doing. Getting a plumber to set one up will be near impossible.

Translation;"Domestic BMS is not currently available.

Simple and easy to a man of your many indisputable talents. Completely beyond us mere mortals and, probably, the OP.

As mud.

[IMM]

Correction: they are.

Look at the makers web sites.

I know that.

are >connected, and the module to a P/C and the appropriate software. Simple and >easy. Not rocket science. The skill is setting it up, which the avearge DIYer will have a hard time doing. Getting a plumber to set one up will be near impossible.

I know that.

Not much else I can do for you.

[John Aston]

Yes, that's what I want.

I don't quite understand this. Is the UFH flow temperature a function of the internal temperature and the external temperature, or just the external temperature only?

Please could you explain what a weather compensator with room influence is?

Thanks.

[John Aston]

Yes, that's what I want.

I don't quite understand this. Is the UFH flow temperature a function of the internal temperature and the external temperature, or just the external temperature only?

Please could you explain what a weather compensator with room influence is?

Thanks.

[John Aston]

Isn't the problem, though, that the radiators have a faster response than the UFH. Assuming it's cold enough outside for the radiators to be on, they will raise the room temperature relatively quickly. The UFH will see that the room temperature is rising and throttle back on its output. The eventual "steady state" position will be determined by how quickly the radiators and UFH respond to the demand for heating.

What about using a boiler with a modulated output flow temperature (a cooler flow temperature when it's warmer outside)? The radiators would have no thermostatic valves but would simply have a flow temperature defined by the external temperature, acting as a boost to the UFH. The UFH would be controlled by the room thermostat.

The max primary flow temperature could fall off linearly from 70 deg to 50 deg, say, as the outside temperature increases from 0 to 10 deg C. (A drawback of this would be that the domestic hot water cylinder would take longer to heat, I guess, but then that must be true of any boiler with a modulating heat output.)

[Andy Hall]

In effect, this is what a modulating boiler with outside weather compensation does.

Some have settings for max flow temperature and the outside temperature above which the flow temp falls.

For example on mine there is

85 degree mode with flow starting to fall when outside exceeds +2 degrees and reaching 20 degrees flow at 22 degrees outside 70 degree mode with flow starting at 70 and reaching same end point. 50 degree mode for UFH, same as above.

The start point for sloping can be adjusted from about -8 to +8 outside temperature.

No, because what happens is if there is a hot water priority arrangement, the cylinder thermostat/sensor signals the boiler which switches the motorised valve(s) over to the cylinder and winds the boiler up to full power. Once the cylinder is reheated, the boiler switches the valve back and drops the flow temperature.

TRVs are still needed because it's desirable to have different settings per room. However, they only start closing near the set point for the room quite gently because the average flow temperature has fallen quite a bit anyway. For example on my radiator based system, the flow temperature can be around 40 degrees sometimes with the boiler running at very low output.

[Aidan]

In terms of temperature re-set, this is known as boiler reset, i.e., the boiler flow temperature is reset according to the outside air temperature. However, you would need two separate heating circuits operating at different flow temperatures, serving the UFH and the radiator systems. The UFH should work with a much lower flow temperature than the rads. I do not think that a boiler re-set system would be practical for your application.

If you have the space, system re-set would be better for this, i.e., two heating circuits each with it's own pump, mixing valve and flow temperature sensor. The flow and return connections of the 2 heating circuits, the boiler circuit and the DHWS circuit are all connected to a thermal store/buffer vessel. The boiler is operated by an immersion sensor in the buffer vessel and keeps it at, say, 82 degC. The heating circuits' flow temperature is lower than and independent of both the store temp and the other heating circuit's temp.

Yes, but you could probably accommodate this by having a system with two separate control loop modules for the rads and the UFH. The speed of response of the loops' output to a change in the input signal (air temperature) depends on the value of the PID constants programmed into the loop. So the rad heating could have a rapid response and the UFH a slow response. The rads' flow temperature should drop rapidy as the UFH output crept up.

You'd still need TRVs or stats & zone valves on both the rads and the UFH circuits.

Many UFH system are equipped with a manually set thermostatic mixer to control the flow temperature. The flow temperature remains constant for days, unless the occupants constantly adjust it. Despite this, the heat output is usually satisfactory. A system with slow automatic adjustments to the UFH flow temperature should be an improvement on this.

[John Aston]

Thanks, that would seem a promising solution. Some trial and error with the boiler's "outside compensation" settings, and some empirical testing with the TRVs would provide the commensurate boost heating on cold days.

Theoretical set-up procedure....On a cold day, with the radiators held off and the underfloor heating fully on, I could measure the difference between the internal temperature and the external temperature (say 9°C, as shown by calculations of the house's thermal performance). With the radiators on and the u.f.h. held off, I could experiment to find the boiler flow compensation slope that maintain the rooms at 9°C below the desired internal temperature. Then, switch on the radiators AND the u.f.h. and hopefully I'll get close to the control that I require (i.e. u.f.h. as the primary background heat source and radiators as a second stage boost on cold days).

A room thermostat could shut off the radiators' zone valve once the setpoint is reached.

The response of the above system will be slow compared to a house heated by radiators alone, but I can live with that. Hopefully, the boiler would have a manual override that enables you to wick up the flow temp to maximum if the outside temperature drops very rapidly.

On occasions when I want to lower all the room temperature setpoints (e.g. when I go on holiday), I'll have to change the compensation slope and turn down all the TRVs if I want the u.f.h. to be the predominant heat source. I probably won't bother. I'll just turn down the room thermostats and let the rads take precedence over the u.f.h. while I'm away.

[The Natural Philosopher]

Don't go there. IMM's command of the English language is less even than his comprehension of heating issues.

[IMM]

It is called compensation, as the boiler temperature is reset to the outside temperature

For his needs a two stage controller sensing the inside temperature with each stage operating a modulating valve would the best option.

Ourtside weather compensation would work on the UFH.

Two temp boilers are available. ECO-Hometec make one. It is basically one boiler with two boiler run stats.

That is one option. You can go direct from the boiler instead of having an intermediate thermal store. A weather compensator on the UFH modulating a valve to maintain the UFH temp. A blending valve set to say 50C is still required as a safety measure. UFH floors do not like being at 80C for long periods. The boiler can be set to 70-80C feeding the rads and the UFH system modulating away When the UFH valve hit its full travel (fully open) the rads may come in. A stat needs to be in the room to limit the room temp and cut out the rads. An outside stat could also be inserted into the wiring to hold off rads say over 12 to 15C outside.

Only if it has room influence or control.

The rads should only come in when UFH is at full belt.

[IMM]

This function is only on the top boilers and really seen on the mid-price boilers. Ravenheat (a cheap and nasty, has it)

compensation

[IMM]

If you have something useful to add please do, otherwise keep quiet. This man is after a cost effective solution and all you do is act the goat.

[John Aston]

Thanks for your reply, Aidan.

I had planned to install two separate heating circuits as you describe. The radiators would be heated (to 70°C max.) directly by the boiler flow, and the u.f.h. would have a thermostatic mixing valve to keep the underfloor pipes at 55°C. Could you expand a little bit on why it would be impractical to modulate the radiator output with a boiler reset? [Remember that the radiators and u.f.h. are not installed in separate locations. Each room has both u.f.h. AND radiators. The radiators should come on only as a boost when the underfloor heating on its own is not capable of keeping the room at 21°C (i.e. generally when it drops below 12°C outside)]

Aidan, please could you explain the purpose of the vessel and pumps. Why would you do this instead of just using the boiler pump and one zone valve per circuit? Is it because the boiler pump has insufficient residual head? Or perhaps because the pumps are variable speed and so you get some proportional control rather than just the on/off control of valves?

Is "system reset" as efficient as the "boiler reset" method that you talked about? For most of the year, a primary flow temperature of 55°C will be sufficient to heat the house. A condensing boiler will be operating more efficiently at 55°C than at 82°C. It seems a shame to heat the flow on/off at 82°C and then modulate the u.f.h. circuit down to 55°C. Wouldn't it be better to heat the boiler flow continuously at 55°C. I may be missing something here...

I can't quite see why the rads' flow temperature should drop rapidly as the UFH output creeps up. Basically, there are two control systems in parallel with different response times but with the same setpoint (room temp. = 21°). The outside temperature is constant. What part of the control loop would make the radiators respond rapidly but then drop off when the u.f.h. gets up to speed?

OK

Although, on warm days, the "boiler reset" method would result in lower flow temperatures than the thermostatic mixer setpoint, which seems desirable.

Aidan, I've seen some commercial buildings with "low velocity mixing headers" (simply looks like a big pipe with input & output ports) and pumps for each heating zone rather than zone valves. Are they used commonly in domestic situations, too? What advantages do they offer over the standard boiler pump and zone valves arrangement? Cheers.

[IMM]

Look at:

is a two stage controller. It requires a 24v supply so a 230v x 24v transformer is needed. Look at the connection diagram at the bottom. A room temp sensor (B1 in the diagram), NI 1000 range, is needed. The outputs are Y1 (first stage [UFH]) and Y2 (2nd stage [rads]). The output voltage is

0 - 10V A timer (S1) can switch it on or off.

You will also require two suitable 0 -10 volt modulating valve actuators complete with valve (22mm).

On the second stage (Y2) you could have an outside stat set to 12 to 15C, to hold off the second stage. I can't quite recall now, but if the Y2, 0 - 10v line to the valve actuator is cut, the valve will move to off (0V). That needs checking. In reality there will be no need for the outside stat as the controller will bring in either UFH or rads and modulate either to maintain the room setpoint temp. It may settle on UFH fully on (Y1 will be

10v) and the rads just on (Y2 at 1V) to give an even temperature in the room or house.

Not cheap, but worth pricing up.

1. Polygyr RWC62 controller

1 Suitable 230v x 24v transformer to power the controller and two valve actuators.

1. Mounting box for the controller. This is best a central box for all the system with all wires going in and out of the box.
2. NI 1000 range, wall mounted room temp sensor.
3. Suitable 0 - 10V actuators and valves 22mm. (Landis and Staefa will advise on this)

Sorted on the control side.

You will need 3 pumps. 1 for UFH, one for the rads and one for the primary circuit. You will need a blending valve for the UFH circuit to ensure it does not go over 50C.

When DHW calls for heat it switches in the primary pump, if not already called by the UFH and rads. Boiler temp will need to be enough to maintain DHW temp. A boiler with integral load compensation control will modulate the burner up and down to the temp differential of the flow and return. It roughly works out the heat load of the building by doing this.

When looking at expensive UFH heating controls and the extra control for the rads, this may well end up a cheaper option and seamless control of the two systems. Polygyr is an excellent control system.

[Lurch]

On Mon, 8 Nov 2004 12:49:52 -0000, "IMM" strung together this:

Still makes me larf....

[IMM]

A brainless snot makes you larf? My, oh , my!

[IMM]

The equiv to the above control are made by other makers. Intensys, make similar and probably much cheaper. Look at

make do similar stuff and the room sensor is built in. They also do

1st stage modulating and 2nd stage switched. Modulate the UFH and switch in the rads for boost.

A cheaper option: The Polygyr RWC82 controller is a two stage switch controller. Each stage does not modulate it switches. So the first could switch a zone valve to allow heat into a UFH circuit, the 2nd switches a rad pump. The RWC62 and RWC82 controllers are about £250 each. When switched, the RWC82, no expensive modulating valves and actuators need be bought.

So if using a thermal store/heat bank, a UFH zone valve is switched on/off and a CH rad pump is also switched on-off. Each zone will have set temperatures, not modulated, the UFGH pump will run all the time on the time clock. These controllers are very accurate and have a wide span of adjustment, and will maintain the room temperature very well.

[Pete C]

Hi,

Why not have two themostats, one for the rads (TMV) and one for the UFH. Set the UFH one slightly higher than that for the rads so it's preferenced.

cheers, Pete.