CH Questions

[Doctor Drivel]

You should look in more tanks then.

[<me9]

You could always get the quote, frame it, hang it on the wall, and then get more realistic quotes from local dogs.

[Dave Plowman (News)]

But dribble would obviously 'see' lots of such things so he can recommend his magnetic water conditioners.

[Doctor Drivel]

....On the good ship Plowman they all did sail ....he was the skipper to no avail

....he sent the crew atop the mast ....until he spouted to them at last

....I'll keele haul ye, I will be jabers ....so cut the masts and make em into cabers

....because of cabers the ship did wreck ....I have my cabers said cap'n, so what the heck

... the breaking of law the police did pursue ....Cap'n Plowman was dragged into public view

....the judge did say in all his day ....a case so bad he did not sit ...."the mans a fool I will commit"

....the judge, he screamed, "an idiotic fool" ....the capn' now resides as uk.d-i-y newsgroup tool

[Roger]

The message from "Set Square" contains these words:

Presumably you have it set to a particular temperature over night. Could it be that which triggered a heating cycle?

[Set Square]

I did wonder about that. It's controlled at 22 degrees until nearly midnight, and then set back to 12 degrees until 07:30. I would be surprised if my well insulated house *actually* got down to 12 degrees - but I can't be certain that it didn't.

[Andy Hall]

That could be.

The installer guide says 3hrs max for the optimised start

Danfoss Randall TP7000 has settable curves - i.e. you can control the start time limit.

said that, this does seem a bit dubious, because the thermostat should have a learned history of how long it takes for the house to warm up.

However, it is limited because the thermostat can only take into account the air temperature and elapsed time.

It doesn't know directly about the behaviour of the boiler and radiators - i.e. potential cycling, nor does it know about the outside temperature.

If the rate of change of outside temperature on a day to day basis is relatively small, then this type of thermostat can optimise reasonably well. It can also calculate how long it takes to raise the temperature from a starting temperature to the desired one. This gives an approximation that will be related to the outside temperature since the starting temperature will be lower if the outside temperature is lower.

I have a system where there is an outside temperature sensor and where the room thermostat feeds the temperature via a digital proportional connection to the controller in the boiler - i.e. the boiler knows both interior and exterior temperatures. The boiler then also monitors flow and return temperatures and controls burn rate (via fan speed) and pump speed. I can get an output of all of this lot as data or a graph on a PC.

The thermostat has a night setback which is normally set to 14 degrees.

The radiators have been sized to achieve the design room temperatures for -3 degrees outside with 70 degree flow, 50 degree return which is the recommendation to improve condensing boiler efficiency.

However, on the boiler, I can set a maximum temperature as well as the slope related to compensation for outside temperature.

If I set the max boiler temperature to 85 degrees, if it is really cold, the boiler will wind itself up to that temperature quite quickly and the house will warm up quickly as well. The boiler does back off as the room temperature set point is neared to avoid overshoot. However, the boiler will be running less efficiently during the later part of the warm up as the flow gets to 85 degrees.

If I set the max boiler temperature to 70 degrees, for an equivalent outside temperature, it will come on earlier to achieve the required temperature at the time required. It will run more efficiently too since the flow never exceeds 70 degrees.

I logged the data for a few days each way for the same outside temperatures. I measured gas use by looking at the figures for the fan speed (relates to burn rate) each minute from the PC log.

It worked out to about 5-8% less gas use for the 70 degree setting during the warm up phase for a given outside temperature around 0 degrees compared with 85 degree use. At higher outside temperatures, the difference is less since less time is spent in the warm up phase. So from a saving perspective, it is not a lot - however, there is optimisation to avoid overshoot, and I think that that is one of the main areas where there is an effective waste in many systems.

I've found from a comfort perspective, I would prefer that the boiler starts raising the temperature from the setback to the set point over a longer period - which is one of the effects of a lower flow temperature. This is because I tend to sometimes get up earlier than others and I would prefer for the house to have warmed to 16-18 degrees a bit earlier than to have a steeper rate of change nearer to the set point time.

I think it's a matter for the individual. Some people are perhaps more tolerant of the set point not having been achieved when they get up but that the temperature continuing to rise. Others may prefer having a something earlier.

[Doctor Drivel]

It may be in a very cold hall or room that drop sharply in temperature to the rest of the house, sensing temperatures that are representative.

[Doctor Drivel]

It may be in a very cold hall or room that drop sharply in temperature to the rest of the house, sensing temperatures that are unrepresentative.

[Set Square]

Thanks for that Andy. It turns out that I must be blind - because I have that leaflet but hadn't previously been able to find the reference to 3 hours!

Do you have any views about the likely effects of adjusting any of the variable parameters on a CM67 - particularly Minimum ON Time, Cycle Rate, and Preportional Bandwidth?

Minimum ON Time defaults to 1 minute - but can be changed to up to 5 minutes. One minute seems a very short time to me and, even with a reasonably well balanced system, some radiators hardly get hot in that time. However, a longer time may result in more risk of overshoot? I think mine is currently set to 2 minutes and, in mildish weather, the heating will run for a bit - with some overshoot - and will then be off for a long time, with some rooms feeling quite chilly before it comes on again. [I haven't currently got any TRVs].

Cycle Rate defaults to 6 per hour - but can be set to 3, 6, 9 or 12,

Proportional Bandwidth defaults to 1.5 deg but can be increased up to 3. The leaflet seems to suggest that increasing this may reduce the overshoot in some circumstances.

Any thoughts about a methodology for arrivinmg at the best combination of these parameters?

[Andy Hall]

I think that there are several things here.

You could probably make the biggest difference to overall comfort levels by fitting TRVs. Even if the system is well balanced, it looks as though heat loss from some for whatever reason is at a greater rate than others - possibly those with more window and outside wall area? After adding TRVs, you can arrange that the radiator where the room thermostat is located is turned down somewhat and that therefore the other rooms will come up to temperature first and TRVs will start to back off.

With a simple room thermostat (bimetal or electronic) on the way up, the temperature is allowed to increase above the set point and on the way down below the set point by a certain amount. This hysteresis is to reduce the amount of cycling, but does lead to the room temperature moving up and down between these settings. Some thermostats do have a setting for the width of this.

With a proportional thermostat, the hysteresis range is , in effect, replaced by a proportional control band. Below the lower limit, the thermostat turns the boiler switch control off, and above turns it on all the time until the temperature changes to within the band. Within the proportional band, the thermostat cycles the switched power to the boiler with the on/off ratio adjusted. So for example, near the bottom of the band, it might be on for 90%, off for 10% and near the top on for 10%, off for 90%. The idea is that in steady state, the amount of heat being output on average matches the heatloss and the room temperature remains constant - i.e. an on time of 90% on a

20kW boiler equates to an average output of 18kW.

However, there are some practical issues.

The first is the time from switch on of the burner to when the radiators are contributing heat to the room. That depends on a variety of factors including boiler design (cast iron heat exchangers take longer for example), water volume in system, pump speed etc.

The second is whether the boiler would cycle using its internal thermostat sensing the water temperature. That happens if the rate at which the heat is dispersed is less than the burn rate. In extreme cases, it's not very helpful, because it means a loss of efficiency of the boiler. Also, it is effectively adding another control loop which can spoil the behaviour of the proportional controller.

Of course, having the proportional controller also deliberately cycles the boiler. Hence the notion of having adjustable cycle and minimum on times.

I would begin by checking the extent to which the boiler cycles on its internal thermostat with the thermostat full up. If that is at a rate close to or comparable with the minimum and cycle times of the controller, then it should be addressed. The normal ways are to turn up the boiler thermostat and perhaps increase pump speed. The idea is that the controller should be cycling the boiler on and off.

I'd certainly increase the proportional band width up to 3 degrees. This will make other adjustments easier.

Then it's a case of playing with the minimum on and cycle times so that the room temperature climbs to the setpoint within the proportional band reasonably quickly but does not overshoot significantly nor oscillate too much about the set point. Unfortunately, this will be a bit of a compromise because the rate of heat loss will vary with outside temperature. I would start with the lowest cycle rate and try different minimum on times. It's more efficient for the boiler to have fewer cycles, but it's at the expense of poorer maintenance of room temperature. It's also better to let the boiler get hot on each cycle to avoid unwanted condensing if it's a conventional boiler.

The adjustment's going to be iterative, I'm afraid.

I would not be at all concerned that the radiators only get slightly warm sometimes while this is working. Remember that the point of proportional control with an on/off boiler is to control the average heat output. It's therefore completely reasonable for the radiators not to get that hot when the room temperature is in the proportional band and the outside temperature is mild. Remember that below the proportional band, the boiler should be firing anyway.

On my own system, which is effectively fully proportional because the boiler modulates, the boiler can be running continuously on a flow temperature of 40 degrees sometimes. I just checked it, and it's at

50 degrees with +2 outside.

[Set Square]

Many thanks for your comprehensive reply. I'll have a go!