Gas on constantly or on and off...?

stevemcol

Cardew wrote: »

This then raises the usual retort.

If you go away from your house for, say, 10 years, is it more efficient to keep the house at a 'base' temperature while you are away?

How about a year? A month? A week? 24 hours? 12hours?

At what point do the Laws of Thermodynamics not apply?

Cardew. That may be the usual retort but it's not a particularly intelligent one. It is obvious that temperatures in a house won't continue falling indefinitely. Certainly they won't fall indefinitely below OAT.
I agree that nautural laws govern this subject. However, in this context there are other features at play. A significant one for my house certainly, is the ability of the heating to recover the temperature at all (or before bed time at least) if the house temperture is allowed to fall below a base value.
Now I accept that leaving the heating on to maintain a mean temperature during the day is not the answer. It's a matter of knowing how early, for varying conditions, to have your heating come on in order to recover a comfortable temperature. It almost needs a 'learning' control programmer.

Cardew

stevemcol wrote: »

Cardew. That may be the usual retort but it's not a particularly intelligent one. It is obvious that temperatures in a house won't continue falling indefinitely. Certainly they won't fall indefinitely below OAT.
I agree that nautural laws govern this subject. However, in this context there are other features at play. A significant one for my house certainly, is the ability of the heating to recover the temperature at all (or before bed time at least) if the house temperture is allowed to fall below a base value.
Now I accept that leaving the heating on to maintain a mean temperature during the day is not the answer. It's a matter of knowing how early, for varying conditions, to have your heating come on in order to recover a comfortable temperature. It almost needs a 'learning' control programmer.

The 'on 24/7 or timed' debate revolves around this Urban Myth that somehow it uses more energy to heat a house up from cold than to keep it at a constant temperature. This was the original question:

Someone has recently told me that it is cheaper to leave the boiler on and on the thermostat all day even when we are out of the house as this avoids the system heating the whole house from freezing and simply tops it up to maintain the temperature.

What is your experiences of this, will this result in a cheaper bill?

We go through this loop with a similar query several times a year.
Apart from the first sentence;) there is nothing in your post to dispute. However my ‘retort’ is relevant to the OP’s question, whereas your post doesn’t address the exam question!

Obviously how anyone runs their heating is a personal decision and is usually a compromise between convenience and cost. Your situation is typical, let the house get too cold and it takes too long to get it up to temperature.

The only issue I am disputing is the urban myth that ‘it is cheaper to leave a property at a constant temperature as it costs less to constantly top up, than heating up from cold.’

How people operate their system is of course totally their prerogative.

Hooloovoo

stevemcol wrote: »

It's a matter of knowing how early, for varying conditions, to have your heating come on in order to recover a comfortable temperature. It almost needs a 'learning' control programmer.

Exactly!

All of the current heating control systems are rubbish. Even the latest programmable thermostats like the one I have fitted to control my combi boiler. I shouldn't have to set the time at which I want the heating to come on. I want to be able to set the time at which I want the house to be at a certain temperature.

So if I know I am getting in from work at 6pm that's the time I want the house to be up to temperature. Who knows what time that means the heating should come on? Well, the thermostat should know. It should know how long it took to warm up yesterday, so it should know roughly what time to fire the heating today. The OAT doesn't change that drastically on a day-to-day basis so it's easy to make a reasonably accurate prediction model of the heating system.

Outside the house I have a weather station that monitors OAT, humidity, wind velocity, and rainfall. The base station is in my living room and also monitors the indoor temperature and humidity. It's connected to a mediacentre PC that logs all the data. So I can plot how the temperature of my house varies throughout the day, and see exactly how fast my heating increases the temperature of the house, and how fast it cools when it switches off. It only takes me a second to look at the data and see what time the heating needs to come on and switch off in the evening, and I can adjust my programmable thermostat accordingly. It also means if I set a base temperature of 16C while I'm out at work, I can look back and see how often the boiler kicked in to maintain that temperature, and estimate the duty cycle.

A lot of my job involves writing data analysis software and control systems. I'm pretty confident that I could write a learning thermostat that would easily beat any of the current systems for both comfort *and* efficiency. But I don't work in the heating trade. So I'll just have to wait until thermostat technology catches up. Or somehow get my PC to control my heating.

stevemcol

Hooloovoo

I hope to see you on Dragon's Den in the coming months.
I'll be your first customer. Just so you know where I'm at 'I'm in'.

Hooloovoo

stevemcol wrote: »

I hope to see you on Dragon's Den in the coming months. I'll be your first customer. Just so you know where I'm at 'I'm in'.

Worcester are *almost* there, but not quite.

I have their "Optimiser" Digistat. For the first temperature change of the day after the night setting, it aims to get the temperature up to the requested level within one hour of your setpoint time. So if the setpoint goes up from 12C to 18C at 0700 it aims to reach 18C by 0800. If it managed to get to 18C by 0730 the previous day, then it holds off firing the heating until 0730 because it knows it can get there by 0800, saving you money by running the heating for half an hour less than you requested.

But there is a major flaw in this design. Even with some very large radiators my house only increases in temperature by around 2C/h. If the temperature has dropped to 12C overnight, there's no way my heating is ever going to get up to the set point in only one hour never mind any quicker so in practise the "optimising" feature of the thermostat never gets to do anything.

They need to do three things. First they need to extend the "optimising" look-ahead period to way more than one hour, they need to change the setpoint time from heating-on to heating-completed, and finally make the optimiser function work on all setpoints not just the morning one.

This programmer is around four years old now. Maybe they have something new that actually does what I want and I'm just not up to date.

DVardysShadow

stevemcol wrote: »

Cardew. That may be the usual retort but it's not a particularly intelligent one. It is obvious that temperatures in a house won't continue falling indefinitely. Certainly they won't fall indefinitely below OAT.
I agree that nautural laws govern this subject. However, in this context there are other features at play. A significant one for my house certainly, is the ability of the heating to recover the temperature at all (or before bed time at least) if the house temperture is allowed to fall below a base value.
Now I accept that leaving the heating on to maintain a mean temperature during the day is not the answer. It's a matter of knowing how early, for varying conditions, to have your heating come on in order to recover a comfortable temperature. It almost needs a 'learning' control programmer.

Actually, Cardew's retort is quite intelligent, because it does address the problem you have. Cardew is effectively asking what is the thermal time constant of a house. In your case, it looks like it is longer that the off period you would program.

But your problem is not whether it is more economical to leave the heating on or off during the day, it is that your system is not capable of getting the place warm. When it is necessary to leave the heating on all day, this will always be because otherwise you cannot achieve a desired level of comfort - and never because it is more economic.

Cardew

This subject should be a 'sticky' and simply state:

'Having heating(or hot water) timed is cheaper than having it on constantly!'

Anyone who needs to ask the question -i.e. which is cheaper - quite clearly does not have a scientific or engineering background and should be given an uncomplicated answer that they can understand.

Yet every thread seems to develop into 'I have both sets of bedridden grandparents in one bedroom and new born triplets in the other bedroom of an uninsulated house and need heating on 24/7'!!!!

This plaintive cry is followed by scores of posts describing how they set their heating for their particular set of circumstances.

stevemcol

I'm getting nearer a solution. My heating progammer though not intelligent allows 6 periods each day. I've set a warm up profile which is sufficient to get things warm on a cold day but shouldn't waste too much energy on a warmer day. Programming 15° at 4pm (first on time), 17° at 5, 19 at 6. I've also wound up the rad stat on a large radiator in a room we don't use much which has improved overall warm up time.

I can't believe there aren't more theromostats and programmers out there that can learn.

jem16

Hooloovoo wrote: »

They need to do three things. First they need to extend the "optimising" look-ahead period to way more than one hour, they need to change the setpoint time from heating-on to heating-completed, and finally make the optimiser function work on all setpoints not just the morning one.

I have the Honeywell CM927 programmable thermostat.

I have its optimising feature set to on. So far it's been working well - heating varies when it comes on so it's at the set temperature by the time specified. I've seen it vary from around 15 minutes to around 90 minutes so far.

I only had it installed in April so this is the first winter period. So afr so good though.

Hooloovoo

jem16 wrote: »

I have its optimising feature set to on. So far it's been working well - heating varies when it comes on so it's at the set temperature by the time specified. I've seen it vary from around 15 minutes to around 90 minutes so far.

Is that only on the morning setpoint though, or all of them?

For example, when it goes from the daytime setpoint to the evening setpoint ready for you to get home from work, does it optimise that one as well?