Our home heating system sucks!

Sometimes rooms are cold and don’t heat up when you’d expect – sometimes they are too hot (though I suspect this may be due to my wife and I turning the radiator valves all the way up in a desperate attempt to solve the first problem).  Anyhow I’m working towards understanding of the dynamics of our rather over-complex heating system and ultimately I’d like to improve it somehow. We have a conventional (for Edinburgh, UK) home, built around 1860, with large windows and relatively poor (by modern standards) insulation. We’ve done a lot to improve its thermal efficiency but it is particularly hard to maintain a sensible temperature in key rooms throughout the year.

The “up-to-temperature” Problem

It seems to me that the normal UK system of thermostats in the common areas (hallways in our case) and thermostatic radiator valves (TRVs) doesn’t cope well with our situation. Sometimes the thermostats are indicating up-to-temperature but some of the radiator TRVs are wanting hot water to be circulating so they can get their room temperature up. Sometimes the TRVs are all closed down but the pumps and boiler/furnace are working away furiously.

I can understand that in a modern house with good insulation the temperature from room to room might not vary so much and the system might work well but in our place that’s anything but the truth.

Problems and Questions

I’ve talked this problem though with a number of people – most recently with Pilgrim Beart – founder of 1248.io and of AlertMe  (the company behind the British Gas Hive product) – who knows more than a little about this stuff. He suggested that ultimately I might need to ditch the hall-based thermostats and control the boiler and pumps based on broader room temperature information.  Which sounds great in principle – but how would I fit this into an existing system?

So now I’m starting to investigate ways to achieve this. The problems and questions seem to be:

  • How do I (discretely) monitor temperature in the key rooms and get the data back to a place where actions can be taken? Wireless probes exist but there are batteries to change and we all know that wireless solutions aren’t infallible – I have so much wireless stuff in the house already I do wonder how much of the time it spends drowning out other wireless systems.
  • Will it be possible to control TRVs so that room temperatures can be controlled at different times of day – e.g. bedrooms cooler at night – or not?
  • How do I manage the control loop – turning on and off the boiler and pumps when needed? Actually this seems the simplest part – just run a simple control loop and pretend to be a thermostat I guess.
  • Is all of this worthwhile – i.e. will I save gas and money by doing this – or at least improve the temperature regulation in the house?

Possible Steps to a Solution

I think firstly I’m going to concentrate on the last question – is it worth it? I’ve spent a lot of time messing around with home automation and some of it has been distinctly unremarkable. I really don’t want to engage in this if there is a simple commercial solution or if it really isn’t going to improve things much.  So my idea right now is:

Step 1 : Implement a monitoring system to see just how much gas is being used heating each zone in the house.

Step 2 : Find a discrete way to monitor temperature in key rooms.

Step 3: Investigate TRVs some more – Danfoss OpenTRV I have found

Step 4: Investigate commercial systems if any can be found – most seem to use proprietary signalling and can’t be interfaced with. While this might not be a show-stopper I don’t want to go down a dead-end route either.