In looking at what is available on the market, I have been trying to categorise the different types of controller and what they offer. To begin at the beginning, there was a boiler connected to radiators fitted with manual valves and a simple mechanical thermostat on the wall to turn the boiler on and off. Unless the user spent considerable effort adjusting radiators and the boiler this was very wasteful of energy.
Then the next level of sophistication is a timeswitch to override the thermostat when heating is not required but this still relied on manual control of the individual radiators if the user could be bothered.
Then comes the TRV, which allows different rooms to be controlled individually. This takes away a lot of the pain of trying to balance the radiators to stop one room overheating while another is still warming.
None of this is particularly ‘smart’, just mechanical switches and valves. So the next stage in to introduce electronic, wireless timeswitches that allow remote access from a smartphone or other device. These may even look up weather forecasts from the internet to modify the times and temperatures the boiler runs at. But the rooms themselves are still controlled by the mechanical TRVs. There are programmable eTRVs but these often talk neither to the controller nor the boiler. Some will detect an open window with an additional sensor or one built in, but can’t tell if the room is occupied or not.
So, to my mind a truly ‘smart’ controller needs to do three things:
- Detect the rooms that need heating and follow some heuristic to determine the appropriate temperature.
- Command the eTRV to set the radiator to the appropriate temperature.
- Control the boiler to provide sufficient heating for the demand from the radiators. (Can include weather compensation to keep boiler at optimum efficiency.)
These are all fine, but in practice the user will be forever walking into unheated rooms and have to wait while the radiator warms up. So we need either a time machine to anticipate where the user will be in half an hour’s time, or the system will need to learn patterns of occupancy and have a sensible setback temperature to stop frequently used rooms from becoming too cold.
Having established these criteria I shall endeavour to categorise various products. From what I have seen of the project so far, the OpenTRV system appears to be getting the closest to the ideal.