Heating a room nobody is in is the loss controls exist to stop

Controls are about time and place rather than temperature

The temperature you choose is one question, and it has a physics of its own. What controls decide is different: which parts of the building get heat, and during which hours. A house can be run at a perfectly modest setpoint and still waste a great deal by delivering that temperature everywhere, all day, including to rooms nobody enters.

That framing helps because the two questions have different answers. Lowering the setpoint reduces loss everywhere and is felt everywhere. Improving the controls removes heat only where it was doing nothing useful, which means the saving arrives without the household noticing a change in comfort. That is the more attractive kind of saving.

It also explains why control upgrades are oversold. In a small, well-occupied house there is little heat going anywhere useless, so the sophisticated system has almost nothing to work with. The saving belongs to the household’s pattern, not to the equipment.

A radiator valve is a throttle, not a switch

A thermostatic radiator valve contains a capsule of wax or liquid that expands as the room warms and pushes a pin down onto the valve seat, progressively restricting flow. It is a proportional device: it does not click off, it closes gradually as the room approaches the temperature the setting corresponds to.

The numbers on the head are not degrees, and that’s the source of most misuse. Turning a valve to its maximum does not heat the room faster; it simply raises the temperature at which the valve begins to close, so the room overshoots later rather than warming sooner. The rate is set by the radiator and the water temperature.

Nor does a valve turn off the heat source. Close every valve in the house and the boiler or heat pump still runs, circulating water that has nowhere to go, which is why a system needs a room thermostat or an equivalent control somewhere that can actually stop it. Valves shape distribution; something else has to end the call for heat.

Zoning splits the system so that parts of it can genuinely stop

A zone is a section of the system with its own control and motorised valve, able to be satisfied and shut while the rest continues. The classic split is upstairs and downstairs, which works because the two are occupied at different times and neither wants heating while the other is in use.

This is a real saving in the right building. A large house with an unused wing, or a household whose hours differ sharply between floors, has heat going somewhere that does not need it, and zoning is the mechanism that stops it.

It is also easy to overdo. Every zone adds a valve, a control, wiring and a further thing to go wrong, and past a point the extra zones separate rooms that were always used together anyway. The saving is in the rooms genuinely left cold, not in the number of zones.

Short cycling is the failure mode that over-zoning creates

Heat sources have a minimum output, and a boiler or heat pump modulating down to its floor still produces more heat than one small zone can absorb. So it fires, satisfies the zone in a few minutes, and stops. Then the zone cools and it fires again. This is short cycling, and it is worth recognising.

Every start costs something. Efficiency is lowest in the first minutes of operation, before the appliance has settled, and repeated starting wears the components that do the starting. A system cycling rapidly can consume more than the same system running steadily for a longer period at a lower output.

The practical consequence is that zoning and heat source sizing are the same decision. A system split into many small zones needs a heat source able to turn down a long way, or a buffer to absorb the excess. Fitting fine-grained zoning to an oversized appliance is how a control upgrade makes things worse.

Scheduling does more work than sophistication does

The single largest control saving in most households is simply not heating the house during the hours it is empty or everybody is asleep. A basic time clock does that, and it doesn’t need an app. A device with an app does it too, along with a great deal else, and the incremental saving from the great deal else is usually modest.

Some of the extras earn their place. Weather compensation, which lowers the water temperature as the outside gets milder, is genuinely valuable on a low-temperature system. Occupancy detection and location-based triggering suit irregular households and do very little for regular ones. Learning algorithms are pleasant and hard to attribute a saving to.

Which leads to the unfashionable recommendation. If the existing programmer works and the schedule reflects how the house is used, replacing it will not pay for itself, whatever the packaging suggests.

Most control failures are neglect rather than technology

Thermostatic valves seize when left in one position for years, usually stuck open, and the room they serve quietly overheats every winter afterwards. Exercising them occasionally through their full travel, particularly at the end of the heating season, keeps the pin free. It is a two-minute job per radiator and it is almost never done.

Sensor placement is the other recurring fault. A room thermostat above a radiator, in a draught, in direct sun, or in a hallway colder than everywhere else will govern the whole house from a reading that describes nowhere. And schedules outlive their circumstances, carrying on for years after the pattern that shaped them stopped. Reviewing the times, checking the valves and confirming the thermostat sits somewhere representative costs nothing and reliably beats buying another controller.