Heating is started early enough that comfort is achieved before occupancy, but not too early. A start is optimal if comfort is achieved at least energy cost.
This means an early start when it is cold, and a later start when it gets warmer, for two related reasons:
1) On a cold night the building cools faster to a lower temperature.
2) On a cold morning the rate of warming is slower, because losses are greater
 |
| On a cold night the optimum start is earlier |
Fortunately the mathematical calculations are done by a control system so an energy manager does not have to worry.
Unfortunately, many control systems are very bad at this "optimum search".
Unfortunately, Optimum Start controllers often fail.
The worst thing is that they fail expensively and when they fail it is hard to know - some error is expected every day, and you have to "be there to see it ".
Happily a remote energy auditing tool like kWIQly can show and even score the possibility of start failure on a continual basis...
 |
| Optimum Start Failure |
Optimum start is most important for older massive, airy, unventilated buildings (churches, castles) that respond slowly. Failure can easily add 20% to a fuel bill per annum.
Knowing it is working is important and early failure identification easy given the right tools.