Six call-outs that were not faults

The system is doing exactly what it was told. It was set back over the weekend and told to start at, say, six — and on a January morning in Colorado six is not early enough to bring a cold building up before the doors open.

The building then recovers by mid-morning, which is what makes it look intermittent and is why it gets reported as a fault three Mondays in a row before anybody looks at the controller.

• Look at the start time on the controller and compare it with the time people actually arrive.
• Check whether the start time is the same for Monday as for the other days. It very often is not, and it should be earlier on Monday, not later, because the building has been cold for two days.
• Check whether the setback is deeper at weekends than overnight. If it is, the Monday recovery is longer by design.

If the building is not recovering by mid-morning, or if the recovery is getting slower month by month with the same settings, that is a capacity or a performance problem and it is worth a visit. The give-away is the trend, not the Monday.

A holiday or exception schedule that was entered once, years ago, and is still in the controller — or one that was entered for last year’s dates and never updated. The system is off because somebody told it to be off.

The second most common version: the controller is in a hand-off-auto state that somebody left in off after a previous visit, or a service switch that was never put back.

• Look for an exception, holiday or override list in the controller, and look at the dates on it.
• Check whether the system will run when you force it on manually. If it will, this is a schedule question and not an equipment one.
• Ask whether anybody was in the plant room recently, including other trades.

If it will not run when forced on manually, stop and call. And if the building was unoccupied and unheated through a hard freeze, the priority changes entirely — the question stops being why it did not run and becomes what froze while it was off. That is worth a visit the same day.

Three ordinary causes. Optimum start, which deliberately brings the system on early enough to hit temperature by occupancy — so on the coldest night of the year it will start very early, and that is correct behaviour rather than a fault. Freeze protection, which overrides the schedule to stop coils or pipes freezing. Or a setback that is not actually set back, so the night setting and the day setting are close enough that it simply keeps running.

• Compare the occupied and unoccupied setpoints. If they are within a degree or two of each other there is no setback to speak of.
• Check whether the nights it ran were the coldest nights. If they map onto the weather, it is optimum start or freeze protection doing its job.
• Find the sensor that is driving the zone. A sensor above a fridge, in a server cupboard, in direct sun, or in a space nobody uses will keep a whole zone running for nothing — and this is the single most common cause of an inexplicable overnight bill.

If it runs all night in mild weather, with a genuine setback programmed and the sensor in a sensible place, something is calling that should not be. That is a visit, and it is usually a stuck valve, a failed sensor or a control fault — all real and all worth finding, because it is costing money every hour.

The other tenants prove the plant works, so something is different about that space or about what reaches it. In order of how often we find it: a partition built after the fit-out that blocked a diffuser or a return path; a space heater or a server cupboard added since; a sensor in the wrong place; and a flexible duct crushed when somebody re-laid a ceiling.

There is also a fifth, which nobody likes: the tenant is right and the space was never designed for what they now do in it. That is not a fault either, but it is a real problem and it needs a different conversation.

• Ask when it started, and what else happened around then. A fit-out two suites away is a common answer.
• Ask what has been added to the space — people, equipment, partitions, blinds taken down.
• Look at where the thermostat or sensor for that zone actually is.

If nothing has changed in the space and the complaint is new, it is worth measuring the airflow at that diffuser. A number settles it either way, and it is a short visit.

Controllers with a dead backup battery lose their schedule and their clock and restart on a default programme — which is usually 24-hour operation or none at all. Some equipment does not restart on its own by design, and waits for a manual reset. And breakers or motor protectors that tripped during the event stay tripped.

The power cut is rarely the fault. What it exposed usually was. A controller battery that failed two years ago causes no symptom at all until the day the power goes off.

• Check the time and date on the controller. If they are wrong, the schedule is almost certainly gone too.
• Walk the panel and look for anything tripped.
• Check whether the equipment is simply waiting for a reset rather than failed.

Anything that will not reset, anything that trips again after resetting, and anything that smells hot — stop there and call. A breaker that trips twice is not a nuisance, it is a message.

The controller did not follow the change, or followed it using an old rule. Everything then happens an hour early or an hour late — the building warms an hour after people arrive, or the system shuts down an hour before they leave, and both get reported as faults in the same week.

It is worth knowing this one exists, because it is the easiest thing on this page to diagnose and the easiest to pay somebody to diagnose for you.

• Look at the clock on the controller against the clock on your phone.
• Check whether the complaint is exactly an hour out. If it is, this is almost certainly it.
• Check every controller in the building, not just one. They do not all change together.

If the clock keeps drifting after being set, the controller has a battery or a board problem and that is worth attending to — not because the drift matters this week, but because the next power cut will lose everything.