Energy Management & IoT

Energy consumption cannot be managed while unmeasured, and the worth of a measurement depends on where the meter sits: a total reading at the main board explains the bill and changes no behaviour. Analysers on riser feeds, floor boards and large consumers tie spending to whoever answers for it; only then can a deviation be traced to its device.

A power analyser does more than count kilowatt-hours: it measures current, voltage, power factor and harmonic distortion. Poor power factor lands on the bill as a reactive charge, while harmonic distortion heats the conductor needlessly and shortens transformer life. Because these quantities move before a fault appears, measurement stops being an accounting record and becomes early warning.

IoT sensors across the site complete the picture with quantities other than electricity: room temperature, relative humidity, how long a cold room door stays open, flow in a water main. What sets the life of a battery-powered device is how often it talks: a sensor reporting every reading at once is soon spent, one that speaks only on a change threshold stays in the field far longer. A light messaging arrangement such as MQTT carries those short packets over a weak link.

Baseline and deviation

Proof of a saving is not that consumption fell but that it fell under comparable conditions. A comparison drawn without outdoor temperature, occupancy and working calendar misleads: a mild winter produces a lower bill with no improvement behind it. A baseline comes first, and later periods are read against it.

Deviation reads most clearly when the building is empty. Consumption continuing on an empty floor usually signals not a fault but a forgotten time setting or a valve nobody closed by hand. Once the behaviour of the envelope and the mechanical plant shows in these records, the rule handed to the control side is written from measurement, not guesswork.