Most office buildings are designed to use a certain amount of energy, yet many consume noticeably more once they are occupied. The gap often comes from small mismatches: schedules that never changed after a tenant moved out, air handlers running at full flow for a floor that is half empty, or setpoints adjusted by hand during a complaint and never reset. Facility teams usually see only the monthly utility bill, which tells them something is wrong but not where.
A building digital twin addresses this by linking a 3D model to live data. Meter readings for electricity, chilled water and gas are combined with building management system points such as supply air temperature, damper positions, pump speeds and occupancy counts from access badges or Wi-Fi logs. The twin keeps the building's geometry, zoning and equipment list, so every reading can be placed in context: which zone, which air handler, which schedule.
Consider a six-storey office where cooling demand stays high on the north wing after 18:00. The twin can compare the measured load with the expected profile for each zone and highlight that two floors show almost no badge activity while their ventilation still runs at design flow. Operators can then test a revised schedule in the model first, estimating the effect on comfort and load before changing any real setpoint.
The same approach supports commissioning and fault detection. A simulated baseline shows what a healthy chiller plant should consume at a given outdoor temperature, so a persistent deviation can point to a fouled coil, a stuck valve or a sensor reading incorrectly. Because the model updates continuously, teams can check whether a change truly reduced consumption instead of relying on a single week of data.
Results depend on data quality and on people. Sensors need regular calibration, occupancy data must respect privacy rules, and comfort complaints matter more than any kilowatt-hour target. A practical start is one building, a few key meters and a clear question, such as which zones run outside their needed hours. Platforms such as TwinUp work with this kind of data, and their main value is giving facility managers a safe place to test changes before occupants feel them.