An electricity bill tells a building owner how much energy was purchased, but rarely explains where it went. Submetering adds that missing detail by measuring selected circuits or systems separately. Heating, ventilation, lighting and charging equipment can then be compared on the same timeline. The purpose is to turn an unexplained total into a small number of questions that someone can investigate.

The first distinction is between power and energy. Kilowatts describe the rate of electricity use; kilowatt-hours describe the amount used over time. As a simple calculation, a device drawing a constant 2 kW for three hours consumes 6 kWh. A short, high peak and a modest load running all night therefore deserve different responses. Reducing one does not automatically solve the other.

Interval selection changes what the operator can discover. A 15-minute log produces 96 intervals per day, or 35040 over a 365-day year. That is sufficient for many scheduling investigations, although short peaks may disappear inside an average. For example, an 8 kW load operating for 5 minutes uses about 0.67 kWh. Spread across a 15-minute interval with no other consumption, that becomes an average of approximately 2.67 kW. These are calculated values, not readings from a particular building. Protective-device selection requires a different level of analysis from an energy dashboard.

Consider a worked example with a measured overnight load of 1.8 kW. Investigation identifies equipment that can legitimately be switched off, reducing that load to 0.6 kW for 12 hours on each of 250 comparable working days. The calculated reduction is (1.8 − 0.6) × 12 × 250 = 3600 kWh. That figure is only defensible if the equipment really follows the new schedule and the required service is preserved. It must not be applied indiscriminately to refrigeration, safety systems or equipment needed outside office hours. Weekend operation needs its own calculation rather than a convenient assumption.

Meter boundaries are equally important. If a distribution-board meter includes both ventilation and a tenant’s equipment, a fall in its total cannot automatically be assigned to ventilation controls. Before accepting the data, confirm circuit identification, current-transformer ratios where applicable, phase association and the direction of energy flow. Import and export must not be confused when photovoltaics are present. A useful commissioning exercise compares accumulated submeter energy against an appropriate upstream measurement over the same period. Differences may reflect unmetered loads or losses, but can also reveal configuration errors that would distort every later performance report.

Savings verification needs a baseline that reflects the service being delivered. Comparing a full office in January with a half-empty office in April mixes occupancy, weather and control changes. The U.S. Department of Energy’s measurement-and-verification guidance explicitly considers these influences when adjusting a baseline. For a practical project, document operating hours, relevant outdoor conditions, occupied area and any major equipment changes. Keep measured values separate from estimated replacements for missing data. Then report the observed energy change alongside the adjustment method and unresolved uncertainty. A defensible result is more useful than an impressive percentage whose origin nobody can reproduce.

A practical monitoring project starts with a clear question. Does the ventilation continue running after staff leave? Has a heater been left on in a rarely used room? Record operating hours alongside meter readings, and compare days with similar weather and occupancy. Check units, timestamps and gaps before interpreting a graph. A missing reading should never silently become zero consumption, because that can create the appearance of savings that did not occur.

For a small office, a useful first report could show weekday demand, weekend demand and the overnight baseline. Each unusual pattern should have an owner and a follow-up action. After changing a schedule, review comfort and consumption together. Measurement alone does not reduce energy use, but it helps the operator choose a specific change and see whether it worked. That makes metering a foundation for effective automation rather than just another screen of numbers.

Articles