Data Properate calculates itself

Not every figure comes from a meter. Temperature correction, hourly values from accumulated readings, leak indications and baselines are calculated — and they behave differently from measured data.

Some of the most useful numbers in Properate were never measured by anything. They are derived from the data you already have.

This is worth understanding for one practical reason: a calculated value depends on its inputs. When a meter goes quiet, everything computed from it goes quiet too — and that can look like several unrelated faults at once when it is really one.

Making raw readings usable

Hourly values from accumulated meters. Many meters do not report consumption at all — they report a running total, the way a car's odometer does. Properate converts that into consumption per hour. It also handles meters that reset to zero, which would otherwise produce a large negative reading at the moment they roll over.

The energy hierarchy. Main meters and sub-meters are arranged so a building's total and its parts stay consistent, and can be grouped by energy source — electricity, district heating, solar. This is what allows you to drill from a total into the systems behind it without adding numbers by hand or double counting.

Cleaning. Obvious outliers are filtered so a single implausible reading does not distort a monthly figure or a graph's scale.

Making figures comparable

Temperature-corrected consumption, by degree-day method and by a model built from the building's own measured relationship between energy and outdoor temperature. See Weather and climate data for which method answers which question.

Operational energy label and per-square-metre figures, so buildings of different sizes can be compared.

Finding things you did not ask about

Leak indications. A water leak rarely looks dramatic. It looks like consumption that never returns to near zero overnight. Properate watches for exactly that pattern, which is hard to notice by eye and expensive to miss.

Indoor climate aggregates. Minimum, mean and maximum per room and per floor, so a floor can be judged at a glance instead of by opening every room in turn.

Anomaly monitoring. Behaviour that departs from a building's own established pattern, which catches problems nobody thought to write an alarm for.

Flexibility baseline. What a building would have consumed had it not been asked to flex. This is the reference that delivered flexibility is measured — and paid — against. See Energy flexing.

Calculations you build yourself

The same engine is available to you. Virtual sensors compute a new value from existing ones — an efficiency derived from two temperatures, a total that no single meter measures, a figure your organisation reports on but nobody installed an instrument for. Cloud automations take the result a step further and write it back as a setpoint.

Once a virtual sensor exists it behaves like any other value: you can graph it, alarm on it, place it on a technical schema or a dashboard.

Worth knowing

Calculated values are only as good as their inputs. If several derived figures stop at the same moment, look for a single source that went quiet rather than several separate faults.

Calculation problems do not raise building alarms. They appear as notifications on the calculation itself. If a derived value looks stale, the reliable check is when the calculation last ran.

Corrected and uncorrected figures are both right. They answer different questions, and most disagreements about "what did we actually use" come down to two people having different correction methods selected.